You are on page 1of 16

INTERNATIONAL JOURNAL OF ENERGY RESEARCH

Int. J. Energy Res. 2011; 35:1213–1228


Published online 20 July 2011 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/er.1889

REVIEW PAPER

A review on methanol crossover in direct methanol fuel


cells: challenges and achievements
Mahmoud Ahmed1,*,† and Ibrahim Dincer2
1
Department of Mechanical Engineering, Assiut University, Assiut 71516, Egypt
2
Faculty of Engineering and Applied Science, University of Ontario Institute of Technology (UOIT), 2000 Simcoe Street North, Oshawa,
Ontario L1H 7K4, Canada

SUMMARY
Direct methanol fuel cells have the potential to power future microelectronic and portable electronic devices because of
their high energy density. One of the major obstacles that currently prevent the widespread applications of direct methanol
fuel cells is the methanol crossover through the polymer-electrolyte membrane. Methanol crossover is closely related to
several factors including membrane structure and morphology, membrane thickness, and fuel cell operating conditions such
as temperature, pressure, and methanol feed concentration. This work presents a comprehensive overview of the state-of-
the-art technology for the most important factors, affecting methanol crossover in direct methanol fuel cells. In addition, the
current and future directions of the research and development activities, aiming to reduce the methanol crossover are
reviewed and discussed in order to improve the performance of direct methanol fuel cells. Copyright # 2011 John Wiley
& Sons, Ltd.

KEY WORDS
direct methanol fuel cells; methanol crossover; polymer-electrolyte membranes; methanol permeability

Correspondence
*Mahmoud Ahmed, Department of Mechanical Engineering, Assiut University, Assiut 71516, Egypt.

E-mail: aminism@aun.edu.eg

Received 2 March 2011; Revised 28 May 2011; Accepted 28 May 2011

1. INTRODUCTION Figure 1, consists of a flow field (including collector ribs


and channels), a diffusion layer (DL), and a catalyst layer
Fuel cells are nowadays one of the most promising clean en- (CL) on the anode and the cathode sides as well as a
ergy technologies. Direct methanol fuel cells (DMFCs) have polymer-electrolyte membrane (PEM). Reactants enter the
the potential to power future microelectronic and portable cell through the flow channels. The DL is typically made
electronic devices. The direct methanol fuel cells offer some of carbon cloth or paper, which not only tends to uniformly
advantages of effective operation at low temperatures, simple distribute the reactants over the surface of the catalyst layers,
design, and environmentally-benign nature. Moreover, meth- but provides an electrical connection between the CL and the
anol is easier to handle because of its liquid nature at room current collector as well. Electrochemical reactions occur in
temperature. It is of low cost and availability at industrial scale the catalyst layers, which are attached to both sides of the
as well as easy to store, and safe in use and delivery. More membrane. The catalyst layers must be designed in such a
importantly, unlike hydrogen polymer exchange membrane manner to facilitate the transport of protons, electrons, and
fuel cells, DMFCs based on methanol aqueous solution do reactants [4].
not need humidifying system and special thermal manage- Typically, the liquid feed DMFC is operated at a tem-
ment ancillary devices. In addition, the power and energy perature lower than 100 C. The methanol feed solution is
densities are superior, even when compared with indirect fuel forced to flow along the length of the flow channel and
cell and newly developed lithium-ion batteries. Possible to penetrate the DL to arrive at the anode CL, where the
applications of the DMFC include video cameras, electric methanol is electrochemically oxidized to form protons,
wheelchairs, portable-powered briefcases, notebook compu- electrons, and CO2. The produced CO2 in the CL, then
ters, intelligent transportation systems (road signs, traffic moves backward through the DL to the flow channel, and
lights, etc.), and military communications [1–3]. it is swept by the stream of liquid solution towards the exit
In general, a fuel cell works by converting chemical en- of the flow channel. The protons, formed at the anode CL,
ergy of a fuel into electrical energy. DMFC, as shown in are transported through ion-conducting polymer within the

Copyright # 2011 John Wiley & Sons, Ltd. 1213


M. Ahmed and I. Dincer Methanol crossover in direct methanol fuel cells

reduction in the methanol crossover to improve the DMFCs


performance are reviewed and discussed comprehensively,
and some potential solutions are also discussed.

2. CHALLENGES
Although considerable improvements in DMFCs design and
components have been made over the past years, numerous
challenging technical issues remain to be addressed before
the widespread commercialization of this type of fuel cell.
As previously reported, methanol crossover not only affects
the performance of DMFCs, but influences the volumetric
energy density of DMFCs as well. The transport of protons
takes place with water molecules, and for high-proton conduc-
tivity, PEM should be swollen enough with water, whereas it
should not for low methanol crossover. Methanol crossover
occurs by the absorption of methanol by membranes, diffusion
through membranes, and desorption from membranes. How-
ever, PEM, which satisfies both high-proton conductivity
and low-methanol crossover for DMFC, is not easily devel-
oped. Both proton conductivity and methanol permeability
increase with temperature. Concentration and rate of supplied
methanol affect methanol crossover. Increasing membrane
thickness reduces methanol crossover and increases the flow
resistance. Concentration and rate of supplied fuel/oxidant
Figure 1. Schematic representation of the direct methanol fuel affect methanol crossover [6,9]. These challenges and the
cell (DMFC). respective efforts carried out to reduce methanol crossover
in DMFCs are discussed in the current manuscript.

catalyst layers and the membrane to the cathode, where


they react with oxygen and electrons transported via the 3. METHANOL CROSSOVER
external circuit to form water. The anode, cathode, and
overall cell reactions can be written as follows [5]: When methanol comes into contact with the membrane, it
diffuses through the membrane from anode to cathode. It
Anode : CH3 OH þ H2 O ! 6Hþ þ 6e þ CO2 is dragged along with the hydrated protons under the influ-
Cathode : 1:5 O2 þ 6Hþ þ 6e ! 3H2 O ence of electric current flowing across the cell. Hydraulic
Overall : CH3 OH þ 1:5O2 ! 2H2 O þ CO2 pressure can also result in methanol transport. Therefore,
the methanol crossover through the membrane can be
The most pressing problem associated with DMFC is that driven by three transport mechanisms [10]: electro-osmotic
of fuel crossover because of permeate methanol together with drag by proton transport, diffusion by methanol concentra-
water from the anode to the cathode side. Methanol crossover tion gradient, and convection by the hydraulic pressure
during DMFC operation results in low power output because gradient between the anode and the cathode. It was
of chemical oxidation of methanol at the cathode with the help reported that the diffusive mode of methanol transport
of the cathode catalyst, causing (1) electrode depolarization, dominates when the cell is idle, whereas the electro-
(2) mixed potential, resulting in the open-circuit voltage osmotic drag dominates when the cell is operating [1].
(OCV) of the DMFC below 0.8V (3) consumption of O2, When the convective and electro-osmosis drag modes are
(4) cathode catalyst poisoning by CO (an intermediate of neglected, methanol crossover occurs only because of diffu-
methanol oxidation), and (5) serious water accumulation on sion by methanol concentration gradient. In this case, the
the cathode (water being produced by methanol oxidation), methanol crossover can be defined by using the simple form
which limits O2 access to cathode catalyst sites. Moreover, of Fick's law for diffusion across a membrane as follows:
the overall fuel utilization efficiency of the fuel cell is low-
ered when there is excessive methanol crossover [6–8]. dC
j ¼ D (1)
The objective of this review article is to present a compre- dz
hensive overview of the state-of-the-art technology for the
most important factors affecting methanol crossover in where j is equal to the methanol flux, D is the diffusion coef-
DMFCs. In addition, the current and future directions of ficient, C is the concentration of methanol, and z is the position
the research and development activities that seek the within the membrane. This equation can be integrated to yield.

1214 Int. J. Energy Res. 2011; 35:1213–1228 # 2011 John Wiley & Sons, Ltd.
DOI: 10.1002/er
Methanol crossover in direct methanol fuel cells M. Ahmed and I. Dincer

DH thickness, and the cell operating parameters, such as tem-


j¼ ΔC (2)
L perature and methanol feed concentration. In the next sec-
tion, the influences of these factors on methanol crossover
Here, H is the partition coefficient (or the ratio of con- are reviewed based on the previous published data and
centration in the membrane to that in solution). L is the their true interpretation.
thickness of the membrane, and ΔC is the concentration
difference between the solutions in contact with the mem-
brane. Methanol permeability (DH) is defined as the prod- 4. FACTORS AFFECTING METHANOL
uct of the diffusion coefficient and partition coefficient and CROSSOVER
takes into account both solubility and diffusion [11].
Direct methanol fuel cells are typically run under ambi- 4.1. Membrane material and morphology
ent conditions with a little or no pressure difference across
the cell so that the hydraulic permeation can be ignored. In Because the methanol crossover causes considerable cell
this case, the effect of convective transport through electro- voltage losses in the DMFC, different approaches to mini-
osmotic drag is considered in estimating methanol cross- mize or eliminate methanol crossover have been carried
over. Equation (1) can be modified to incorporate the elec- out such as: modification of Nafion membranes, using
tro-osmotic drag [1] as follows: new alternative types of proton-conducting membrane
materials, or exploring composite polymer materials.
dc Table I presents the development polymer-electrolyte
j ¼ D þcV (3) membranes for the DMFCs that have been reviewed and
dz
discussed in the present work. The reviewed polymer–
where V is the convective velocity introduced by the pas- electrolyte membranes are classified into four categories
sage of current and the effects of electro-osmotic drag. This such as modified Nafion membranes, copolymer membranes,
equation takes into account both diffusive and convective blend membranes, and composite membranes.
contributions. The value of V can be defined as follows:
Modified Nafion membranes. Nafion, a perfluorinated
ED i M sulfonic acid ionomer, has been the standard DMFC mem-

θF brane. It has a highly phase separated morphology that
impart excellent proton conductivity. On the contrary,
Here, ED is the electro-osmotic drag coefficient, i is the using Nafion in DMFC applications causes a relatively
current density, M is the methanol molar fraction at the in- high methanol crossover. Therefore, many modifications
terface between the anode CL, and the membrane, F is of Nafion have been attempted to reduce the methanol
Faraday's constant, and θ is the volume fraction of metha- crossover. These include: palladium (Pd) impregnated
nol within the membrane. [17], laminating with poly(ethylene glycol)/silica (PEG/
When all methanol crossover mechanisms are consid- SiO2) [18], polymerization [19–22], deposit Pd films on
ered [10], the methanol crossover per unit area can be writ- the surface by sputtering [23], incorporating nano-
ten as particles fillers into the polymers [24–26], casting
[27–33], silica coating [34], intercalation [35], composite
DHΔC C2 KΔP ED Mi membrane [36–40], blending membranes with tetraethoxy
j¼ þ þ (4)
L mL F orthosilicate (TEOS) and a variety of organic silane
[41,42], high temperature processing [43], partial substitu-
where ΔC represents the difference between methanol con- tion of the sulfonic acid groups [44,45], and electrochem-
centration at the interface between the anode CL and the ical modification [46,47]. Unfortunately, decreasing
membrane and between the membrane and the cathode methanol permeability was always accompanied by a loss
CL, respectively; ΔP is the difference in the liquid pressure in proton conductivity.
across the membrane. In addition, there are other several
mathematical formulas that are used in estimating the Copolymer membranes. A number of copolymer mem-
methanol crossover [12–16]. branes have been synthesized to reduce methanol crossover
Based on the above given equations (1– 4), the rate of and improve proton conductivity for DMFCs via partially
methanol crossover can be reduced by decreasing the dif- sulfonated [48–50], direct copolymerization [51–61], sul-
fusivity, increasing the membrane thickness or reducing fonation [62–66], post sulfonation [67,68], the aromatic
the methanol concentration at the anode CL. The methanol nucleophilic polycondensation [69–73], plasma polymeri-
crossover is closely related to the methanol permeability, zation [74], and direct free radical polymerization [75].
and cell performance is directly related methanol cross-
over. Although the exact relationship between permeability Blend membranes. Blends research on polymer blends
and methanol crossover is complex, several important fac- includes partially sulfonated polymer blend [79], poly
tors are controlling methanol crossover. These factors in- (vinyl alcohol) and (sulfonated poly ether ether ketone)
clude membrane material and morphology, membrane SPEEK [80–82], poly(vinyl alcohol) crosslinked with

Int. J. Energy Res. 2011; 35:1213–1228 # 2011 John Wiley & Sons, Ltd. 1215
DOI: 10.1002/er
M. Ahmed and I. Dincer Methanol crossover in direct methanol fuel cells

Table I. Developed polymer-electrolyte membranes for direct methanol fuel cells.

PEM Abbreviations Reference

Modified Nafion membranes


Pd nanophases in a Nafion polymer. Pd/ Nafion 17
Organic–inorganic hybrid-laminated Nafion 115. Nafion115/Hy 18
Nafion/polypyrrole composite. Nafion/P 19
Nafion/sulfonated phenol formaldehyde resin composite. sPFR/Nafion 20
Nafion/ crosslinked poly(2-Acrylamido-2-methyl-1-propanesulphonic acid). Nafion/AMPS 21
Nafion/polymerized dopamine nano-layer. Nafion 22
Pd-layered Nafion. Nafion117-Pd 23
Nafion/(3-mercaptopropyl) methyldimethoxysilane nanocomposite. Nafion/MPMDMS 24
Nafion/nano-size sulfated titanium dioxide solid superacid composite. Nafion/TiO2/SO42 25
Nafion/2-acrylamido-2-methylpropanesulfonic acid -modified montmorillonite. AMPS-MMT/Nafion. 26
Nafion–sulfonated organosilica nano-composite. Nafion/sDDS 27
Composite silica/Nafion. SiO2/Nafion 28
Polyimide/Nafion composite. PI/ Nafion 29
Nafion/acid functionalized zeolite beta nanocrystals nanocomposite. Nafion/AFB 30
Nafion/titanium dioxide Composite. Nafion/TiO2 31
Nafion/Nano-sized Fe2O3-SO24 solid superacid composite. Fe2O3/SO24/Nafion 32
Porous layered oxide/Nafion nanocomposite. Nafion/APO 33
Nano-silica layered Nafion composite. Nafion/silica 34
Nafion/organo-clay nanocomposite. Nafion/clay 35
Nafion/ternary composite materials. Nafion115/PI–PVA TSPS 36
Nafion/sol-gel-derived alkaline phosphate glass. STP/ Nafion 37
Nafion/zirconium meta-sulfonphenyl phosphonic acid-incorporated composite. Zr-msPPA/Nafion 38
Nafion/polyphenylene oxide with phosphomolybdic acid (PMA) composite. Nafion/PPO-PMA 39
Nafion/poly(diallyldimethylammonium)/PM multilayer films. PDDA/PM/ Nafion 40
Blended Nafion/Sulfonated poly(ether ether ketone)s. Nafion/SPEEK 41
Nafion/ORMOSIL hybrid. TEOS/DEDMS 42
Heat-treated Nafion. HTN 43
Nafion/sulfated b-cyclodextrin composite. Nafion/sb-CD 44
Blended Nafion/sulfonated poly(ether ether ketone). Nafion/SPEEK 45
Organic silica/Nafion composite. Nafion /SILCPM 46
Nafion/polyaniline. Nafion /PA/PY 47
Copolymer membranes
Partially sulfonated polystyrene block poly(ethylene -ran-butylene) SSEBS 48
block polystyrene copolymers.
Partially sulfonated block copolymers. SEBS 49
Disulfonated poly(arylene ether benzonitrile) Copolymers. 6FCN-35/BPSH-40 50
Sulfonated co-polyimide. Co-SPIs 51
Sulfonated polyimides. SPI 52
Sulfonated naphthalene dianhydride based polyimide copolymers. BAPS/ ODA 53
sulfonated poly(phthalazinone ether sulfone). SPPES 54
Sulfonated cardo poly(aryl ether sulfone). SPES-C 55
Sulfonated naphthalenic polyimides. SPI-K 56
Crosslinked sulfonated poly(arylene ether ketone) (SPAEKs). SPAEKs 57
Epoxy-based cross-linked sulfonated poly(arylene ether ketone). Cr-SPAEK 58
New comb-shaped poly(arylene ether sulfone) copolymers. S-PAES 59
Crosslinked sulfonated poly(arylene ether sulfone). C-ABPSH40 60
Sulfonated poly(arylene ether ketone) copolymers. C-SPAEK 61
Sulfonated polyethersulfone Cardo. SPES-C 62
A triblock copolymer ionomer, sulfonated poly(styrene-isobutylene-styrene). S-SIBS 63
Sulfonated polyimide*. NTDA/DAPPS/ODA 64
Sulfonated poly(styrene-b-butadiene-b-styrene) triblock copolymers. SBS (scSBS) 65
Sulfonated poly(ether ether ketone). SPEEK 66
Sulfonated poly(ether ether ketone) and crosslinked. SPEEK-C 67
Aromatic Poly(ether ketone)s with Pendant Sulfonic Acid Phenyl. Me-SPEEKDK& Ph-SPEEKDK 68

(Continues)

1216 Int. J. Energy Res. 2011; 35:1213–1228 # 2011 John Wiley & Sons, Ltd.
DOI: 10.1002/er
Methanol crossover in direct methanol fuel cells M. Ahmed and I. Dincer

Table I. (Continued)

PEM Abbreviations Reference

Polymerized polypyrrole on sulfonated poly(arylene ether ketone). SPAEK-C-(PPY/PWA)n 69


Sulfonated poly(arylene ether sulfone) copolymer. BPSH 70
Cross-linked sulfonated poly (arylene ether ketone). SPAEK-C/PVA 71
polyaniline on sulfonated poly(arylene etherketone) SPAEK-C (PANI/PWA)n 72
Cross-linked sulfonated poly(arylene ether ketone). C-SPAEK 73
Sulfonated plasma polymerized. SPPM 74
Zwitterionic crosslinked proton exchange. NaSS-4VP 75
Sulfonated poly(styrene). SPS 76
Sulfonated poly(arylene ether ether ketone ketone) copolymers. SPAEEKK 77
Sulfopropylated ethylene-vinyl alcohol copolymers. s-EVOHs 78
Blend Membranes
Partially sulfonated polymer blend. sPS/sPPO 79
Poly(vinyl alcohol) and its ionic blends with sodium alginate and chitosan. PVA/SA/CS 80
Poly(ether ether ketone)/Polyvinyl alcohol blend. SPEEK/PVA 81
A polyvinyl alcohol/p-sulfonate phenolic resin PVA/SP 82
Crosslinked poly(vinyl alcohol) containing sulfosuccinic acid. PVA/SSA 83
Novel covalent-ionically cross-linked. C-SPEEK 84
Polyelectrolyte complex (PEC). CS/(P(AA-AMPS) 85
Poly(2,6-dimethyl-1,4-phenylene oxide)-based acid–base polymer blend. PPO 86
Blends of ethylene-propylene-diene terpolymer (EPDM) and organophilized silicas. EPDM 87
Composite Membranes
Crosslinked sulfonated poly(ether ether-ketone) /silica hybrid. SPEEK/SiO2 88
Poly(vinyl alcohol) sulfosuccinic acid silica hybrid. PVA/SiO2 89
Poly(vinyl alcohol) with embedded phosphotungstic acid. PVA/PWA 90
Chitosan/titanate Nanotube Hybrid. CS/TNT 91
a-zirconium phosphate and silicotungstic acid. PVA/ZrP/SWA 92
Inorganic–organic hybrid–poly siloxane/poly (mal eicimide-co-styrene) network. AESA-Na 93
Poly(vinyl alcohol)/sulfated b-cyclodextrin. PVA/sulfated -cyclodextrin 94
Zeolite beta-filled chitosan. CS/Beta/SO3H 95
Poly(vinyl alcohol) with polyimide/ 8-trime thoxy- silyl pro pyl glycerine ether-1, PI-PVA-TSGEPS 96
3,6-pyrenetrisulfonic acid composite.
Organic-inorganic hybrid alkaline by epoxide ring. AEMs s 97
Polyvinyl alcohol pervaporation. PVA 98
Sulfonated poly(styrene-ran-ethylene) and porous polyimide matrix. PI/SPES 99
Self-cross-linkable sulfonated poly(ether ether ketone). SPEEK 100
Phosphorylated titanate nanotubes embedded nanocomposite. CS/PTNT 101
Solid superacid embedded chitosan hybrid. CS/STiO2 102
Organic–inorganic hybrid. CS/silica 103
Sulfonated poly(ether ether ketone) /tung stopho sphoric acid/ mesoporous SPEEK/TPA/MCM-41 104
materialcomposite.
sulfonated mesoporous benzene-silica incorporated poly(ether ether ketone) sPEEK-OMB 105
composite.
Poly(vinyl alcohol)-based hybrid Nanocomposite. PVA/SSA/sPBS 106
Sulfonated poly(ether ether ketone) sub-layer and polyvinyl alcohol sub-layer. SPEEK/ PVA 107
Functionalized carbon nanotube-poly(arylene sulfone) composite. PtRu/CNT-sPAS 108
Sulfonated polystyrene/polytetrafluoroethylene composite. SPS/PTFE 109
Dodecylbenzene sulfonic acid-doped/ polyethylene glycol/SiO hybrid. DBSA-PEG / SiO2 110
Chitosan-Poly (Vinyl Alcohol) and Calcium Oxide Composite. CPV 111
Macroporous polyimide composite. PAMPS 112
A cross-linked hybrid. SPAEK 113
Sulfonated and hydrogenated styrene/butadiene blocks copolymer PE-SHSBS 114
and plasma-treated microporous polyethylene.
Benzimidazole grafted poly(ether ether ketone and sulfonated poly(ether SPEEK-PEEK-BI 115
ether ketone) composite.
SPEEK/PVDF/PWA 116

(Continues)

Int. J. Energy Res. 2011; 35:1213–1228 # 2011 John Wiley & Sons, Ltd. 1217
DOI: 10.1002/er
M. Ahmed and I. Dincer Methanol crossover in direct methanol fuel cells

Table I. (Continued)

PEM Abbreviations Reference

Sulfonated poly(ether ether ketone), poly(vinylidene fluoride) and


phosphotungstic acid.
Nitrated sulfonated poly(ether ether ketone). Nitrated SPEEK 117
Sulphonated tetramethyl poly(ether ether ketone)/epoxy/sulphonated STMPEEK/TMBP/PNBS 118
phenol novolac semi-IPN.
*polymer
PEM, polymer-electrolyte membrane.

sulfosuccinic acid [83], and crosslinked (sulfonated poly zirconium meta-sulfonphenyl phosphonic acid (Zr-msPPA)
ether ether ketone) [84]. [38], and Nafion-infiltrated composite membrane [29]
reports conductivities of the same order of magnitude as
Composite membranes. A number of composite mem- Nafion 117 and significant reduction of methanol perme-
branes have been synthesized to reduce methanol crossover ability. These works are shown as (points #17, and #27)
and improve proton conductivity for DMFCs such as inor- in Figure 2(a). The reduction in methanol permeability is
ganic–organic material [88–97], composite with organic attributed to the nano crystalline layer-developed
[98–100], composite with inorganic [101–104], nanocom- interior of the Nafion 117 membrane [38]. The preceding
posite [105,106], and multilayer composite [107]. discussion highlights the range of transport properties of
modified Nafion membranes reported in previous literatures.
Figure 2(b) shows the variation of selectivity versus
4. 2. Transport properties of membranes proton conductivity of modified Nafion membranes for
DMFCs as listed in Table I. The most favorably industrial
An electrolyte with high proton conductivity and low region lies to the top right of Figure 2(b) (i.e., high proton
methanol crossover is the key element of direct methanol conductivity and selectivity). The selectivity parameter
fuel cells. To gain understanding of the methanol crossover determines the ratio of proton conductivity to methanol
process in PEMs, transport phenomena in membranes must permeability of a membrane and the higher selectivity
be considered. Therefore, in this section transport proper- value leads to a better membrane performance. Based on
ties related to methanol crossover such as transport of pro- the figure, the maximum selectivity value of 170104
ton (proton conductivity) and transport of methanol Sscm3 was achieved by both Nafion/zirconium meta-
(methanol permeability) are discussed. As previously sulfonphenyl phosphonic acid-incorporated [38], and
reported, the reviewed membranes are classified into four Nafion-infiltrated composite membranes [29] which is sig-
categories such as modified Nafion membranes, copolymer nificantly (43 folds) higher than that of 4.0104 Sscm-3 for
membranes, blend membranes, and composite membranes. Nafion117. It is followed by Nafion/2-acrylamido-2-
methylpropanesulfonic acid-modified mont-morillonite
membrane (AMPS-MMT/N) [26]. It was found that selec-
4. 2. 1. Modified Nafion membranes tivity value of about 90104 Sscm-3 is achieved at 3.0wt.%
Figure 2(a). shows the results of proton conductivity and of AMPS-MMT. In fact, such a remarkably improved se-
methanol permeability of modified Nafion membranes for lectivity parameter of Nafion matrices is owing to the ef-
DMFCs as listed in Table I. The most favorably industrial re- fective role of AMPS-modified nanolayers in restricting
gion lies to the top left of Figure 2(a) (i.e., high proton conduc- methanol permeation while maintaining essential proton
tivity and low methanol permeability). Based on the figure, conduction properties.
a significant achievement was observed in both proton con-
ductivity and methanol permeability. The proton conductivity
improved from 0.0032Scm-1 (point #1) for Pd nanophases 4. 2. 2. Copolymer membranes
in a Nafion polymer [28] to 0.117Scm-1 (point #30) for Figure 3(a) shows the results of proton conductivity and
Nafion/porous layered oxide nanocomposite [33], that is, methanol permeability of copolymer membranes for
increasing by about 36 times, which is remarkably higher DMFCs as documented in Table I. Based on the figure, a
than that of 0.026Scm-1 for Nafion 117. In addition, the meth- significant achievement was observed in both proton con-
anol permeability decreases from 7.2610-6 cm2 s-1 (point #7) ductivity and methanol permeability for different copolymer
for Nafion/nano-size sulfated titanium dioxide solid membranes. It was found that the proton conductivity
superacid composite [25] to 0.021710-6 cm2 s-1 (point #2) improved from 0.00025Scm-1 (point #4) for sulfonated
for organic–inorganic hybrid-laminated Nafion 115 [18], polyethersulfone Cardo [62] to 0.181Scm-1 (point #13)
that is, decreasing by about 333 times. A work on Nafion/Zirco- for sulfonated plasma polymerized [74], that is, increasing
nium meta-sulfonphenyl phosphonic acid-incorporated com- by about 724 times, which is significantly higher than that
posite membrane with different concentration percentage of of 0.026Scm-1 for Nafion117. In addition, the methanol

1218 Int. J. Energy Res. 2011; 35:1213–1228 # 2011 John Wiley & Sons, Ltd.
DOI: 10.1002/er
Methanol crossover in direct methanol fuel cells M. Ahmed and I. Dincer

Figure 2. (a) Variations of the measured proton conductivity ver- Figure 3. (a) Variations of the measured proton conductivity ver-
sus methanol permeability for modified Nafion membranes. (b) sus methanol permeability for copolymer membranes. (b)
Variation of selectivity (conductivity/permeability) versus proton Variation of selectivity (conductivity/permeability) versus proton
conductivity for modified Nafion membranes from the litera- conductivity for copolymer membranes from the literatures:
tures: ( ) Pd/Nafion [17], ( ) Nafion115/Hy [18], ( ) Nafion/ ( ) SSEBS [48], ( ) Co-SPIs [51], ( ) SPI [52], ( ) SPES-C
P [19], ( ) Nafion117-Pd [23], ( ) Nafion/sDDS [27], ( ) N/ [62], ( ) S-SIBS [63], ( ) SPS [76], ( ) SEBS [49], ( ) BAPS/
MPMDMS [24], ( ) TiO2/SO4 2[25], (X) SiO2/N [28], (+) sPFR/ ODA [53], ( ) 6FCN-35/BPSH-40 [50], ( ) NTDA/DAPPS/ODA
Nafion [20], () Nafion/silica [34], ( ) Nafion/PPO-PMA [39], [64], ( ) SPAEEKK [77], ( ) SPAEK-C-(PPY/PWA)n [69], ( )
( ) Nafion/AMPS [21], ( ) Nafion/clay [35], ( ) N/SPEEK SPPM [74], ( ) SBS (scSBS) [65], ( ) SPEEK [66], ( ) BPSH
[41], ( ) TEOS/DEDMS [42], ( ) HTN [43], ( ) PI/Nafion [29], [70], ( ) SPPES [54], ( ) SPEEK-C [67], ( ) NaSS-4VP [75],
(*) N/SILCPM [46], (^) Nafion/AFB [30], (&) Nafion [22], ( ) ( ) SPES-C [55], ( ) SPI-K [56], (a) SPAEKs [57], (C) SPAEK-C/
Nafion/TiO2 [31], ( ) Nafion/sb-CD [44], ( ) AMPS/MMT/N PVA [71], (@) SPAEK-C (PANI/PWA)n [72], (#) Cr-SPAEK [58],
[26], ( ) PDDA/PM/Nafion [40], ( ) STP/Nafion [37], ( ) ($) C-SPAEK [73], (+)s-EVOHs [78], (*) S-PAES [59], (&)Me-
Fe2O3/SO24/Nafion [32], ( ) Zr-msPPA/Nafion [38], (#) Nafion / SPEEKDK & Ph-SPEEKDK [68], (^)C-ABPS H40 [60], (X)
PA/PY[47], ($) N/SPEEK [45], (C) N/APO [33], (@) N115/PI–PVA– C-SPAEK [61].
TSPS [36].

permeability decreases from 6.4610-6 cm2 s-1 (point #19) thin polyaniline (PANI) and phosphotungstic acid (PWA)
for zwitterionic crosslinked proton exchange [75] to multilayers coated [72], Polymerized polypyrrole on sulfo-
0.000810-6 cm2 s-1 (point #3) for sulfonated polyimides nated poly(arylene ether ketone) by deposit polypyrrole
[52], that is, decreasing by about 8100 times. A work on (PPY) and phosphotungstic acid (PWA) layer by layer
polyaniline on sulfonated poly(arylene ether ketone) with [69], and Sulfonated Poly(arylene ether ether ketone ketone)

Int. J. Energy Res. 2011; 35:1213–1228 # 2011 John Wiley & Sons, Ltd. 1219
DOI: 10.1002/er
M. Ahmed and I. Dincer Methanol crossover in direct methanol fuel cells

copolymers with phthalazinone [77] reports conductivities


of the same order of magnitude as Nafion and a significant
reduction of methanol permeability. These works are
shown as (points #24, #12, and #11) in Figure 3(a). The re-
duction in methanol permeability is attributed to the thin
polyaniline, and phosphotungstic acid multilayers coated
[72], and deposit polypyrrole and phosphotungstic acid
layer by layer [69]. The above discussion presents the
available range of transport properties of copolymer mem-
branes reported in previous literatures.
Figure 3(b) shows the variation of selectivity versus
proton conductivity of copolymer membranes for DMFCs
as listed in Table I. Based on the figure, several copolymer
membranes achieved higher values of selectivity, the max-
imum selectivity value of 311104 Sscm3 was achieved
by Sulfonated polyimides with sulfonation level 16mol %
(SPI). However, the proton conductivity is one order of mag-
nitude less than that of Nafion 117 [52]. It is followed by
Polymerized polypyrrole on sulfonated poly (arylene ether
ketone) [69] with selectivity of 247104 Sscm3, and sul-
fonated plasma polymerized [74] with selectivity of 243
104 Sscm3. In addition, it was found that selectivity value
of about 202104 Sscm3 is achieved by polyaniline on
sulfonated poly(arylene ether ketone) [72].

4. 2. 3. Blends membranes
Figure 4(a). shows the results of proton conductivity
and methanol permeability of blends and composite blends
membranes for DMFCs as listed in Table I. Based on the
figure, a significant achievement was observed in both pro-
ton conductivity and methanol permeability. The proton
conductivity improved from 0.002Scm-1 (point#7) for
poly(ether ether ketone)/polyvinyl alcohol blend [81] to
0.11Scm-1 (point #1) for crosslinked poly(vinyl alcohol)
containing sulfosuccinic acid [83], that is, increasing by
about 55 times, which is remarkably higher than that of Figure 4. (a) Variations of the measured proton conductivity ver-
0.026Scm-1 for Nafion117. In addition, the methanol per- sus methanol permeability for blends membranes. (b) Variation
meability decreases from 11.7810-6 cm2s-1 (point # 7) of selectivity (conductivity/permeability) versus proton conduc-
for poly(ether ether ketone)/polyvinyl alcohol blend [81] tivity for blends membranes from the literatures: ( ) PVA/SSA
to 0.05810-6 cm2s-1 (point # 6) for Novel covalent- [83], ( ) sPS/sPPO [79], ( ) PVA/SA/CS [80], ( ) CS/(P(AA-
ionically cross-linked [84], that is, decreasing by about AMPS) [85], ( ) PPO [86], ( ) C-SPEEK [84], ( ) SPEEK/PVA
203 times. A work on Poly(2,6-dimethyl-1,4-phenylene [81], ( ) EPDM [87], ( ) PVA/SP [82].
oxide)-based acid–base polymer blend [86], and polyelec-
trolyte complex (PEC) [85], reports conductivities of the on the figure, the maximum selectivity value of 81104
same order of magnitude as Nafion and a significant reduc- Sscm3 was achieved by Poly (2,6-dimethyl-1,4-pheny-
tion of methanol permeability. These works are shown as lene oxide)-based acid–base polymer blend [86], which is
(points #5, and #4) in Figure 3(a). The reduction in meth- significantly higher than that of 4104 Sscm3 for
anol crossover is attributed to the solution of Nafion117. In fact, such a remarkably improved selectivity
bromomethylated-poly (2,6-dimethyl-1,4-phenylene oxide) parameter is owing to the effective role of (BrPPO) in
(BrPPO) [86]. The above reported discussion emphasizes restricting methanol permeation while maintaining essen-
the range of transport properties of modified Nafion mem- tial proton conduction properties.
branes reported in previous literatures.
Figure 4(b). shows the variation of selectivity versus 4. 2. 4. Composite membranes
proton conductivity of blends and composite blends mem- Figure 5(a) shows the results of proton conductivity and
branes for DMFCs as listed in Table I. The selectivity pa- methanol permeability of composite membranes for
rameter determines the ratio of proton conductivity to DMFCs as reported in Table I. Based on the figure, a sig-
methanol permeability of a membrane and the higher selec- nificant achievement was noticed in both proton conductiv-
tivity value leads to a better membrane performance. Based ity and methanol permeability. The proton conductivity

1220 Int. J. Energy Res. 2011; 35:1213–1228 # 2011 John Wiley & Sons, Ltd.
DOI: 10.1002/er
Methanol crossover in direct methanol fuel cells M. Ahmed and I. Dincer

for sulfonated poly (ether ether ketone) sub-layer and poly-


vinyl alcohol sub-layer [107] to 0.00910-6 cm2s-1 (point
#7) for Polyvinyl alcohol pervaporation [98], that is, de-
creasing by about 1333 times. A work on Sulfonated poly-
styrene/polytetrafluoroethylene composite (SPS/PTFE)
[109], Sulfonated poly(styrene-ran-ethylene) and porous
polyimide matrix (PI/SPES) [99], and Sulfonated and hy-
drogenated styrene/butadiene block copolymer and
plasma-treated microporous (PE-SHSBS) [114] reports
conductivities of the same order of magnitude as Nafion
and a significant reduction of methanol permeability.
These works are shown as (points #4, points #11, and
#24) in Figure 5(a). The reduction in methanol permeabil-
ity in (SPS/PTFE) is due to increasing DVB (a crosslinker)
content. Increasing crosslinking density results in the in-
crease of the diffusion resistance of methanol over water,
finally, resulting in the suppression of methanol perme-
ation or methanol crossover [109].
Figure 5(b) shows the variation of selectivity versus
proton conductivity of composite membranes for DMFCs
as listed in Table I. The selectivity parameter determines
the ratio of proton conductivity to methanol permeability
of a membrane. Based on the figure, the maximum selec-
tivity value of 73104 Sscm3 was achieved by sulfonated
poly(styrene-ran-ethylene) and porous polyimide matrix
[99], which is significantly higher than that of 4104
Sscm3 for Nafion 117. It is followed by sulfonated
polystyrene/polytetrafluoroethylene composite membrane
[109], where the selectivity value is about 62104
Sscm3. On the other hand, the ratio of styrene/divinylben-
zene (styrene/DVB) of 85/15 attains the lowest methanol
permeability. In fact, such a remarkably improved of selec-
tivity parameter is due to the sharp decrease in methanol
permeability with increasing crosslinking density. It was
found that crosslinking is restricting methanol permeation
Figure 5. (a) Variations of the measured proton conductivity ver- while maintaining essential proton conduction properties
sus methanol permeability for composite membranes. (b) Varia- [109].
tion of selectivity (conductivity/permeability) versus proton To conclude based on the above discussion, there is a
conductivity for composite membranes from the literatures: significant potential for the development of membrane
( ) SPEEK/SiO2 [88], ( ) PVA/SiO2 [89], ( ) CS/(P(AA-AMPS) architectures that can allow more independent control over
[108], ( ) SPS/PTFE [109], ( ) DBSA-PEG / SiO2 [110], ( ) the proton conductivity, and methanol permeability charac-
PVA/PWA[90], ( ) PVA [98], (X) CS/PTNT [101], ( ) CS/STiO2
teristics. Consequently, achieving of highly selective
[102], ( ) CS/silica [103], ( ) PI/SPES [99],( ) CS/TNT [91],
PEMs is very encouraging because PEMs with greater se-
( ) sPEEK-OMB [105], ( ) PVA/SSA/sPBS [106], ( ) SPEEK/
lectivity have potential for improved DMFC performance.
PVA [107], (+) CPV [111], () SPEEK/TPA/MCM-41[104], () PAMPS
It is worth mentioning that just because selectivity of a
[112], () PVA/ZrP/SWA [92], ( ) AESA-Na [93],( ) SPEEK [100],
( ) SPAEK [113], (@) PVA/sulfated b-cyclodextrin [94] ($) PE-
PEM is higher than Nafion, does not mean improved
SHSBS [114], (a) SPEEK-PEEK-BI [115],(C) SPEEK/PVDF/PWA
DMFC performance can be achieved. One limitation of se-
[116],(#) CS/Beta/SO3H [95], (&) Nitrated SPEEK [117], (^) lectivity as a gauge of DMFC performance is due to the
STMPEEK/TMBP/PNBS [118], (*) PI-PVA-TSGEPS [110], () fact that a minimum conductivity is still required for
AEMss [97]. DMFC operation, regardless of how low methanol perme-
ability is.

improved from 0.00013Scm-1 (point #7) for polyvinyl 4.3. Membrane thickness
alcohol pervaporation [98] to 0.25Scm-1 (point #19) for
macroporous polyimide composite [112], that is, increas- It was reported that the thickness of the Nafion membrane
ing by about 1920 times, which is significantly higher than is one of the most essential factors in determining the
that of 0.026 Scm-1 for Nafion117. In addition, the metha- methanol crossover rate. The methanol crossover decreases
nol permeability decreases from 1210-6 cm2s-1 (point # 15) with the increase of thickness, whereas increasing the

Int. J. Energy Res. 2011; 35:1213–1228 # 2011 John Wiley & Sons, Ltd. 1221
DOI: 10.1002/er
M. Ahmed and I. Dincer Methanol crossover in direct methanol fuel cells

thickness reduces the power density because of it high


ohmic loss [119]. However, this trend changes at the highest
current densities because membrane resistance to transport
phenomena became dominant. This fact is related to the re-
duction of resistance of charge transfer from the anode side
to the cathode side, and the reduction of concentration polar-
ization in the polymer with the decreasing of membrane
thickness. Thus, the electrical performance of fuel cell with
membranes of various thicknesses shows that the perfor-
mance trend is probably determined by the combined effects
of proton conductivity and fuel crossover. Something similar
occurs with the influence of the equivalent weight in the
performance of the cell. But at highest density currents, the
electrical performance of a DMFC may be decrease when
the equivalent weight of the membrane decreases [120].
It was found for composite Nafion-based membranes
that both the conductance and the methanol permeation
rate show a significant decrease with increasing thickness.
Figure 6. Variation of Methanol permeability versus Methanol
The optimization of these parameters regarding favorable
concentration (M) for different types of membranes reported in
properties is difficult. Moreover, for thinner membranes,
the literatures: ( ) SPEEK/SPVA [123] ,( ) SMMT/SPEEK
the permeation rate changes by a factor of four to five with [124], ( ) SPEEK/ PVA [107], ( ) PAMPS[112], ( ) Nafion/
a temperature change from 25 to 65 C, whereas in the hydroxyapatite composite [125], ( ) CS/Silica [103], ( ) Nafion
thicker membranes this factor is about three. At lower [22], ( ) PVA/sulfated b-cyclodextrin [94] ,( ) CS/Beta/SO3H
temperatures, the variation of the methanol flux through [95],( ) Nafion 115 [94], ( ) Nafion117 [95], ( ) Co-SPIs
the membranes versus thickness is small compared with [51], ( ) NTDA–BAPBDS/DABI [126], ( ) PVA/PSSA–MA [127].
that at 65 C. Also, a strong decrease of permeation with
thickness is observed [121]. It was found that the methanol
crossover limiting currents significantly decreases with the methylpropanesulfonic acid polymer (PAMPS) was reduced
membrane thickness for disulfonated polyarylene ether because of its effect on the size and distribution of hydro-
benzonitrile copolymers and Nafion at 80 C [122]. philic channels responsible for methanol permeation. Also,
the permeability of Nafion/hydroxylapatite composite mem-
4.4. Methanol concentration brane [125], and chitosan/silica [103] decrease by the in-
crease of methanol concentration. Work on poly (vinyl
In fact, a higher methanol concentration is desirable in alcohol)/sulfated b-cyclodextrin [94], and chitosan/Beta/
DMFCs because of higher energy density. Therefore, SO3H [95] indicated that a reduction of methanol cross-
many researchers are investigated the influence of metha- over was observed, as methanol concentration increased.
nol concentration on methanol crossover to the cathode This is attributed to the excellent methanol resistance of
side. Figure 6 shows the variation of methanol permeabil- hydroxyls in poly(vinyl alcohol) [94]. The incorporated
ity versus methanol concentration (M) for different types zeolite beta particles can extend the diffusion path length
of membranes as reported in previous literatures. It was of methanol in the membrane, rigidify the chitosan chains,
found that the methanol permeability of sulfonated poly and compress the volumes among chitosan chains. These
ether ether ketone [123], sulfonated poly ether ether ketone reasons caused the decrease in methanol crossover of
(SPEEK55) [107], Nafion [22], Nafion 115 [94], and chitosan/Beta hybrid membranes [95].
Nafion117 [95] increases with the increase of methanol A work on 1,4,5,8-naphthalenetetracarboxylic dianhy-
concentration. However, the methanol permeability of sul- dride (NTDA), based sulfonated polyimides (SPIs) [126]
fated poly (vinyl alcohol) decreased as the concentration of shows the methanol permeability decreases with increasing
methanol increases. Also, adding sulfated poly (vinyl alco- methanol concentration. Also, the methanol permeability
hol) to sulfonated poly(ether ether ketone) membranes can of poly(vinyl alcohol)/poly(styrene sulfonic acid-co-maleic
suppress the methanol crossover especially at high metha- acid) (PVA/PSSA–MA) membrane [127] is at least seven
nol concentration [123,107]. The methanol permeability times lower than those of Nafion 117 in all methanol
decreases with increasing methanol concentration of the concentration ranges. In addition, the increase of methanol
acid treated sulfonated montmorillonite / sulfonated poly concentration leads to a slight decrease of methanol perme-
(ether ether ketone) nanocomposite membranes [124]. ability through PVA/PSSA–MA membrane. On the other
A work on three-dimensionally ordered macroporous hand, in the work on sulfonated co-polyimides (co-SPIs),
(3DOM) polyimide matrix composite [112] indicated that the methanol permeability is hardly depended on methanol
the methanol permeability decreased with increasing meth- concentration [51].
anol concentration. It was found that by applying the 3DOM In summary, the methanol permeability increases with in-
matrix, the methanol permeability of 2-acrylamido-2- creasing the methanol concentration for SPEEK [123], and

1222 Int. J. Energy Res. 2011; 35:1213–1228 # 2011 John Wiley & Sons, Ltd.
DOI: 10.1002/er
Methanol crossover in direct methanol fuel cells M. Ahmed and I. Dincer

Nafion [22,94,95] membranes. However, a significant is because the swelling of the filling polymer electrolyte
decrease in methanol crossover was reported by increasing of the PF membrane was prevented, and the solvent con-
the methanol concentration for SPEEK/SPVA [123], tent of the PF membrane was kept lower than that of
SMMT/SPEEK [124], SPEEK/ PVA [107], PAMPS Nafion117 membrane [129]. A work on Pd and Pd–Cu
[112], Nafion/ hydroxyapatite composite membrane alloy deposited Nafion membranes indicated that the
[125], CS/Silica [103], PVA/sulfated b-cyclodextrin [94], methanol crossover current is significantly greater for
CS/Beta/SO3H [95], NTDA–BAPBDS/DABI [126], and unmodified membrane (point #4), and increases with in-
PVA/PSSA–MA [127]. No significant change of methanol creasing methanol concentration. The difference in
permeability by increasing methanol concentration for methanol crossover current density is more pronounced
Co-SPIs [51] was observed. between the Pd-sputtered and bare Nafion membranes
Figure 7 shows the variation of methanol crossover with increasing concentration. This is indicating the
limiting current density (mAcm-2) versus methanol con- effective suppression of methanol crossover by the Pd-
centration (M) for different types of membranes. For the sputtered membrane for higher methanol concentrations.
Pd/Nafion 117 membranes, the methanol crossover limit- The possible mechanism leading to the lower methanol
ing current results at 20 C (point # 1), and 60 C (point crossover rate in the Pd and Pd–Cu alloy sputtered
#2) with 1, 2, and 5M methanol solution are shown in membranes is attributed to the effect of the coated Pd
the figure . It was found that methanol crossover increased or Pd–Cu alloy, which is physically blocked by the pas-
with the increase of methanol concentration because of sage of methanol molecules [130].
higher concentration gradient. In addition, the value of For Nafion 112 and 1135 membranes (point #5) at
methanol crossover limiting current density increased with 60 C, the crossover current density is approximately line-
increasing cell temperature. This is caused by a severe arly proportional to the methanol concentration, with the
membrane swelling with the higher temperature methanol Nafion 112 membrane featuring higher crossover rate.
solution. It was noticed that methanol crossover was sup- The difference in crossover current density between the
pressed by incorporating of Pd into the membrane, and
two membranes diminishes with methanol concentration;
lower crossover was seen with a higher Pd loading mem-
for example, the difference decreases from 32mAcm-2 at
brane [128]. More importantly, the methanol crossover of
2M to11mAcm-2 at 4M. Thicker membranes have higher
Nafion 117 membrane with 1 and 5M methanol was signif-
cell internal resistance but lower methanol crossover.
icantly reduced by impregnating Pd.
Therefore, the cell using Nafion 1135 has the best electro-
A work on a pore-filling (PF) polymer electrolyte mem-
chemical performance, and its power density is insensitive
brane (point # 3) shows that the methanol crossover
to the anode stoichiometry [131].
through Nafion117 membrane increased with methanol
The methanol crossover currents density at 25 C (point
concentration, whereas methanol crossover through PF
#6), and 80 C (point #7), measured with Nafion/silica
membrane saturated at high methanol concentration. This
hybrid membranes as a function of methanol concentra-
tion are shown in the figure. It was found that the methanol
crossover current increases with increasing methanol con-
centration [132]. The methanol crossover rate for celtec-V
was measured at different methanol feed concentrations
and cell temperatures at 60 C (point #8), and 90 C (point
#9). Methanol permeation increases in an approximately
linear correlation with the methanol concentration. Also,
methanol crossover increases with the increase of the
temperature. This is of utmost importance from a perfor-
mance point of view. Furthermore, a lower crossover rate
is associated with higher fuel efficiency for the same con-
centration or a gain in volumetric energy density of the
fuel. Comparing the methanol crossover with Nafion,
Nafion shows significantly large values [133].
The relationship between limiting current density and
methanol concentration can be estimated from the
 
following equation JI ¼ 6FD L Cin . Where JI is a steady
state limiting current density; Cin is the methanol concen-
tration at the feed edge. This equation was developed by
Figure 7. Variation of Methanol crossover current versus Meth- assuming that the methanol crossover is only because of
anol concentration (M) for different types of membranes diffusion [10]. Based on the above equation, the variation
reported in the literatures: ( – ) Pd/Nafion [128], ( ) pore- of limiting current density is linearly increased with the
filling membranes [129], ( ) Pd-Cu[130], ( ) Nafion [131], methanol concentration, if the diffusion coefficient is con-
( – ) Nafion/Silica Hybrid [132], ( - ) PBI/PVPA [133]. stant. However, increasing the temperature leads to an

Int. J. Energy Res. 2011; 35:1213–1228 # 2011 John Wiley & Sons, Ltd. 1223
DOI: 10.1002/er
M. Ahmed and I. Dincer Methanol crossover in direct methanol fuel cells

increase of diffusion coefficient. This will result in increasing Nafion 117, and Nafion 112 membranes over the tempera-
the limiting current density as the temperature increases. ture range 25–80 C. The different methanol permeation of
these membranes and Nafion membranes can be explained
by the difference in their microstructures.
4.5. Fuel cell temperature
To conclude, the methanol permeability increases with
Figure 8 shows the methanol permeability as a function of the increase of temperature, and an Arrhenius-type depen-
temperature for different types of membranes reported in dency of methanol permeability on temperature exists for
the literatures: Nafion 117 and Phosphonated polypho- all these membranes. The methanol permeability of SPES-
sphazene [134], sulfonated poly(arylene ether sulfone) co- C, SPS, S-SIBS, SPEEK, SPPES, and SPEEK/ PVDF /
polymer (BPSH) [70], sulfonated polyethersulfone Cardo PWA is much less than that of Nafion.
(SPES-C) [62], Sulfonated poly(ether ether ketone)
(SPEEK) [66], sulfonated poly(styrene) (SPS) [76], sulfo-
nated poly(styrene-isobutylene-styrene) (S-SIBS) [63], 5. ACHIEVEMENTS
sulfonated co-polyimide (Co-SPIs) [51], Nafion112 [51],
partially fluorinated ionomeric [135], sulfonated poly(ary- Although great improvements in fuel cell design and com-
lene ether sulfone)–silica nanocomposite [136], sulfonated ponents have been made over the past years, several issues
poly(phthalazinone ether sulfone) (SPPES) [54], and sulfo- remain to be addressed before PEM fuel cells can become
nated poly(ether ether ketone)/poly(vinylidene fluoride)/ competitive enough to be used commercially. Polymer-
phosphotungstic acid (SPEEK/PVDF/PWA) [116]. Based electrolyte membranes in DMFC should transport protons
on the figure, the methanol permeability of all membranes as an electrolyte and prevent the methanol crossover. To
increases with the increase of temperature. In addition, an reduce methanol crossover, different approaches are fol-
Arrhenius-type dependency of methanol permeability on lowed. The most important approaches considered in this
temperature exists for all these membranes. It is worth not- manuscript are (1) to modify the Nafion membrane, (2) to
ing that the methanol permeability of the SPES-C [62], develop alternative membranes, or (3) exploring composite
SPS[76], S-SIBS[63], membranes are considerably smaller polymer materials.
than that of both Nafion 117, and Nafion 112 membrane
over the temperature range measured. We also observed 1. A work on modified Nafion membranes such as Zir-
the similar results of methanol permeability in SPEEK conium meta-sulfonphenyl phosphonic acid (Zr-
[66], SPPES [54], and SPEEK/PVDF/PWA membranes msPPA)/Nafion composite membranes showed a
[116]. It is important to note that the methanol permeability considerable reduction of methanol permeability
of these membranes is considerably smaller than that of with increasing Zr-msPPA content. This is because
of the Zr-msPPA nano conductors acted as crystal-
line barriers to methanol permeation [137]. However,
the proton conductivity also decreased with increas-
ing Zr-msPPA content, but its effect was not as con-
siderable as with methanol permeation because of the
inherent, high conductivity of Zr-msPPA. The meth-
anol permeability reduced by two orders of magni-
tude (48 times) compared with that of Nafion117,
whereas the proton conductivity is at the same order
of magnitude as Nafion117 [38]. Moreover, the poly-
imide/Nafion (PI/Nafion) composite membrane has a
very low methanol crossover compared with a
Nafion 112 membrane because of the fact that the
composite membrane mechanically prevents Nafion
from swelling and consequently reducing the metha-
nol crossover. The methanol permeability is eighty
times lower than that of Nafion 112. In addition,
the proton conductivity is the same as Nafion 112.
It was reported that there is a significant improve of
Figure 8. Variation of methanol permeability versus cell temper- the performance of a DMFC with PI/Nafion compos-
ature for different types of membranes reported in the litera- ite membrane because of the use of a very thin mem-
tures: ( ) Nafion 117 [134] ,( ) Phosphonated brane and a high methanol concentration [97].
polyphosphazene [134], ( ) BPSH [70], ( ) SPES-C [62], 2. A work on alternative membranes such as polymer-
( ) SPEEK [66], ( ) SPS [76], ( ) S-SIBS [63], ( ) Co-SPIs ized polypyrrole on sulfonated poly (arylene ether
[51], ( ) Nafion 112 [51], ( ) Partially Fluorinated Ionomeric ketone) SPAEK-C-(PPY/PWA)n [20] shows a con-
[135], ( ) Sulfonated oly(arylene ether sulfone)–silica nanocom- siderable decrease of methanol permeability with in-
posite [136], ( ) SPPES[54], ( ) SPEEK/PVDF/PWA [116]. creasing the number of bilayers compared with the

1224 Int. J. Energy Res. 2011; 35:1213–1228 # 2011 John Wiley & Sons, Ltd.
DOI: 10.1002/er
Methanol crossover in direct methanol fuel cells M. Ahmed and I. Dincer

Nafion 117. It is reported that the SPAEK-C mem- Table II. Transport properties and selectivity of the modified
branes having two regions hydrophilic one, and hy- polymer-electrolyte membranes for direct methanol fuel cells.
drophobic one. Methanol diffuses primarily through PEM Relative Relative Relative Reference
the hydrophilic water-rich regions. When the surface conductivity permeability selectivity #
of SPAEK-C is coated by the polypyrrole (PPY) and
phosphotungstic acid (PWA) films, the electrostatic (Zr-msPPA)/ 0.984–0.45 0.08–0.02 12.3–22.5 38
interactions between the polyacid/poly base pairs Nafion
block the methanol transport pathways [20]. However, (PI/Nafion)* 0.95 0.012 78 97
increasing the number of bilayers of (PPY/PWA) SPAEK-C- 0.91–0.24 0.0313– 28.8–32.6 69
shows a slightly negative effect on proton conductiv- (PPY/PWA)2–10 0.0073
SPPM 1.17 0.0358 32.5 74
ity. The methanol permeability reduced by three
SPAEK-C- 1.224 0.046 26.6 72
orders of magnitude compared with that of Nafion
(PANI/PWA)5
117, whereas, the proton conductivity is the same or-
PI/SPES** 1.47 0.05 29 99
der of magnitude as the Nafion 117. Similarly, a
PPO 0.77 0.047 16.3 86
work on polyaniline on sulfonated poly(arylene ether
ketone) SPAEK-C (PANI/PWA)n[ 72] shows a low *Reference values of Nafion 112.
**Reference values of Nafion 1135.
methanol permeability because of close pores by PEM, polymer-electrolyte membrane.
forming of polyelectrolyte multilayered films on the For the list of the definitions for the aforementioned items, please
substrate surface. The methanol permeability is two refer to Table 1.
orders of magnitude less than that of Nafion 117, and
proton conductivity is higher than that of Nafion 117.

In addition, a work on sulfonated plasma polymerized • The up to date modified membranes with the highest
(SPPM)[74] shows lower permeability of plasma mem- selectivity are as follows: (i) the modified Nafion
branes because of highly crosslinked structure of the membranes such as (Zr-msPPA)/Nafion, and (PI/
plasma polymerized membranes and their water binding Nafion); (ii) the developed alternative membranes
characteristics, which could effectively restrict methanol such as SPAEK-C-(PPY/PWA)n, SPPM, SPAEK-C-
diffusion. The methanol permeability is about 28 times less (PANI/PWA)n, PI/SPES and PPO.
than that of Nafion 117, and the proton conductivity is • The methanol permeability increases with increasing
slightly higher than that of Nafion 117. A work on Sulfo- the methanol concentration for SPEEK and Nafion
nated poly(styrene-ran-ethylene) and porous polyimide membranes. However, a significant decrease in meth-
matrix (PI/SPES) composite shows a significant reduction anol crossover was reported by increasing the metha-
of methanol permeability because of pore-filling of the nol concentration for SPEEK/SPVA, SMMT/SPEEK,
porous PI matrix with SPSE polymer. It results in sup- SPEEK/ PVA, PAMPS, Nafion/HA, CS/Silica, PVA/
pressed water absorption, and less dimensional change sulfated b-cyclodextrin, CS/Beta/SO3H, NTDA–
may indicate less methanol crossover though the mem- BAPBDS/DABI, and PVA/PSSA–MA. No signifi-
brane [99]. A work on poly(2, 6-dimethyl-1,4-phenylene cant change of methanol permeability was observed
oxide)-based acid–base polymer blend shows a consider- by increasing methanol concentration for Co-SPIs.
able reduction of methanol permeability because of the ef- • For all reviewed membranes, increasing the methanol
fective role of (BrPPO) in restricting methanol permeation concentration results in an increasing of the methanol
while maintaining essential proton conduction properties. crossover current density.
The methanol permeability is 21 times less than that of • The methanol permeability increases with the increase
Nafion 117 and the proton conductivity is in the same or- of temperature, and an Arrhenius-type dependency of
der of magnitude as Nafion 117 [86]. methanol permeability on temperature exists for all of
In summary, the relative transport properties with these membranes. The methanol permeability of
respect to Nafion 117 such as conductivity, methanol per- SPES-C, SPS, S-SIBS, SPEEK, SPPES, and SPEEK/
meability, and selectivity of the best modified membranes PVDF /PWA are much less than that of Nafion.
are shown in Table II. The modified Nafion membranes • The methanol crossover decreases with the increase of
are (Zr-msPPA)/Nafion, and (PI/Nafion). The developed membrane thickness.
alternative membranes are SPAEK-C-(PPY/PWA)n,
SPPM, SPAEK-C-(PANI/PWA)n, PI/SPES, and PPO.
REFERENCES
6. CONCLUSIONS
1. Zhao TS, Kreuer KD, Nguyen TV. Advances in Fuel
In this review paper, the development of new PEMs for the Cells, Vol. 1. Elsevier Ltd, 2007.
DMFC has been reviewed and the following important 2. Srinivasan S. Fuel Cells: From Fundamentals to
points are extracted. Applications. Springer, 2006.

Int. J. Energy Res. 2011; 35:1213–1228 # 2011 John Wiley & Sons, Ltd. 1225
DOI: 10.1002/er
M. Ahmed and I. Dincer Methanol crossover in direct methanol fuel cells

3. Kamarudin SK, Daud WRW, Ho SL, Hasran UA. 27. Lia C, Sun G, Ren S, Liu J, Wang Q, Wu Z, Sun H,
Journal of Power Sources 2007; 163:743–754. Jin W. Journal of Membrane Science 2006; 272:
4. Deluca N, Elabd YA. Journal of Polymer Science: 50–57.
Part B: Polymer Physics 2006; 44:2201–2225. 28. Jiang R, Kunz HR, Fenton JM. Journal of Membrane
5. Ahmad H, Kamarudin SK, Hasran UA, Daud WRW. Science 2006; 272:116–124.
International Journal of Hydrogen Energy 2010; 29. Nguyen T, Wang X. Journal of Power Sources 2010;
35:2160–2175. 195:1024–1030.
6. Zaidi SM, Matsuura T. Polymer Membranes for Fuel 30. Holmberg BA, Wang X, Yan Y. Journal of Mem-
Cells. Springer Science: NY, USA, 2009. brane Science 2008; 320:86–92.
7. Li N, Fane A, Ho WS, Matsuura T. Advanced Mem- 31. Barbora L, Acharya S, Verma A. Macromolecular
brane Technology and Applications. John Wiley & Symposia 2009; 277:177–189.
Sons, Inc.: Hoboken, New Jersey, USA, 2008. 32. Sun L, Wang S, Jin W, Hou H, Jiang L, Sun G. Inter-
8. Zhao TS, Xu C, Chen R, Yang WW. Progress in national Journal of Hydrogen Energy 2010;
Energy and Combustion Science 2009; 35:275–292. 35:12461–12468.
9. Kim H, Cho J, Yoon J, Chang H. 2001 ECS Confer- 33. Hudiono Y, Choi S, Shu S, Koros WJ, Tsapatsis M,
ence 2001. Nair S. Microporous and Mesoporous Materials
10. Han J, Liu H. Journal of Power Sources 2007; 2009; 118:427–434.
164:166–173. 34. Kim D, Scibioh MA, Kwak S, Oh I-H, Ha HY.
11. Cussler EL. Diffusion Mass Transfer in Fluid Sys- Electrochemistry Communications 2004; 6:
tems (3rd Edition edn). Cambridge University Press: 1069–1074.
New York, U.S.A, 2007. 35. Song M-K, Park S-B, Kim Y-T, Rhee H-W.
12. Wang ZH, Wang CY. Journal of the Electrochemical Molecular Crystals and Liquid Crystals 2003; 407:
Society 2003; 150–4:A508–A519. 15/[411]–23/[419].
13. Liu W, Wang CY. Journal of the Electrochemical 36. Fang Y, Wang T, Miao R, Tang L, Wang X. Electro-
Society 2007; 154–3:B352–B361. chimica Acta 2010; 55:2404–2408.
14. Chiu YJ. International Journal of Hydrogen Energy 37. Wang Y-J, Colby RH, Kim D. Journal of Membrane
2010; 35:6418–6430. Science 2011; 366:421–426.
15. Ye D, Zhu X, Liao Q, Li J, Fu Q. Journal of Power 38. Hwang M, Ha H-Y, Kim. Journal of Membrane Sci-
Sources 2009; 192:502–514. ence 2008; 325:647–652.
16. Eccarius S, Garcia B, Hebling C, Weidner JW. Jour- 39. Sauk J, Byun J, Kim H. Journal of Power Sources
nal of Power Sources 2008; 179:723–733. 2005; 143:136–141.
17. Kim YM, Park KW, Choi JH, Park IS, Sung YE. 40. Xiang Y, Zhang J, Liu Y, Guo Z, Lu S. Journal of
Electrochemistry Communications 2003; 5:571–574. Membrane Science 2011; 367:325–331.
18. Lin CW, Fan KC, Thangamuthu R. Journal of Mem- 41. Tsai J-C, Cheng H-P, Kuo J-F, Huang Y-H,
brane Science 2006; 278:437–446. Chen C-Y. Journal of Power Sources 2009; 189:
19. Park HS, Kim YJ, Hong WH, Lee HK. Journal of 958–965.
Membrane Science 2006; 272:28–36. 42. Kim YJ, Choi WC, Woo SI, Hong WH. Journal of
20. Wu Z, Sun G, Jin W, Wang Q, Hou H, Chan KY, Xin Membrane Science 2004; 238:213–222.
Q. Journal of Power Sources 2007; 167:309–314. 43. Ramya K, Dhathathreyan KS. Journal of Membrane
21. Choa KY, Jung HY, Shin SS, Choi NS, Sung SJ, Science 2008; 311:121–127.
Park JK, Choi JH, Park KW, Sung YE. Electrochi- 44. Jeon J-D, Kwak S-Y. Journal of Power Sources
mica Acta 2004; 50:589–593. 2008; 185:49–54.
22. Wang J, Xiao L, Zhao Y, Wu H, Jiang Z, Hou W. 45. Tsai J-C, Cheng H-P, Kuo J-F, Huang Y-H, Chen
Journal of Power Sources 2009; 192:336–343. C-Y. Journal of Power Sources 2009; 189:958–965.
23. Yoon SR, Hwang GH, Cho WI, Hong S-A, Ha HY. 46. Li T, Yang Y. Journal of Power Sources 2009;
Journal of Power Sources 2002; 106:215–223. 187:332–340.
24. Ladewig BP, Knott RB, Martin DJ, Costa JD, Lu GQ. 47. Huang QM, Zhang QL, Huang HL, Li WS, Huang
Electrochemistry Communications 2007; 9:781–786. YJ, Luo JL. Journal of Power Sources 2008;
25. Wu Z, Sun G, Jin W, Hou H, Wang S, Xin Q. Jour- 184:338–343.
nal of Membrane Science 2008; 313:336–343. 48. Kim J, Kim B, Jung B. Journal of Membrane Science
26. Hasani-Sadrabadi MM, Dashtimoghadam E, Majedi 2002; 207:129–137.
FS, Kabiri K, Solati-Hashjinb M, Moaddel H. Jour- 49. Kim B, Kim J, Jung B. Journal of Membrane Science
nal of Membrane Science 2010; 365:286–293. 2005; 250:175–182.

1226 Int. J. Energy Res. 2011; 35:1213–1228 # 2011 John Wiley & Sons, Ltd.
DOI: 10.1002/er
Methanol crossover in direct methanol fuel cells M. Ahmed and I. Dincer

50. Kim YS, Sumner MJ, Harrison WL, Riffle JS, 71. Zhao C, Lin H, Na H. International Journal of
McGrath JE, Pivovar BS. Journal of the Electro- Hydrogen Energy 2010; 35:2176–2182.
chemical Society 2004; 151(12):A2150–A2156. 72. Zhao C, Lin H, Zhang Q, Na H. International Jour-
51. Okamoto K-I, Yin Y, Yamada O, Islama MN, Honda nal of Hydrogen Energy 2010; 35:10482–10488.
T, Mishima T, Suto Y, Tanaka K, Kita H. Journal of 73. Zhang Y, Wan Y, Zhang G, Shao K, Zhao C, Li H,
Membrane Science 2005; 258:115–122. Na H. Journal of Membrane Science 2010; 348:
52. Woo Y, Oh SY, Kang YS, Jung B. Journal of Mem- 353–359.
brane Science 2003; 220:31–45. 74. Jiang Z, Jiang Z-J, Yu X, Meng Y. Plasma Process.
53. Einsla BR, Kim YS, Hickner MA, Hong Y-T, Hill Polym. 2010; 7:382–389.
ML, Pivovar BS, McGrath JE. Journal of Membrane 75. Ye Y-S, Chen W-Y, Huang Y-J, Cheng M-Y, Yen
Science 2005; 255:141–148. Y-C, Cheng C-C, Chang F-C. Journal of Membrane
54. Fang J, Shen PK, Liu QL. Journal of Membrane Sci- Science 2010; 362:29–37.
ence 2007; 293:94–99. 76. Carretta N, Tricoli V, Picchioni F. Journal of Mem-
55. Zhang Q, Gong F, Zhang S, Li S. Journal of Mem- brane Science 2000; 166:189–197.
brane Science 2011; 367:166–173. 77. Kim DS, Guiver MD. Journal of Polymer Science:
56. Liu C, Li L, Liu Z, Guo M, Jing L, Liu B, Jiang Z, Part A: Polymer Chemistry 2008; 46:989–1002.
Matsumoto T, Guiverc MD. Journal of Membrane 78. Huang C-H, Wu H-M, Chen C-C, Wang C-W, Kuo
Science 2011; 366:73–81. P-L. Journal of Membrane Science 2010; 353:1–9.
57. Chen X, Chen P, An Z, Chen K, Okamoto K. Journal 79. Jung B, Kim B, Yang JM. Journal of Membrane Sci-
of Power Sources 2011; 196:1694–1703. ence 2004; 245:61–69.
58. Zhang Y, Fei X, Zhang G, Li H, Shao K, Zhu J, Zhao 80. Smitha B, Sridhar S, Khan AA. Journal of Applied
C, Liu Z, Han M, Na H. International Journal of Hy- Polymer Science 2005; 95:1154–1163.
drogen Energy 2010; 35:6409–6417. 81. Yang T. International Journal of Hydrogen Energy
59. Kim DS, Robertson GP, Guiver MD. Macromole- 2008; 33:6772–6779.
cules 2008; 41:2126–2134. 82. Wu C-S, Lin F-Y, Chen C-Y, Chu PP. Journal of
60. Feng S, Shang Y, Xie X, Wang Y, Xu J. Journal of Power Sources 2006; 160:1204–1210.
Membrane Science 2009; 335:13–20. 83. Rhim J-W, Park HB, Lee C-S, Jun J-H, Kim DS, Lee
61. Zhang Y, Cui Z, Zhao C, Shao K, Li H, Fu T, Na H, YM. Journal of Membrane Science 2004; 238:
Xing W. Journal of Power Sources 2009; 191: 143–151.
253–258. 84. Wang J, Zhao C, Zhang G, Zhang Y, Ni J, Ma W,
62. Li L, Wang Y. Journal of Membrane Science 2005; Na H. Journal of Membrane Science 2010;
246:167–172. 363:112–119.
63. Elabd YA, Napadensky E, Sloan JM, Crawford DM, 85. Jiang Z, Zheng X, Wu H, Wang J, Wang Y. Journal
Walker CW. Journal of Membrane Science 2003; of Power Sources 2008; 180:143–153.
217:227–242. 86. Fu R-Q, Julius D, Hong L, Lee J-Y. Journal of Mem-
64. Yin Y, Fang J, Cui Y, Tanaka K, Kita H, Okamoto brane Science 2008; 322:331–338.
K-I. Polymer 2003; 44:4509–4518. 87. Escribano PG, Canovas MJ, Ojeda MC, Delri'o C,
65. Won J, Park HH, Kim YJ, Choi SW, Ha HY, Oh I-H, Sa'nchez F, Acosta JL. Journal of Polymer Science:
Kim HS, Kang YS, Ihn KJ. Macromolecules 2003; Part B: Polymer Physics 2009; 47:1203–1210.
36:3228–3234. 88. Feng S, Shang Y, Liua G, Dong W, Xie X, Xu J,
66. Li L, Zhang J, Wang Y. Journal of Membrane Sci- Mathur VK. Journal of Power Sources 2010;
ence 2003; 226:159–167. 195:6450–6458.
67. Li H, Zhang G, Wu J, Zhao C, Zhang Y, Shao K, Han 89. Kim DS, Park HB, Rhim JW, Lee YM. Journal of
M, Lin H, Zhu J, Na H. Journal of Power Sources Membrane Science 2004; 240:37–48.
2010; 195:6443–6449. 90. Li L, Xu L, Wang Y. Materials Letters 2003;
68. Liu B, Robertson GP, Kim D-S, Guiver MD, 57:1406–1410.
Hu W, Jiang Z. Macromolecules 2007; 40(1934): 91. Geng J, Jiang Z, Wang J, Shi Y, Yang D, Xiao L.
1944. Chemical Engineering and Technology 2010; 33,
69. Lin H, Zhao C, Ma W, Li H, Na H. International No. 2: 244–250.
Journal of Hydrogen Energy 2009; 34:9795–9801. 92. Helen M, Viswanathan B, Murthy SS. Journal of
70. Kim YS, Hickner MA, Dongd L, Pivovar BS, Membrane Science 2007; 292:98–105.
McGrath JE. Journal of Membrane Science 2004; 93. Kuo P-L, Jheng W-H, Liang W-J, Chen W-F. Jour-
243:317–326. nal of Power Sources 2010; 195:6434–6442.

Int. J. Energy Res. 2011; 35:1213–1228 # 2011 John Wiley & Sons, Ltd. 1227
DOI: 10.1002/er
M. Ahmed and I. Dincer Methanol crossover in direct methanol fuel cells

94. Yang T. International Journal of Hydrogen Energy 117. Lin C-K, Kuo J-F, Chen C-Y. Journal of Power
2009; 34:6917–6924. Sources 2009; 187:341–347.
95. Wang Y, Yang D, Zheng X, Jiang Z, Li J. Journal of 118. Fu TZ, Liu J, Cui ZM, Ni J, Zhang G, Yu HB, Zhao
Power Sources 2008; 183:454–463. CJ, Shi YH, Na H, Xing W. Fuel Cells 2009;
96. Fang Y, Miao R, Wang T, Wang X. Pure and 9-5:570–578.
Applied Chemistry 2009; 81-12:2309–2316. 119. Lin C, Wang CY. J. Power Sources 2003; 113:
97. Tripathi BP, Kumar M, Shahi VK. Journal of Mem- 145–150.
brane Science 2010; 360:90–101. 120. Bae B, Kho BK, Lim T, Oh I, Hong S, Ha HY. J.
98. Pivovar BS, Wang Y, Cussler EL. Journal of Mem- Power Sources 2006; 158:1256–1261.
brane Science 1999; 154:155–162. 121. Dimitrova P, Friedrich KA, Vogt B, Stimming U.
99. Nguyen TH, Wang C, Wang X. Journal of Mem- Journal of Electroanalytical Chemistry 2002; 532:
brane Science 2009; 342:208–214. 75–83.
100. Li H, Zhang G, Wu J, Zhao C, Jia Q, Lew CM, Zhang 122. Kim YS, Sumner MJ, Harrison WL, Riffle JS,
L, Zhang Y, Han M, Zhu J, Shao K, Ni J, Na H. Jour- McGrath JE, Pivovar BS. Journal of the Electro-
nal of Power Sources 2010; 195:8061–8066. chemical Society 2004; 151:A2150–A2156.
101. Wang J, Zhao Y, Hou W, Geng J, Xiao L, Wu H, 123. Yang T. Journal of Membrane Science 2009;
Jiang Z. Journal of Power Sources 2010; 195: 342:221–226.
1015–1023. 124. Gosalawit R, Chirachanchai S, Shishatskiy S, Nunes
102. Wang J, Zhang Y, Wu H, Xiao L, Jiang Z. Journal of SP. Journal of Membrane Science 2008; 323:
Power Sources 2010; 195:2526–2533. 337–346.
103. Wang J, Zhang H, Jiang Z, Yang X, Xiao L. Journal 125. Park Y-S, Yamazaki Y. Polymer Bulletin 2005;
of Power Sources 2009; 188:64–74. 53:181–192.
104. Bello M, Javaid Zaidi SM, Rahman SU. Journal of 126. Chen K, Hu Z, Endo N, Fang J, Higa M, Okamoto
Membrane Science 2008; 322:218–224. K-I. Journal of Membrane Science 2010; 351:
105. Cho E-B, Luu DX, Kim D. Journal of Membrane Sci- 214–221.
ence 2010; 351:58–64. 127. Kang M-S, Kim JH, Won J, Moon S-H, Kang YS.
106. Cho E-B, Kim H, Kim D. The Journal of Physical Journal of Membrane Science 2005; 247:127–135.
Chemistry. B 2009; 113:9770–9778. 128. Jiang R, Zhang Y, Swier S, Wei X, Erkey C, Kunz
107. Yang T, Xu Q, Wang Y, Lu B, Zhang P. Interna- HR, Fenton JM. Electrochemical and Solid-State Let-
tional Journal of Hydrogen Energy 2008; 33: ters 2005; 8:A611–A615.
6766–6771. 129. Yamauchi A, Ito T, Yamaguchi T. Journal of Power
108. Joo SH, Pak C, Kim EA, Lee YH, Chang H, Seung Sources 2007; 174:170–175.
D, Choi YS, Park J-B, Kim TK. Journal of Power 130. Prabhuram J, Zhao TS, Liang ZX, Yang H, Wong
Sources 2008; 180:63–70. CW. Journal of the Electrochemical Society 2005;
109. Shin J-P, Chang B-J, Kim J-H, Lee S-B, Suh 152:A1390–A1397.
DH. Journal of Membrane Science 2005; 251: 131. Liu F, Lu G, Wang C-Y. Journal of the Electrochem-
247–254. ical Society 2006; 153:A543–A553.
110. Thangamuthu R, Lin CW. Solid State Ionics 2005; 132. Miyake N, Wainright JS, Savinell RF. Journal of the
176:531–538. Electrochemical Society 2001; 148:A905–A909.
111. Mat NC, Liong A. Engineering Letters 2009; 17 133. Gubler L, Kramer D, Belack J, Ünsal Ö, Schmidt TJ,
(4):1–4. Scherer GG. Journal of the Electrochemical Society
112. Munakata H, Yamamoto D, Kanamura K. Journal of 2007; 154:B981–B987.
Power Sources 2008; 178:596–602. 134. Zhou X, Weston J, Chalkova E, Hofmann MA, Am-
113. Lin H, Zhao C, Jiang Y, Ma W, Na H. Journal of bler CM, Allcock HR, Lvov SN. Electrochimica Acta
Power Sources 2011; 196:1744–1749. 2003; 48:2173–2180.
114. Navarro A, Ri C, Acosta J. Journal of Polymer Sci- 135. Tricoli V, Carretta N, Bartolozzi M. Journal of the
ence: Part B: Polymer Physics 2008; 46:1684–1695. Electrochemical Society 2000; 147:1286–1290.
115. Li H, Zhang G, Ma W, Zhao C, Zhang Y, Han M, 136. Lee CH, Min KA, Park HB, Hong YT, Jung BO, Lee
Zhu J, Liu Z, Wu J, Na H. International Journal of YM. Journal of Membrane Science 2007; 303:
Hydrogen Energy 2010; 35:11172–11179. 258–266.
116. Xue S, Yin G, Cai K, Shao Y. Journal of Membrane 137. Boutry DT, Geyer AD, Guetaz L, Diat O, Gebel G.
Science 2007; 289:51–57. Macromolecules 2007; 40:8259–8264.

1228 Int. J. Energy Res. 2011; 35:1213–1228 # 2011 John Wiley & Sons, Ltd.
DOI: 10.1002/er

You might also like