Journal of Membrane Science 185 (2001) 29–39

On the development of proton conducting polymer membranes for

hydrogen and methanol fuel cells
K.D. Kreuer∗
Max-Planck-Institut für Festkörperforschung, Heisenbergstr 1, D-70569 Stuttgart, Germany
Received 17 January 2000; received in revised form 14 July 2000; accepted 14 July 2000

Abstract
The transport properties and the swelling behaviour of NAFION and different sulfonated polyetherketones are explained
in terms of distinct differences on the microstructures and in the pKa of the acidic functional groups. The less pronounced
hydrophobic/hydrophilic separation of sulfonated polyetherketones compared to NAFION corresponds to narrower, less
connected hydrophilic channels and to larger separations between less acidic sulfonic acid functional groups. At high water
contents, this is shown to significantly reduce electroosmotic drag and water permeation whilst maintaining high proton
conductivity. Blending of sulfonated polyetherketones with other polyaryls even further reduces the solvent permeation (a
factor of 20 compared to NAFION), increases the membrane flexibility in the dry state and leads to an improved swelling
behaviour. Therefore, polymers based on sulfonated polyetherketones are not only interesting low-cost alternative membrane
material for hydrogen fuel cell applications, they may also help to reduce the problems associated with high water drag
and high methanol cross-over in direct liquid methanol fuel cells (DMFC). The relatively high conductivities observed for
oligomers containing imidazole as functional groups may be exploited in fully polymeric proton conducting systems with no
volatile proton solvent operating at temperatures significantly beyond 100◦ C, where methanol vapour may be used as a fuel
in DMFCs. © 2001 Elsevier Science B.V. All rights reserved.
Keywords: NAFION; Polymer membrane; Direct liquid methanol fuel cell; Proton conductivity; Electroosmotic drag; Permeation; Proton
diffusion

1. Introduction brought into focus. This puts new demands on the

materials being used, and is in particular true for the
The concept of polymer electrolyte membrane separator membrane material, which, traditionally, is a
fuel cells (PEM-FC) has been well established since hydrated perfluorosulfonic polymer such as NAFION.
early 1960s, and PEM-FCs are successfully com- Although such membranes show superior perfor-
mercialised for niche applications, such as electrical mance in fuel cells operating at moderate temperature
power sources in space crafts and submarines. Dur- (<90◦ C) and high relative humidity with pure hydro-
ing the last decade, however, the use of PEM-FCs as gen as a fuel, their inherently high cost of production
power sources in mass products, such as the electri- makes their use in cost critical applications, such as
cal vehicles and portable electrical devices, was also fuel cells for electrical vehicles, unlikely. In addi-
tion the properties of such polymer membranes are
∗ Tel.: +49-711-689-1772; fax: +49-711-689-1722. insufficient if higher temperatures and/or fuels differ-
E-mail address: kreuer@chemix.mpi-stuttgart.mpg.de ent from pure hydrogen are used. Currently, there is
(K.D. Kreuer). an increasing interest in using hydrogen-rich gases

0376-7388/01/$ – see front matter © 2001 Elsevier Science B.V. All rights reserved.
PII: S 0 3 7 6 - 7 3 8 8 ( 0 0 ) 0 0 6 3 2 - 3
30 K.D. Kreuer / Journal of Membrane Science 185 (2001) 29–39

produced by reforming methanol or even gasoline. acid/polymer ratio, while proton conductivity still
Such gases contain traces of different gases in par- remains relatively high [13]. Due to the low transport
ticular CO, which reduces the activity of platinum or coefficients of other species in the highly viscous phos-
platinum-alloys (e.g. [1]), which are generally used phoric matrix and the low solubility of methanol at
as anode catalysts. The CO tolerance, however, in- high temperatures, methanol cross-over is drastically
creases with increasing temperature, and therefore, reduced compared to hydrated acidic polymers [14].
fuel cell operation at somewhat higher temperature is In this paper, two approaches towards membrane
desirable. Also the direct electrochemical oxidation materials operating at higher temperatures will be
of methanol is significantly promoted with increas- presented. The first is based on the modification of
ing temperature. For direct liquid methanol fuel cells the microstructure of sulfonated polyaryls by blend-
(DMFC), however, water and methanol cross-over ing with other polymers. It will be demonstrated that
significantly increases with temperature, which re- water cross-over can drastically be reduced while
duces the fuel efficiency and requires an expensive maintaining high proton conductivity, which is of sig-
water management systems (e.g. [2]). nificant importance for applications in direct liquid
In short, limited operation temperature due to the methanol fuel cells. In the second approach, the use
humidification requirements, water and methanol of heterocycles as proton solvating species will be
cross-over and cost are severe disadvantages of plain examined. This may lead the way to fully polymeric
perfluorosulfonic polymers. proton conductors, which may allow further exten-
These limitations have already stimulated a variety sion of the operation temperature in low humidity
of approaches in the development of alternative poly- environments.
meric proton exchange membranes (for a recent re- Apart from already published results (see refer-
view see [3]). Most of them are sulfonated polymers, ences), some recent, as yet unpublished, results are
but also sulfamides have been used as thermally more included in this general discussion. Experimental and
stable acid functional groups [4]. In any case, proton methodological details will be published elsewhere.
conductivity relies on proton solvation by water at
high water activities, and the limitations of their use in
fuel cells are generally similar to those of NAFION. 2. Hydrated polymers based on sulfonated
Modification of such membranes by the inclusion polyetherketones
of small inorganic particles such as silica [5–7] or
zirconiumphosphates sulfophenylphosphonates [7] The choice of polyaryls, in particular polyetherke-
leads to some improvement of the performance of tones instead of perfluorinated polymer backbones,
such membranes especially in pressurised fuel cells was mainly based on cost and stability considera-
operating at temperatures up to 140◦ C [5]. Also the tions [15]. Although sulfonated polyaryls have been
morphological stabilisation of acidic polymers by ei- demonstrated to suffer from hydroxy radical initiated
ther acid/base blending or covalent cross-linking [8,9] degradation [16] we found sulfonated polyetherke-
appears to reduce swelling and water and methanol tones to be durable under fuel cell conditions over
cross-over. But reduced conductivity and brittleness several thousand hours. Polyetherketones (PEEKK
in the dry state emerge as new problems. and PEEK) functionalised by electrophilic sulfonation
Distinctly different approaches are based on the with sulfuric acid (similar to the procedure described
complexation of basic polymers, such as polyben- in [17]) were supplied by Hoechst–Aventis. The hy-
zimidazole, with oxo-acids (e.g. [10,11]) especially dration behaviour and the transport of protonic charge
for phosphoric acid, high acid to polymer ratios lead carriers and water have been examined and com-
to conductivities close to that of the pure acid. As in pared to the corresponding properties of NAFION.
the case of pure phosphoric acid [12], proton con- The distinct differences are qualitatively explained by
ductivity of such adducts is predominantly carried by differences in the microstructures and the acidity of
structure diffusion, i.e. proton transfer between phos- the sulfonic acid functional groups [15,18,19]. The
phate species and phosphate reorientation. Phospho- variation of these features by chemical modification,
ric acid is successively immobilised with decreasing blending or cross-linking may provide space for the
 K.D. Kreuer / Journal of Membrane Science 185 (2001) 29–39 31

adaptation of the polymer properties to particular fuel sulfonic acid functional group is less acidic and there-
cell applications. fore, also less polar) and the smaller flexibility of the
Perfluorosulfonic polymers naturally combine, polymer backbone, the separation into a hydrophilic
in one macromolecule, the extremely high hydro- and a hydrophobic domain is less pronounced.
phobicity of the perfluorinated backbone with the This can directly be inferred from the results of
extremely high hydrophilicity of the sulfonic acid small angle X-ray scattering (SAXS) experiments.
functional groups. Especially in the presence of wa- For a hydrated sulfonated polyetherketone compared
ter, this gives rise to some hydrophobic/hydrophilic to NAFION the ionomer peak is broadened and shifted
nano-separation. The sulfonic acid functional groups towards higher scattering angles and the scattering
aggregate to form a hydrophilic domain. When this intensity at high scattering angles (Porod-regime) is
is hydrated, protonic charge carriers form within in- higher (Fig. 1). This indicates a smaller characteristic
ner space charge layers by dissociation of the acidic separation length with a wider distribution and a
functional groups, and proton conductance assisted larger internal interface between the hydrophobic and
by water dynamics occurs. While the well connected hydrophilic domain for the hydrated sulfonated poly-
hydrophilic domain is responsible for the transport of etherketone. The SAXS data and water self-diffusion
protons and water, the hydrophobic domain provides coefficients obtained by pulsed-field-gradient (PFG)-
the polymer with the morphological stability and NMR have been used to consistently parameterise a
prevents the polymer from dissolving in water. simple model for the microstructure, which is based
The situation in sulfonated polyetherketones was on a cubic hydrophilic channel system in a hydropho-
found to be distinctly different with respect to both bic matrix. Data for channel diameter, channel sepa-
transport properties and morphological stability. As ration, degree of branching and number of dead-end
a result of the smaller hydrophilic/hydrophobic dif- channels have been obtained for both type of poly-
ference (the backbone is less hydrophobic, and the mers [19]. As schematically illustrated in Fig. 2, the

Fig. 1. Small angle X-ray scattering spectra of hydrated NAFION and a hydrated sulfonated polyetherketone. The characteristic hy-
drophobic/hydrophilic separation lengths are obtained from the position of the ionomere peaks while the internal hydrophobic/hydrophilic
interfaces are obtained from the intensities in the Porod-regimes [19].
32 K.D. Kreuer / Journal of Membrane Science 185 (2001) 29–39

Fig. 2. Schematic representation of the microstructures of NAFION and a sulfonated polyetherketone (derived from SAXS experiments
[19]) illustrating the less pronounced hydrophobic/hydrophilic separation of the latter compared to the first.

water filled channels in sulfonated PEEKK are nar- (DH2 O ) are comparable (Fig. 4). With decreasing
rower compared to those in NAFION. They are less water volume fraction, however, the water diffu-
separated and more branched with more dead-end sion coefficient decreases more rapidly in sulfonated
“pockets”. These features correspond to the larger hy- PEEKK compared to NAFION. The mobility of pro-
drophilic/hydrophobic interface and, therefore, also to tonic charge carriers (Dσ ), as obtained from conductiv-
a larger average separation of neighbouring sulfonic ity data via the Nernst–Einstein relationship assuming
acid functional groups. The stronger confinement of full dissociation of the sulfonic acid functional groups,
the water in the narrow channels of the aromatic poly- shows a similar behaviour since this is roughly related
mers leads to a significantly lower dielectric constant to the water diffusion coefficient (Fig. 4). At very high
of the water of hydration (about 20 compared to al- water contents, however, Dσ is somewhat higher than
most 64 in fully hydrated NAFION [20,21] (Fig. 3)). DH2 O indicating some intermolecular proton transfer
For high water volume fractions the percolation in being involved in the mobility of protonic charge car-
the hydrophilic domain is similar in both microstruc- riers as in the case of dilute aqueous solutions of acids
tures, and since the water/polymer interaction is small (structure diffusion [22–24]). At very low water con-
in this regime, also the water self-diffusion coefficients tents, the opposite is true. As a result of the decreasing
 K.D. Kreuer / Journal of Membrane Science 185 (2001) 29–39 33

Fig. 3. Dielectric constant (ε total ) for hydrated NAFION and a sulfonated polyetherketone obtained at 5 Ghz as a function of the degree of
hydration [20,21]. The dielectric constant of the water of hydration (εH2 O ) has been obtained by extrapolating the data to a water volume
fraction of unity (pure water).

degree of dissociation of the acidic functional group pared to DH2 O with decreasing water volume fraction.
and the decreasing dielectric screening of the anionic While this effect is negligible for NAFION, it is quite
counter charge, the excess protons tend to be more pronounced in sulfonated PEEKK as a consequence
localised in the vicinity of the sulfonic acid functional of the higher pKa of the acidic functional group (ap-
groups, which leads to a stronger decrease of Dσ com- proximately −1 compared to −6 for the superacid
NAFION as calculated by the program “pKa database
4.0” [25]) and the lower dielectric constant of the
water of hydration (see also Fig. 3), which allows only
for a weak dielectric screening of the negative charge
of the sulfonic acid anion. Percolation and proton
localisation effects explain why the proton conduc-
tivity in sulfonated aromatic polymers decrease much
more severely with decreasing hydration levels than
in NAFION.
While the mobility of protonic charge carriers, the
self-diffusion and even the chemical diffusion of water
are random walk processes on a molecular scale with
no or only small drift velocity superimposed, the elec-
troosmotic drag and the water permeation correspond
to the collective flow of water through the membrane,
either as a result of a proton flux or a total pressure
gradient. As opposed to the first, the latter transport
coefficients depend in a direct way on the size of
the channels (Hagen–Poiseuille-type problem [19]).
Fig. 4. Water self-diffusion coefficient (DH2 O ) and proton mobility
An important consequence of the narrow channels in
(Dσ ) as a function of the water volume fraction in NAFION and sulfonated polyetherketones is that the electroosmotic
a sulfonated polyetherketone (data taken from [15]). drag (Fig. 5) as well as the permeation coefficient
34 K.D. Kreuer / Journal of Membrane Science 185 (2001) 29–39

Fig. 5. Electroosmotic drag coefficient obtained by electrophoretic-NMR as a function of the water content n = [H2 O]/[–SO3 H] [26].

(Table 1) are distinctly lower than in well separated the water management under operating conditions
perfluorosulfonic polymers for a given water content [27–29], and the low solvent (water and methanol)
[19,26]. cross-over, which has contributions from electroos-
In the dry state, such membranes are quite brittle, motic drag and permeation, is of advantage especially
while they become soft in the presence of water. Fig. 6 in direct methanol fuel cells (e.g. [2]). On the other
shows the swelling behaviour of sulfonated polyether- hand, the stronger decrease of the conductivity upon
ketones (PEEKK) with different degrees of sulfona- dehydration allows only operation close to the dew
tion compared to that of NAFION. For a sulfonation point of water. While this is not a severe disadvan-
level of about 70% per repeat unit (1.40 meq./g), which tage in many cases, the brittleness of pure sulfonated
corresponds to a similar concentration of acid func- polyetherketones makes their handling difficult and
tional groups as in NAFION, the onset of exaggerated may cause mechanical membrane failure during
swelling is about 50◦ C below that of NAFION. For operation. Also the lower maximum operation tem-
degrees of sulfonation higher than 80% (1.78 meq./g), perature of about 80◦ C, compared to about 140◦ C for
PEEKK even becomes water soluble. NAFION (see Fig. 6) is a severe limitation for fuel
As membrane materials for fuel cells, plain cell applications.
sulfonated polyetherketones apparently have some Based on the above analyses, however, the follow-
advantages and disadvantages compared to perfluoro- ing chemical and microstructural modifications of sul-
sulfonic polymers. The low drag coefficients facilitate fonated polyetherketones should lead to a combination

Table 1
Water permeation coefficients (PH2 O ) for different membrane materials immersed in water (T = 300 K, 1p = 20 hPa).

Membrane material Titrated ion exchange Water content n/ PH2 O /(mol cm−1 s−1 105 Pa−1 )
capacity/(meq./g) ([H2 O]/[–SO3 H])
NAFION 117 0.91 16 1.52 × 10−9
PEEKK (65% sulfonated) 1.43 19 8.9 × 10−10
PEEK (54% sulfonated) + 10 w/o PES 1.39 11 7.42 × 10−11
PEEK (100% sulfonated) + 21 w/o PBI 1.40 7 2.19 × 10−11
 K.D. Kreuer / Journal of Membrane Science 185 (2001) 29–39 35

Fig. 7. Proton mobility in hydrated blends of sulfonated PEEK

with PES and PBI, respectively. The proton mobilities in NAFION
and plain sulfonated polyetherketone is shown for comparison.
Fig. 6. Swelling in water for NAFION, plain sulfonated polyether-
ketones (PEEKK) with different degrees of sulfonation and a blend
of a sulfonated PEEK and PES (ion exchange capacities are given
in brackets). blend of 54% sulfonated PEEK with 10% PES takes
up only 20 H2 O/–SO3 H in water at 90◦ C (Fig. 6)). For
high water volume fractions, the proton mobility (Dσ )
of some advantages of both type of polymers. is virtually identical to that of the plain sulfonated
polyetherketone (Fig. 7). Since the material shows
1. In order to reduce the drastic decrease of the proton
very low swelling at low temperatures, pretreatment
conductivity with decreasing water content, the
in water (e.g. at 90◦ C) is required to achieve water
sulfonic acid functions should be more acidic
volume fractions, which allow the very high proton
(superacid) and their concentration in the hydro-
conductivities required especially for hydrogen fuel
philic domain should be high, i.e. their average
cell applications. For DMFC applications, however,
separation should be low.
the reduced swelling and the absence of any signifi-
2. A well-connected system of very narrow chan-
cant hydrophobic/hydrophilic separation, as indicated
nels should allow for low solvent permeation and
by the virtual disappearance of the ionomer peak in the
electroosmotic drag coefficient along with high
SAXS spectrum, are essential advantages, since these
mobility of the protonic charge carriers.
features are expected to reduce the electroosmotic drag
3. Since narrow channels imply the absence of any
as well as the methanol permeation (see above). A
significant hydrophilic/hydrophobic separation, the
significant reduction of water permeation has already
morphological stability must be achieved by other
been verified experimentally (Table 1) while measure-
means, e.g. by blending and/or cross-linking [18].
ments of drag coefficients by E-NMR are underway.
Blends of sulfonated polyetherketones with other Another approach, as suggested by Kerres et al.
aromatic polymers already provides some of these [8,30], relies on the acid/base interaction between a
features. sulfonated polyetherketone and another basic poly-
One approach is based on the entanglement of a sul- mer, preferentially polybenzimidazole (PBI), which is
fonated PEEKK and an unsulfonated polyethersulfone well known to be durable under fuel cell conditions as
(PES). Blends of the two polymers show higher flexi- constituent of PBI-phoshoric acid adducts [11,14]. In
bility in the dry state and an improved morphological these blends, only part of the acidic functional groups
stability as indicated by less swelling in water (e.g. a are neutralised by the basic amine groups of PBI.
36 K.D. Kreuer / Journal of Membrane Science 185 (2001) 29–39

The proton (charge) transfer between the two types of 3. Polymers with immobilised heterocycles as
polymers leads to the formation of an inner salt, i.e. proton solvents
cross-linking by means of coulomb interaction, which
also reduces swelling in water. For a blend of a fully Since the very high proton conductivity of hydrated
sulfonated PEEK with 10 w/o PBI (corresponding to polymers relies on the presence of liquid water as
an equivalent weight of a 60% sulfonated PEEKK) thermodynamically distinct phase in the hydrophilic
about 39 H2 O/–SO3 H are absorbed at 80◦ C. Surpris- domain [15,18], the maximum operation temperature
ingly, there is no further swelling at higher temper- is approximately given by the boiling point of water,
atures (up to 120◦ C). This high but constant water i.e. 100◦ C at p = 105 Pa. It has been shown, how-
level gives rise to the high proton conductivity of this ever, that substitution of water by heterocycles, such
blend. The proton mobility is in fact similar to plain as imidazole, pyrazole or benzimidazole leads to pro-
sulfonated PEEKK (70% sulfonation) for high water ton conductivities between 150 and 250◦ C which are
contents (Fig. 7). It should, however, be mentioned comparable to the conductivities of hydrated poly-
that acid/base cross-linking is not strong enough to mers [31,32]. Indeed, such heterocycles form similar
prevent the highly sulfonated constituents of such hydrogen bond networks to water, and the transport
blends to be slowly leached out in water. Of course, properties in the liquid state are similar to water for
this can be prevented by reducing the degree of sul- a given temperature relative to the melting point (e.g.
fonation of the acid component (sulfonated polyether- for a mixture of benzimidazole with 10 m/o H3 PO4 a
ketone), but the price to be paid is a somewhat lower conductivity of 5 × 10−2 S cm−1 is observed at 200◦ C
conductivity (Fig. 7). In this respect, blends with PES, [32]). While such adducts are of high interest for
discussed above, behave more advantageously. For an applications in closed electrochemical cells, such
initial ion exchange capacity of 1.39 meq./g the ion as supercapacitors and electrochromic devices, the
exchange capacity slowly decreases in water only for volatility of the heterocycles prevents them from be-
temperatures higher 90◦ C. ing used in open electrochemical systems, such as

Fig. 8. Proton conductivity of pure monomeric heterocycles and oligomers terminated by imidazole [32,34].
 K.D. Kreuer / Journal of Membrane Science 185 (2001) 29–39 37

fuel cells. In contrast to using water as the proton As a first attempt, we have therefore, prepared
solvent, which is usually supplied to the membrane oligomers consisting of a short polyethylenoxide
by humidifying the anode and cathode gases and is segment terminated by imidazole groups [34]. Such
produced by the electrochemical reaction itself at oligomers are highly viscous oils which locally ag-
the cathode, the use of heterocycles as the proton gregate in such a way, that strong hydrogen bonds
solvent requires the immobilisation of the solvent are formed between terminating groups of different
in the polymer membrane in such a way that high oligomers. Like monomeric heterocycles [31,32],
mobility of the protonic charge carriers is still guar- these oligomers show significant self-dissociation, i.e.
anteed. While proton mobility in hydrated polymers formation of protonic charge carriers in the absence
has large contributions from the diffusion of hydrated of an explicit acid, and a proton conductivity, which
protons [15,18], proton mobility in an environment is only slightly lower than for the monomers (Fig. 8).
of immobilised heterocycles must completely rely on On the basis of these encouraging results, the effects
structure diffusion (Grotthuss-type mechanism [24]) of polymerisation of such oligomers, their grafting to
comprising proton transfer between heterocycles inert polymeric networks and the effects of acid doping
and solvent reorganisation. This is generally a very are currently under investigation.
complex process requiring the thermally activated
accessibility of quite different configurations, such
as very short and elongated hydrogen bonds [33]. 4. Summary and conclusions
As a typical many-particle feature, such a situation
is preferred in systems with high concentration of The clue to the understanding the different mec-
solvent molecules with minimised restrictions for the hanical and transport properties of hydrated perflu-
local degrees of freedom (compared to the pure liquid orosulfonic polymers (here NAFION) and low-cost
solvent). sulfonated polyaryls (here sulfonated polyether-

Fig. 9. Proton conductivity of different fully hydrated acidic polymers [15,18] (for ion exchange capacities see Table 1), and a liquid, an
adduct, and an oligomer containing heterocycles as proton solvent [31,32,34].
38 K.D. Kreuer / Journal of Membrane Science 185 (2001) 29–39

ketones) appears to be differences in the pKa of the beyond 100◦ C, where methanol vapour may be used
sulfonic acid functional groups and of the microstruc- as a fuel in DMFCs.
tures. The less pronounced hydrophobic/hydrophilic
separation of sulfonated polyetherketones compared
to NAFION corresponds to narrower, less connected Acknowledgements
hydrophilic channels and to larger separations be-
tween the less acidic sulfonic acid functional groups. The author thanks M. Ise for many fruitfull discus-
On one side, this leads to a disadvantageous swelling sions, A. Fuchs for technical assistance, J. Clauß and
behaviour and a stronger decrease of water and Th. Soczka-Guth (Hoechst–Aventis, Frankfurt) for
proton transport coefficients with decreasing water supplying sulfonated polyetherketones, J. Kerres (Uni-
content; on the other hand side, however, the hydro- versity of Stuttgart) for supplying S-PEEK/PBI-blends
dynamic flow of water, i.e. electroosmotic drag and and R. De Souza for reading the proofs. The work on
water permeation, is reduced compared to NAFION, hydrated polymers has been supported by the BMBF
which is an essential advantage, especially for DMFC under the contract number 0329567 and the work on
applications. hyterocyclic systems by the DFG under the contract
Blending of sulfonated polyetherketones with inert number KR 794/7-2 and ME 1495/2-2 (SPP 1060).
polymers (here PES) or basic polymers (here PBI)
significantly improves the swelling behaviour without References
reducing the high proton conductivity at high water
contents (Fig. 9), blending with PES even yields flex- [1] R. Ianniello, V.M. Schmidt, U. Stimming, J. Stumper, A.
ibility in the dry state. Blending also further reduces Wallan, CO adsorption and oxidation on Pt-Ru alloys —
the hydrophilic/hydrophobic separation and therefore, dependence on substrate composition, Electrochim. Acta 39
also hydrodynamic solvent transport (water and prob- (1994) 1863.
[2] K. Sundmacher, K. Scott, Direct methanol polymer electrolyte
ably also methanol permeation). Therefore, polymers fuel cell: analysis of charge and mass transfer in the
based on sulfonated polyetherketones are not only in- vapour–liquid–solid system, Chem. Eng. Sci. 54 (1999) 2927.
teresting low-cost alternative membrane materials for [3] O. Savadogo, Emerging membranes for electrochemical
hydrogen PEM fuel cell applications, they may also systems. I. Solid polymer membranes for fuel cell systems,
J. New. Mater. Electrochem. Syst. 1 (1998) 47.
help to reduce the problems associated with high wa- [4] D. DesMarteau, S. Creager, B. Thomas, S. Savett, J. Atkins,
ter and methanol cross-over in DMFCs using aqueous S. Zhou, l. Ford, M. Bollinger, G. Shafer, Perfluorinated
solutions of methanol as a fuel. ionomers containing the sulfonimide function: promising
A completely different approach for obtaining membrane and electrocatalyst materials for high temperature
high proton conductivity at temperatures significantly PEM fuel cells, presented at the 11th annual meeting of the
North American Membrane Society, Boulder, 2000.
higher than 100◦ C is based on heterocycles as proton [5] A.S. Aricò, P. Cretı́, P.C. Antonucci, V. Antonucci, Compari-
solvent. Their transport properties are surprisingly son of ethanol and methanol oxidation in a liquid-feed solid
close to these of water for a given temperature relative polymer electrolyte fuel cell at high temperature, Electrochem.
to the respective melting point (about 90◦ C for imida- Solid State Lett. 1 (1998) 66.
[6] R.A. Zoppi, I.V.P. Yoshida, S.P. Nunes, Hybrids of perfluoro-
zole). High proton conductivities are easily achieved sulfonic acid ionomer and silicon oxide by sol–gel reaction
for liquid systems and adducts of acidic polymers and from solution: morphology and thermal analysis, Polymer 39
monomeric heterocycles (e.g. imidazole intercalated (1997) 1309.
into a sulfonated polyetherketone) (Fig. 9). [7] B. Bonnet, D.J. Jones, J. Rozière, L. Tchicaya, G. Alberti,
M. Casciola, L. Massinelli, B. Bauer, A. Peraio, E.
While these conductivities still have large contri-
Ramunni, Hybrid organic–inorganic membranes for a medium
butions from the mobility of solvated protons, e.g. temperature fuel cell, J. New. Mater. Electrochem. Syst. 3
imidazolium (vehicle mechanism [35]), the rela- (2000) 87.
tively high conductivity of pure oligomers such as [8] J. Kerres, A. Ullrich, F. Meier, T. Häring, Synthesis and char-
I-(PEO)n -I (see Fig. 9) indicates also the presence of acterisation of novel acid–base polymer blends for application
in membrane fuel cells, Solid State Ionics 125 (1999) 243.
structure diffusion which may be exploited in fully [9] W. Zhang, G. Dai, J. Kerres, Synthesis of cross-linked
polymeric proton conducting systems with no volatile poly(sulfone) ion exchange membranes, Acta Polym. Sin. 5
proton solvent operating at temperatures significantly (1998) 608.
 K.D. Kreuer / Journal of Membrane Science 185 (2001) 29–39 39

[10] J.C. Lassègues, Mixed inorganic–organic systems: the acid/ [23] M.E. Tuckerman, D. Marx, M.L. Klein, M. Parrinello, On
polymer blends, in: Ph. Colomban (Ed.), Proton Conductors: the quantum nature of the shared proton in hydrogen bonds,
Solids, Membranes and Gels — Materials and Devices, Science 275 (1997) 275.
Cambridge University Press, Cambridge, 1992, pp. 311–328, [24] K.D. Kreuer, Proton conductivity, materials and applications,
Chapter 20. Chem. Mater. 8 (1996) 610.
[11] J.S. Wainright, J.-T. Wang, R.F. Savinell, M. Litt, H. Moaddel, [25] pKa Database 4.0 by Advanced Chemistry Development Inc.
C. Rogers, Acid doped polybenzimidazole, a new polymer (Toronto), http://www.acdlabs.com/products/phys chem lab/
electrolyte, Proc. Electrochem. Soc. 94 (1994) 255. pka.
[12] Th. Dippel, K.D. Kreuer, J.C. Lassègues, D. Rodriguez, [26] M. Ise, K.D. Kreuer, J. Maier, Electroosmotic drag in polymer
Proton conductivity in fused phosphorc acid: a 1 H/31 P membranes: an electrophoretic NMR study, Solid State Ionics
PFG-NMR and QNS-study, Solid State Ionics 61 (1993) 41. 125 (1999) 213.
[13] A. Bozkurt, M. Ise, K.D. Kreuer, W.H. Meyer, G. Wegner, [27] M. Eikerling, Y.I. Kharkats, A.A. Kornyshev, Y.M.
Proton-conducting polymer electrolytes based on phosphoric Volfkovich, Phenomenological theory of electroosmotic effect
acid, Solid State Ionics 125 (1999) 225. and water management in polymer electrolyte proton-conduc-
[14] J.T. Wang, S. Wasmus, R.F. Savinell, Real-time mass ting membranes, J. Electrochem. Soc. 145 (1998) 2684.
spectrometric study of the methanol cross-over in a direct [28] T. Okada, G. Xie, M. Meeg, Ion and water transport charac-
methanol fuel cell, J. Electrochem. Soc. 143 (1996) 1225. teristics of NAFION membranes as electrolytes, Electrochim.
[15] K.D. Kreuer, On the development of proton conducting Acta 43 (1998) 2141.
materials for technological applications, Solid State Ionics 97 [29] X.M. Ren, W. Henderson, S. Gottesfeld, Electroosmotic drag
(1997) 1. of water in ionomeric membranes — new measurements
[16] G. Hubner, E. Roduner, EPR investigation of HO radical employing a direct methanol fuel cell, J. Electrochem. Soc.
initiated degradation of sulfonated aromatics as model 144 (1997) L267.
compounds for fuel cell proton conducting membranes, J. [30] W. Cui, J. Kerres, G. Eigenberger, Development and
Mater. Chem. 9 (1999) 409. characterisation of ion exchange polymer blend membranes,
[17] M. Rehahn, A.D. Schlüter, G. Wegner, Soluble poly(p- Sep. Purif. Technol. 14 (1998) 145.
phenylenes). 3. Variation of the length and the density of [31] K.D. Kreuer, A. Fuchs, M. Ise, M. Spaeth, J. Maier, Imidazole
the solubilised side chains, Macromol. Chem. 191 (1990) and pyrazole-based proton conducting polymers and liquids,
1991. Electrochim. Acta 43 (1998) 1281.
[18] K.D. Kreuer, Th. Dippel, J. Maier, Membrane materials [32] K.D. Kreuer, New proton conducting polymers for fuel cell
for PEM fuel cells: a microstructural approach, Proc. applications, in: B.V.R. Chowdari, et al. (Ed.), Solid State
Electrochem. Soc. 95-23 (1995) 241. Ionics: Science and Technology. Singapore: World Science
[19] M. Ise, Polymer Elektrolyt Membranen: Untersuchungen Publishing Co., 1998, pp. 263–274.
zur Mikrostruktur und zu den Transporteigenschaften für [33] K.D. Kreuer, On the complexity of proton conduction
Protonen und Wasser, Ph.D. Thesis, University of Stuttgart, phenomena, in: Proceedings of SSI-12, Solid State Ionics, in
2000. press.
[20] G. Bender, Fuel cell application of poly-ether-ether-keto: [34] M. Schuster, G. Scharfenberger, W.H. Meyer, H.G. Herz, M.
elctrochemical reaction kinetics and dielectric constant, Ise, K.D. Kreuer, Proton mobility in polymer-bound proton
Master Thesis, University of Stuttgart, 1999. solvents I: imidazole immobilised via flexible spacers, in:
[21] St.J. Paddison, G. Bender, K.D. Kreuer, N. Nicoloso, Th. Proceedings of SSPC-10, Solid State Ionics, submitted for
Zawodzinski, The microwave region of the dielectric spectrum publication.
of hydrated NAFION and other sulfonated membranes, J. [35] K.D. Kreuer, A. Rabenau, W. Weppner, Vehicle mechanism,
New. Mater. Electrochem. Syst. 3 (2000) 293. a new model for the interpretation of the conductivity of fast
[22] N. Agmon, The Grotthuss mechanism, Chem. Phys. Lett. 244 proton conductors, Angew. Chem. Int. Ed. Engl. 21 (1982)
(1995) 456. 208.