Side chain crosslinking of aromatic polyethers for high temperature polymer electrolyte membrane fuel cell applications

Article abstract of DOI:10.1002/pola.25006

Novel aromatic polymers bearing polar pyridine units in the main chain and side chain crosslinkable hydroxyl and propargyl groups have been successfully synthesized. The polymers have been investigated in terms of their critical properties related to thei

Full text of DOI:10.1002/pola.25006

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
Side Chain Crosslinking of Aromatic Polyethers for High Temperature
Polymer Electrolyte Membrane Fuel Cell Applications
Andrea Vo¨ ge,1,2 Valadoula A. Deimede,1,3 Joannis K. Kallitsis1,2,3
1Department of Chemistry, University of Patras, Patras 26500, Greece
2Foundation of Research and Technology-Hellas, Institute of Chemical Engineering and High
Temperature Processes (FORTH-ICE/HT), Patras 26504, Greece
3Advent Technologies S.A., Research and Development Company, Patras Science Park, Stadiou Street, Patras 26504, Greece
Correspondence to: A. Voge (E-mail: andreavoege@online.de)
¨
Received 29 July 2011; accepted 1 September 2011; published online 4 October 2011
DOI: 10.1002/pola.25006
ABSTRACT: Novel aromatic polymers bearing polar pyridine
units in the main chain and side chain crosslinkable hydroxyl
and propargyl groups have been successfully synthesized. The
polymers have been investigated in terms of their critical prop-
erties related to their application in high temperature polymer
electrolyte membrane fuel cells, such as doping ability, me-
chanical properties, and thermal stability. Crosslinked mem-
branes were prepared by direct crosslinking of hydroxyl side
chain groups with decafluorobiphenyl used for the first time as
crosslinked polymers after thermal curing. Both types of cross-
linked membranes exhibited higher glass transition tempera-
tures as well as lower doping levels when doped in phosphoric
acid compared with the non crosslinked analogs, confirming
C
V
the formation of a successfully crosslinked network.
2011
Wiley Periodicals, Inc. J Polym Sci Part A: Polym Chem 50:
207–216, 2012
a
crosslinking agent. However, further functionalization of
KEYWORDS: aromatic polyethers; copolymerization; crosslink-
hydroxyl groups to the propargyl derivative has also led to
ing; high temperature PEM fuel cells; membranes
INTRODUCTION Fuel cells are
a more environmentally
based on aromatic polyethers bearing main chain and/or
side chain pyridine units have been developed by our
group.11–15 These copolymers possesses good mechanical
properties and excellent thermal, oxidative as well as chemi-
cal stability. Furthermore, the polar pyridine units interact
with and/or retain phosphoric acid molecules, thus resulting
in proton conductive materials.
friendly and efficient alternate power source compared with
existing power sources. Polymer electrolyte membrane fuel
cells (PEMFCs) have exceptional advantages hence they find
appliance in stationary and mainly vehicular applications.1
The proton exchange membrane (PEM) is the core of the
fuel cell, which must fulfil several requirements, such as high
ionic conductivity, high thermal, mechanical and chemical
stability, as well as low fuel or gas permeability. Up to now,
Nafion is the mostly well-known perfluorinated polymer
However, the high temperature PEMs are mostly designed
for residential application where durability up to 40.000 h is
required. In addition, operation at higher temperatures and
more specifically in the range of 200–220 ꢀC will open new
horizons because of the prospect to join the fuel cell with a
methanol reformer in a combined single cell.16–18 Hence, the
development of new electrolytes with improved mechanical
and thermal properties as well as adequate ionic conductiv-
ity still remains a challenge. An effective approach to
improve PEM properties is the crosslinking method, which
includes the ionic and covalent crosslinking. Ionic crosslink-
ing is based on interaction forces between different types of
ionomers, such as acid-base polymers.19–22 However, ionic
crosslinking is not effective at high temperatures because of
membrane’s mechanical integrity loss. Covalent crosslinking
is more favorable because in this case membranes can
2–4
ꢀ
used in PEMFCs operating up to 80 C.
Besides its several
advantages, such as excellent chemical stability and high per-
formance, a main drawback is the low proton conductivity at
temperatures higher than 100 ꢀC, since dehydration of the
membrane occurs. Recently, research has been devoted in
the development of PEMFCs operating at temperatures above
100 ꢀC.5,6 Operation at high temperatures offers the distinct
advantages of high carbon monoxide tolerance, easier ther-
mal management, and enhanced kinetics on both electrodes
resulting in higher energy efficiency. Polybenzimidazole (PBI)
doped with phosphoric acid is the most extensively studied
polymer electrolyte used in high temperature PEMFCs due to
its high ionic conductivity and excellent thermal stability.7–10
Alternate polymer electrolytes for high temperature PEMFCs
C
V
2011 Wiley Periodicals, Inc.
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 207–216
207

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
analogs after ether cleavage. These polymers were used
either directly for covalent crosslinking or by further
functionalization to triple bonds that were crosslinked by
thermal treatment.
Aromatic poly(ether sulfone)s have been synthesized via
nucleophilic aromatic substitution (SNAr) of a bis(aryl fluo-
ride) with phenolate-derivatives. Alternate routes for the
synthesis of poly(ether sulfone)s and ketones using bis(aryl
chloride)s instead of fluorides, bearing in mind the economi-
cal and ecological aspects, have been described.47–50
SCHEME 1 Synthesis of 1,4-bis(4-hydroxyphenyl)-2,5-dimethoxy-
benzene (4).
For the successful preparation of the aforementioned poly-
mers via nucleophilic aromatic substitution, the correspond-
ing diols 4 and 8 have been prepared in large quantities and
good yields. The synthetic procedures followed for the syn-
thesis of the two new aromatic diols 4 and 8 are outlined in
Scheme 1 and 2. Both monomers were prepared by Suzuki
coupling, while similar Suzuki couplings with dialkoxy dibor-
onic acids have already been reported.51,52 In the case of
diol 4 the 2,5-dimethoxy-1,4-phenylene diboronic acid (1)
reacted with the THP-protected 4-bromophenol (2) resulted
in the precursor of the diol 3 (Scheme 1). Coupling of 4-
methoxyphenylboronic acid (6) with THP-protected 2,5-
dibromohydroquinone (5) led to compound 7 (Scheme 2).
Following deprotection of the alcohols under acidic condi-
tions in both cases resulted in diol 4 and 8, respectively.
preserve their properties even at elevated temperatures.
There are several possibilities for covalent crosslinking, e.g.,
the crosslinking of double bonds with benzoyl peroxide,23–26
bisazides,27–32 and photochemical crosslinking.33–35 The
crosslinking of triple bonds has been described as well, e.g.,
the thermal crosslinking of polymers containing endcapped
ethynylphenyl36–41 or phenylethynyl42 moieties by trimeriza-
tion of the alkynes to stable benzene systems. Another exam-
ple for covalent crosslinking is the crosslinking of PBI with
dihalides, such as p-xylene dichloride43 and p-xylene
dibromide.44
In this work, we report the synthesis, crosslinking and char-
acterization of new aromatic polymers bearing pyridine units
in the main chain and side cross-linkable hydroxyl and pro-
pargyl groups. The triple bonds were thermally crosslinked
since this method leads to highly stable benzene systems. On
the other hand, side hydroxyl groups offer various possibil-
ities for functionalization and crosslinking, because of their
ability to act after deprotonation by a base as nucleophiles
in nucleophilic substitutions. Reaction of these nucleophilic
species with di- or polyhalides leads to crosslinked systems.
Decafluorobiphenyl, a compound that has been widely used
Novel dimethoxy-based polymers (I, II, V) have been synthe-
sized via nucleophilic aromatic substitution of the new diols
4 or 8, 2,5-bis(4-hydroxyphenyl)pyridine and 3,30,5,50-tet-
ramethyl[1,10-biphenyl]-4,40-diol with bis(4-fluorophenyl)sul-
fone, as shown in Schemes 3 and 4. The chemical structures
of the new dimethoxy-based polymers were verified by 1H
NMR spectroscopy. High molecular weights were obtained in
all cases as proven by gel permeation chromatography
(GPC), which are shown in Table 1. The high molar percent-
age (50–60 mol %) of 2,5-bis(4-hydroxyphenyl)pyridine
monomer used in all synthesized polymers (I, II, V) should
guarantee the sufficient doping ability. On the other hand,
different molar ratios of hydroxyl groups were used in the
case of terpolymers V to study the effect of the amount of
hydroxyl groups on the crosslinking since high molar per-
centage of hydroxyl groups can lead to efficient crosslinking.
All dimethoxy bearing polymers were soluble in solvents,
such as CHCl3, dimethylacetamide (DMAc), dimethylforma-
mide (DMF), and N-Methylpyrrolidon (NMP) and showed
excellent film forming properties.
in
several
polycondensation
reactions
as
mono-
mer37,39,42,45,46 has been chosen in this report as crosslink-
ing agent because of its high thermal, chemical, and oxidative
stability. Furthermore, this compound was used to increase
membrane’s hydrophobicity, thus enabling the phase separa-
tion.4 The effect of the amount of the decafluorobiphenyl
crosslinker on the thermal and mechanical properties as well
as on acid doping ability of the crosslinked membranes was
studied. In all cases, crosslinked membranes showed
improved mechanical in terms of glass transition tempera-
tures (Tgs) as well as thermal properties and lower doping
levels when impregnated with phosphoric acid compared
with the non crosslinked ones.
RESULTS AND DISCUSSION
Aromatic polyethers bearing main chain and/or side chain
pyridine units,11–15 are promising candidates for application
in high temperature PEMFCs. However, to further improve
the mechanical strength and the thermal properties, which
are critical parameters of PEMs, and thus to increase the
long-term durability at temperatures of 200 ꢀC and above,
novel aromatic polyethers bearing side cross-linkable groups
have been synthesized. In this work, side dimethoxy-based
polymers have been prepared and modified to the hydroxyl
SCHEME 2 Synthesis of 1,4-bis(4-methoxyphenyl)-2,5-dihydroxy-
benzene (8).
208
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 207–216

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
SCHEME 3 Synthesis of dimethoxy-based copolymers (I, II) and dihydroxy-based analogs (III).
The aforementioned polymers were all modified to the side
hydroxy-based analogs by ether cleavage with boron tribro-
mide. However, in Table 1 only the dihydroxy-based poly-
mers (III, VI) that were used for crosslinking are listed. The
successful ether cleavage was confirmed by 1H NMR spec-
troscopy. A representative example for the ether cleavage of
dimethoxy bearing copolymers is as follows: the dimethoxy-
based precursor II exhibited a characteristic peak located at
3.81 ppm attributed to the methyl protons (not shown here).
After ether cleavage this peak is disappeared and a new
characteristic peak rose at 9.07 ppm which is attributed to
the hydroxyl groups (Fig. 1).
temperatures led to insoluble gelatine like material; a sur-
prising result in view of the fact that copolymer I and III
have very similar chemical structures. Becuase of this solu-
bility problem, crosslinking of copolymer I with decafluorobi-
phenyl was not feasible as well. 1H NMR spectrum of IV
showed the disappearance of the peak located at 9.07 ppm
attributed to hydroxyl protons and the formation of two new
peaks at 4.74 ppm and 3.50 ppm. These peaks are assigned
to the methylene protons and the terminal alkyne proton,
respectively, confirming the successful substitution of the
hydroxyl groups with propargyl moiety (Fig. 2). Further evi-
dence for a successful substitution is given by IR spectros-
copy, where a weak band at 3290 cmꢁ1 could be observed,
characteristic for terminal alkynes.
Propargyl-based copolymer IV has been synthesized by
nucleophilic substitution reaction of propargyl bromide with
dihydroxy copolymer III (Scheme 5). It should be noticed
that the preparation of propargyl-based copolymers was
only possible by using dihydroxy-based copolymer III, since
deprotonation of dihydroxy-based copolymer I at elevated
As already mentioned, covalent crosslinking results in mem-
branes with high chemical and thermal stabilities.53–59 Side
chain triple bonds can be crosslinked through the formation
of a network of stable benzene systems. Thermal curing of
SCHEME 4 Synthesis of dimethoxy-based terpolymers (V) and dihydroxy-based analogs (VI).
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 207–216
209

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
crosslinked copolymer IVCL is 325 ꢀC, which is considerably
higher than the Tg of copolymer II by 90 ꢀC. Furthermore,
the storage modulus after crosslinking was not decreased
above 370 ꢀC, but it reached a plateau and this behavior is
characteristic for crosslinked networks. The Tg of copolymer
IVCL is among the highest reported values for crosslinked
copolymers containing triple bonds: crosslinking of ethynyl-
phenyl terminated copolymers leads to a maximum Tg of
258 ꢀC,36 while in the case of side chain ethynyl bearing
copolymers there is an increase of the glass transitions tem-
TABLE 1 Characteristics of Synthesized Polymers
Polymer
x/y/z (mol %)
Mn
Mw
I
Copolymers PPy(x)MeO(y)coPSF
Ia
Ib
60/40/0
60/40/0
42,750
39,140
65,980
1.5
1.7
64,600
Copolymers PPy(x)MeOPh(y)coPSF
IIa
IIb
60/40/0
60/40/0
42,000
45,850
76,440
65,220
1.8
1.4
37
ꢀ
perature up to 312 C.
Copolymers PPy(x)OHPh(y)coPSF
III 60/40/0
Copolymers PPy(x)Prop(y)coPSF
IV 60/40/0
–
–
–
–
–
–
An indirect method to confirm a successful crosslinking is to
compare the weight percentage of the soluble fraction of non
crosslinked and crosslinked membranes. The non crosslinked
membranes II and III were completely soluble in DMAc after
heating at 60 ꢀC for 16 h, whereas the crosslinked mem-
brane IVCL was almost insoluble (only 11 wt % was soluble
under the same conditions), denoting the effective
crosslinking.
Terpolymers PPy(x)MeOPh(y)TM(z)coPSF
Va
Vb
Vc
Vd
55/34/11
56/33/11
50/28/22
52/12/36
70,340
53,880
102,680
39,750
111,180
89,080
161,420
65,380
1.6
1.7
1.6
1.7
Another critical parameter that was evaluated is the acid
doping ability of the crosslinked membranes. The doping
ability of the crosslinked membrane IVCL impregnated in
phosphoric acid at 80 and 100 ꢀC is shown in Figure 4. For
comparison reasons, the doping ability of the membrane of
copolymer II is presented as well. In general, it is known
that since crosslinking leads to the formation of more com-
pact chemical structures, the acid doping level as well as the
proton conductivity is decreased. As expected, crosslinked
membrane IVCL showed lower doping levels compared with
the one of copolymer II at 80 ꢀC and 100 ꢀC. More specifi-
cally, the crosslinked membrane IVCL had a doping level of
around 160 wt %, while the non crosslinked one had a dop-
Terpolymers PPy(x)OHPh(y)TM(z)coPSF
VI 52/12/36
–
–
–
copolymer IV led to a covalent crosslinked network which
could be confirmed, i.a. by the results of dynamic mechanical
analysis (DMA) (Fig. 3).
Since thermal crosslinking can be induced even by drying
the membranes under vacuum at 160 C or during the DMA
ꢀ
experiments, comparison of the glass transition temperatures
of the propargyl copolymer IV with the crosslinked one IVCL
is not reliable. The methoxy-based copolymer II can be used
as an alternate for comparison of the mechanical properties,
since it contains no cross-linkable groups. As it is shown in
Figure 3, the crosslinked copolymer IVCL exhibited higher
glass transition temperature (Tg) compared with copolymer
II, as it was expected. The improvement of the Tgs is contrib-
uted to the decreased flexibility of the polymer chains by the
ꢀ
ing level of around 290 wt % after 24 h at 100 C.
At higher temperatures (120 ꢀC), the crosslinked membrane
partially started to dissolve. This behavior could be
explained by the presence of the alkyl ether groups, which
are sensitive for protonation by the acid resulting in the for-
mation of an ionic species that can increase the solubility of
the membrane in phosphoric acid at elevated temperatures.
formation of
a crosslinked network structure through
the thermal curing. The glass transition temperature of the
FIGURE 1 1H NMR of copolymer III in deuterated dimethylsulfoxide with the assignment of the peaks.
210
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 207–216

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
SCHEME 5 Synthesis of propargyl-based copolymer (IV).
FIGURE 3 DMA curves of crosslinked copolymer IVCL com-
pared with dimethoxy-based copolymer II.
Another very promising crosslinking method for the forma-
tion of covalent networks is the nucleophilic substitution of
dihalides or polyhalides, e.g., the crosslinking of PBI with p-
xylene dichloride43 and dibromide,44 respectively. In this
study, decafluorobiphenyl was used as a crosslinker because
of its high thermal, chemical, oxidative stability as well as its
increased hydrophobicity. crosslinking with the decafluorobi-
phenyl compound can lead to more stable aryl ether bonds
compared with alkyl ethers, which would be created in the
case of crosslinking with xylene dihalides. As shown before
in the case of copolymer IVCL, alkyl ether groups could be
the reason for membrane’s dissolution processes in phos-
phoric acid at elevated temperatures due to its protonation.
Furthermore, the method with a crosslinker enables the con-
trol of the crosslinking density by varying the molar ratio of
hydroxyl groups as well as the weight percentage of the
decafluorobiphenyl used as crosslinker.
crosslinking, have been evaluated. Crosslinked membranes
with different weight percentages of decafluorobiphenyl used
as crosslinker have been prepared, as it is shown in Table 2.
As the weight percentage of the crosslinker increases, the
mechanical integrity of the crosslinked membranes as well
as the acid doping ability is decreased. More specifically,
crosslinked membrane III5CL prepared from copolymer III
with composition 60/40 with 5 wt % crosslinker exhibited
good mechanical properties, while crosslinked membrane
III10CL with 10 wt % crosslinker became more fragile. Prepa-
ration of free standing crosslinked membranes with 20 wt %
crosslinker was not possible. Crosslinked membrane VI6CL
prepared from dihydroxy terpolymer VI with composition
52/12/36 and 6 wt % crosslinker showed very good me-
chanical integrity. Crosslinked samples prepared from terpol-
ymer with composition 56/33/11 with 6 and 7 wt % cross-
linker have been prepared as well (not listed in Table 2). As
expected, the doping level of the crosslinked membrane with
the higher weight percentage of decafluorobiphenyl (7 wt %)
was lower. As mentioned before, the effect of the different
Dihydroxy-based polymers (III, VI) reacted with decafluoro-
biphenyl resulted in the formation of a covalently bonding
network. To control the degree of crosslinking, two different
parameters, the weight percentage of the crosslinker as well
as the molar percentage of hydroxyl groups involved in the
FIGURE 2 1H NMR of copolymer IV in deuterated dimethylsulf-
FIGURE 4 Time dependence of doping level (wt %) of II and
IVCL at 80 ꢀC and 100 ꢀC in 85 % H3PO4.
oxide with the assignment of the peaks.
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 207–216
211

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
TABLE 2 Characteristics of Crosslinked and Non Crosslinked Membranes
Soluble
Doping
Level
(wt %)b
x/y/z
Crosslinker
(wt %)
Film
Fraction
(wt %)a
Polymer
(mol %)
Tg (ꢀC)
Quality
III
60/40/0
60/40/0
60/40/0
52/12/36
52/12/36
0
5
250
320
–
þþþ
þþ
100
8
200
180
150
180
150
III5CL
III10CL
VI
10
0
þ
6
284
301
þþþ
þþ
100
11
VI6CL
6
a
Soluble fraction in DMAc after heating at 60 ꢀC for 16 h.
24 h at 100 ꢀC.
b
þþþ Very good.
þþ Good.
þ Bad.
molar percentage of the hydroxyl groups on crosslinking has
been studied as well. Comparing polymers with composition
40 mol % (III5CL) and 12 mol % (VI6CL) of hydroxyl containing
units, respectively, while keeping almost the same amount of
decafluorobiphenyl, it is shown that crosslinking was more
effective in the case of III5CL. This observation is supported by
the DMA data, since in the case of terpolymer with 12 mol %
hydroxyl containing units (VI), the difference in Tg values of
the crosslinked and the non crosslinked membrane was only
16 ꢀC, while in the case of copolymer with 40 mol % hydroxyl
containing units (III) was 70 ꢀC (Table 2). Measurements of
the weight percentage of the soluble fraction of the cross-
linked membranes were also conducted as it can be seen in
(Table 3). In all cases, the crosslinked membranes exhibited
higher thermal stability compared with the corresponding
non crosslinked ones. Tꢀhe crosslinked membranes were ther-
mally stable up to 410 C, while for the noncrosslinked ones,
a small weight loss has been observed at this temperature.
At temperatures above 550 ꢀC, the improved thermal stabil-
ity of the crosslinked membranes is evident, since crosslink-
ing enhances the thermal stability of the polymer main
chain. In details, the crosslinked III5CL showed a Td(5%) at
490 ꢀC while VI6CL had a decomposition temperature Td(5%)
at 340 ꢀC. The higher thermal stability of crosslinked III5CL
compared with VI6CL confirmed the more effective crosslink-
ing as it was also supported by the previous data obtained
by DMA analysis.
Table 2. The crosslinked membranes (III5CL, III10CL, VI6CL
)
The doping ability of the crosslinked membranes as well as
the non crosslinked ones in phosphoric acid at 100ꢀ C is
given in Figure 5. As expected, crosslinked membranes
exhibited lower doping level compared with the non cross-
linked ones. As already mentioned, since crosslinking was
more effective in the case of III5CL, should result in lower
were almost insoluble, only 6–11 wt % was soluble in DMAc
after heating at 60 ꢀC for 16 h, while the non crosslinked
membranes were completely soluble, indicating the successful
crosslinking.
The mechanical properties of the crosslinked membrane in
comparison with the non crosslinked one were studied by
DMA (Table 2). The crosslinked membrane III5CL with 5 wt %
crosslinker exhibited higher glass transition temperature com-
pared with the non crosslinked one III. The Tg of the cross-
linked membrane III5CL was 320 ꢀC, by 70 ꢀC higher compared
with the non crosslinked one III (Tg around 250 ꢀC).
The thermal stability of the crosslinked polymers was eval-
uated by thermogravimetric analysis (TGA). The decomposi-
tion behavior for two different crosslinked (IIICL5 and VI6CL
)
and one non crosslinked membrane III has been compared
TABLE 3 TGA Data of Crosslinked Polymers III5CL, VI6CL, and
the Non Crosslinked One III
Weight
Weight
Loss (wt %)
at 500 ꢀC
Loss (wt %)
at 800 ꢀC
Polymer
Td(5%) (ꢀC)
III
440
490
340
15
5
51
27
42
FIGURE 5 Comparison of the doping ability of the crosslinked
membranes III5CL and VI6CL with the corresponding non cross-
linked ones III and VI with 85% H3PO4 at 100 ꢀC.
III5CL
VI6CL
20
212
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 207–216

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
Instrumentation
1H NMR spectra were obtained on a Bruker Advance DPX
400 MHz spectrometer. The samples were dissolved either in
deuterated chloroform or dimethylsulfoxide with TMS used
as internal standard. GPC measurements were performed on
a Polymer Lab chromatograph equipped with two Plgel 5 lm
mixed columns and a UV detector (254 nm), using CHCl3 as
eluent with a flow rate of 1 mL/min at 25 ꢀC and polysty-
rene standards. DMA were carried out using a solid-state an-
alyzer RSA II (Rheometrics Scientific Ltd.) at 10 Hz. TGA was
performed on a Labsys Tg of Setaram under nitrogen with a
heating rate of 10 ꢀC /min. Fourier transform infrared
(FTIR) spectra were recorded on a Perkin-Elmer 1600 FTIR
spectrometer.
Monomer Synthesis
Synthesis of Monomer 4
FIGURE 6 Temperature dependence of doping ability of III5CL
Synthesis of 1,4-bis(4-(tetrahydropyran-2-yloxyphenyl))-
with 85% H3PO4.
2,5-dimethoxybenzene
(3). 2,5-Dimethoxy-1,4-phenylene
diboronic acid (1) (2.00 g, 8.86 mmol), aqueous solution of
2 M Na2CO3 (35.4 mL) and DMF (65 mL) were degassed
under vacuum. 1-Bromo-(4-tetrahydropyran-2-yloxy)benzene
(2) (9.11 g, 35.43 mmol) and PdCl2(PPh3)2 (0.373 g, 0.53
mmol) were added and the reaction mixture was heated at
155 ꢀC for 5 d. After cooling to room temperature, the mix-
ture was added to ice/water, stirred for 0.5 h, filtered, and
washed with water and hexane. The solid was dissolved in a
small amount of CH2Cl2 and filtered to remove the catalyst.
The solvent was removed, MeOH was added and stirred
overnight. The white product was filtered, washed with
doping levels compared with the crosslinked membrane
VI6CL. In contrary, the crosslinked membrane III5CL showed
higher doping levels (180 wt %) compared with the cross-
linked membrane VI6CL (150 wt %). A possible explanation
for this finding is the higher molar percentage of hydroxyl
containing units (40 mol % for III5CL) that can interact with
phosphoric acid, thus resulting in higher doping levels. The
same observation can be made with the non crosslinked
membranes where membrane III with 40 mol % of hydroxyl
containing units had a doping level of 200 wt %, while the
corresponding VI with 12 mol % had 180 wt %.
ꢀ
MeOH and dried under vacuum at 60 C. 1.62 g (37 % yield)
was received from compound 3.
Although the obtained doping levels of the crosslinked mem-
branes are lower compared with the non cross linked ones,
they are still acceptable for membrane’s use as polymer elec-
trolytes for high temperature PEMFCs. To further enhance
the doping ability and thus the ionic conductivity of the
crosslinked membranes, the doping abilityꢀ has been eval-
uated at higher temperatures (120 and 140 C) as well. As it
is shown in Figure 6, the doping ability of crosslinked III5CL
was improved with increasing temperature. In details, III5CL
showed a maximum doping level of 23ꢀ0 wt % at 140 ꢀC
after 24 h, while the doping level at 100 C was 180 wt %.
1H NMR (CDCl3, ppm): 1.61–1.75 (m, 6 H), 1.87–1.93 (m, 4
H), 2.00–2.08 (m, 2 H), 3.61–3.66 (m, 2 H), 3.78 (s, 6 H),
3.94–3.98 (t, 2 H), 5.48 (s, 2 H), 6.95 (s, 2 H), 7.11–7.13 (dd,
4 H), 7.51–7.53 (dd, 4 H).
Synthesis of 1,4-bis(4-hydroxyphenyl)-2,5-dimethoxyben-
zene (4). To a solution of 3 (1.62 g, 3.30 mmol) in tetrahy-
drofuran (50 mL) was added 37 % HCl (0.8 mL) and then
the mixture was stirred overnight at room temperature. The
solvent was reduced under pressure and the mixture was
added to ice/water. The solid was filtered, washed with
water and hexane and dried under vacuum at 60 ꢀC. 1.04 g
(98 % yield) was received from the desired monomer 4.
Crosslinked membranes obtained according to the here pre-
sented methodology are currently optimized to be applied in
membrane electrode assembly construction and single cell
testing.
1H NMR (DMSO-d6, ppm): 3.74 (s, 6 H), 6.79–6.82 (dd, 4 H),
6.92 (s, 2 H), 7.36–7.38 (dd, 4 H), 9.45 (s, 2 H).
Synthesis of Monomer 8
EXPERIMENTAL SECTION
Synthesis of 1,4-bis(4-methoxyphenyl)-2,5-(tetrahydro-
pyran-2-yloxy) benzene (7). To a degassed mixture of 4-
methoxyphenylboronic acid (6) (4.90 g, 32.24 mmol), aque-
ous 2 M Na2CO3 (27 mL) and DMF (50 mL) were added 2,20-
((2,5-dibromo-1,4-phenylene)bis(oxy))bis(tetrahydropyran) (5)
(4.69 g, 10.75 mmol) and Pd(PPh3)4 (0.75 g, 0.65 mmol). The
reaction mixture was heated at 125 C for 3 d. After cooling to
room temperature, the mixture was added to ice/water, stirred
for 0.5 h, filtered and washed with water and hexane. The solid
Materials
2,5-Bis(4-hydroxyphenyl)pyridine,11 2,20-((2,5-dibromo-1,4-
phenylene)bis(oxy))bis-(tetrahydropyran),14 2,5-dimethoxy-
1,4-phenylene diboronic acid,40 4-methoxyphenylboronic
acid60 and 1-bromo-(4-tetrahydropyran-2-yloxy)benzene61,62
were prepared according to literature. All other chemicals
and solvents were purchased from Aldrich without any fur-
ther purification.
ꢀ
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 207–216
213

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
was dissolved in CH2Cl2 and filtered to remove the catalyst.
The solvent was removed and a small amount of MeOH added
to the residue and stirred overnight. The product was filtered,
washed with MeOH and dried under vacuum at 50 ꢀC. The
crude product 7 was used as received in the next step.
(0.3684g, 1.52 mmol), K2CO3 (0.8116 g, 5.880 mmol),ꢀ tolu-
ene (4.5 mL) and DMF (9 mL) was stirred at 150–160 C for
ꢀ
17 h and at 180 C for 3 h. The viscous solutionꢀwas precipi-
tated in MeOH/H2O 2/1, stirred in water at 60 C and dried
ꢀ
under vacuum at 160 C for 1 d. The results are summarized
in Table 1.
Synthesis of 1,4-bis(4-methoxyphenyl)-2,5-dihydroxyben-
zene (8). To a solution of 7 (3.75 g, 7.65 mmol) in tetrahy-
drofuran (80 mL) was added 37 % HCl (10 mL), and then
the mixture was stirred overnight at room temperature. The
solution was extracted with ethyl acetate, the organic layer
was washed three times with water, dried over MgSO4 and
evaporated. The residue was purified by recrystallization in
toluene and 1.72 g (70 % yield) of the desired monomer 8
was received.
Synthesis of Dihydroxy-Based Polymers
Synthesis of PPy(x)OHPh(y)coPSF copolymers (III)
A representative example for the preparation of copolymers
III is as follows. Boron tribromide (0.35 mL, 3.59 mmol) was
added slowly via a septum to a solution of copolymer II
(1.00 g, 2.00 mmol) in dry CH2Cl2 whereat immediately a
solid was formed. The reaction mixture was stirred overnight
and added to a saturated NaHCO3 solution. Because volumi-
nous agglomerates were formed, after filtration and washing
with water and hexane, for further purification the product
was dissolved in hot DMAc and poured into water to give a
white solid of III in a yield ofꢀ95 % (0.93 g, 1.90 mmol) after
drying under vacuum at 100 C for 1 d. The same procedure
was followed for the preparation of terpolymers VI.
1H NMR (DMSO-d6, ppm): 3.78 (s, 6 H), 6.81 (s, 2 H), 6.96–
6.98 (dd, 4 H), 7.48–7.50 (dd, 4 H), 8.82 (s, 2 H).
Polymer Synthesis
Synthesis of Dimethoxy-Based Polymers
Synthesis of PPy(x)MeO(y)coPSF copolymers (I). A typi-
cal polymerization is as follows. To
a degassed flask
Synthesis of Propargyl-Based Copolymers
equipped with a Dean Stark trap was added a mixture of
2,5-bis(4-hydroxyphenyl)pyridine (0.6126 g, 2.33 mmol), 1,4-
bis(4-hydroxyphenyl)-2,5-dimethoxybenzene (4) (0.5000 g,
1.55 mmol), bis(4-fluorophenyl)sulfone (0.9859 g, 3.88
mmol), K2CO3 (0.622 g, 4.50 mmol), toluene (7 mL) and
DMF (20 mL). The reaction mixture was stirred at 150 ꢀC
for 16 h under argon. After removal of the azeotropic mix-
ture of the formed wateꢀr with toluene, the temperature was
raised gradually at 180 C. The reaction mixture was stirred
at this temperature for 20 h. The obtained viscous solution
was poured into MeOH, where a white solid was precipi-
tated. The solid was washed with MeOH, filtered, stirred in
water at 60 ꢀC and drying at 80 ꢀC under vacuum for 1 d.
Copolymers characteristics are given in Table 1.
Synthesis of PPy(x)Prop(y)coPSF Copolymers (IV)
K2CO3 (0.127 g, 0.92 mmol) was added to a solution of co-
polymer III (0.150 g, 0.31 mmol) in dry DMF (9 mL). After
stirring for 1.5 h at 55 ꢀC, the reaction mixture was cooled
to room temperature (the flask was wrapped with foil to
avoid light because of the light sensitivity of propargyl bro-
mide) and 80 % propargyl bromide in toluene (0.04 mL,
0.37 mmol) was added. After 4 d stirring at room tempera-
ture, the mixture was dropped in an ice/water mixture,
stirred for 1 h, filtered, washed with water and hexane and
ꢀ
dried under vacuum at 30 C for 1 d.
Membrane Preparation
Membranes were prepared by solution-casting method. The
dried polymers were dissolved in DMAc at a concentration 3
ꢀ
Synthesis of PPy(x)MeOPh(y)coPSF copolymers (II). The
same procedure as described for copolymer I was followed
for the synthesis of copolymer II using monomer 8 instead
of 4. A mixture of 2,5-bis(4-hydroxyphenyl)pyridine (0.3676
g, 1.40 mmol), bis(4-fluorophenyl)sulfone (0.5916 g, 2.33
mmol), 1,4-bis(4-methoxyphenyl)-2,5-dihydroxybenzene (8)
(0.3000 g, 0.93 mmol), K2CO3 (0.373 g, 2.70 mmol), toluene
(4.5 mL) and dimethylacetamide (DMAc) (13 mL) was
stirred at 150 ꢀC for 16 h and at 180 ꢀC for 22 h. The vis-
cous solution was precipitated in MeOH. The solid wꢀas
wt %, filtered and poured onto a glass plate at 80 C, where
the solvent evaporated slowly. To remove any traces of the sol-
vent, membranes were dried at 160 ꢀC under vacuum for 3 d.
Crosslinking Procedures
Crosslinking of Copolymer IV by Thermal Treatment
Copolymer IV (0.129 g, 0.25 mmol) was dissolved in DMAc
and casted at 80 ꢀC overnight. The membrane was heated
ꢀ
gradually from 50 to 200 C on a hotplate for 1 h, and then
for 2 h more at 250 ꢀC to complete the crosslinking reac-
tion.37 To remove any solvent residues, the membrane was
ꢀ
stirred in water at 60 C and dried under vacuum at 100 C
ꢀ
dried under vacuum for 3 d at 160 C.
for 1 d. Copolymers characteristics are given in Table 1.
Synthesis of PPy(x)MeOPh(y)TM(z)coPSF terpolymers
(V). The same procedure was followed for the synthesis of
terpolymers V by using three diols, the 1,4-bis(4-methoxy-
phenyl)-2,5-dihydroxybenzene (8), 2,5-bis(4-hydroxyphenyl)-
pyridine and 3,30,5,50-tetramethyl[1,10-biphenyl]-4,40-diol. A
mixture of 2,5-bis(4-hydroxyphenyl)pyridine (0.8000 g, 3.04
mmol), bis(4-fluorophenyl)sulfone (1.2876 g, 5.07 mmol),
1,4-bis(4-methoxyphenyl)-2,5-dihydroxybenzene (8) (0.1631
g, 0.51 mmol), 3,30,5,50-tetramethyl[1,10-biphenyl]-4,40-diol
Crosslinking of Copolymer III with Decafluorobiphenyl
Used as Crosslinker
A representative example of crosslinking with decafluorobi-
phenyl is the following. Copolymer III (0.146 g, 0.30 mmol)
was dissolved in DMAc (3 mL), K2CO3 (0.048 g, 0.34 mmol)
was added and the reaction mixture was stirred at 55–60 C
for 1.5 h. Decafluorobiphenyl (0.0073 g, 0.02 mmol) dis-
solved in DMAc (0.5 mL) was added to the solution, which
ꢀ
was filtered and then poured onto
a glass substrate.
214
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 207–216

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
REFERENCES AND NOTES
Crosslinked membranes were obtained following two-step
procedure: first the glass plate was covered in an oven for
16 h at 80 ꢀC, then the cover was removed and the mem-
brane was obtained by casting at 120 ꢀC. To remove any
excess of the solvent,ꢀ the membranes were dried for 3 d
under vacuum at 160 C.
1 Steele, B. C.; Heinzel, A. Nature 2001, 414, 345–352.
2 Kreuer, K. D. J. Membr Sci 2001, 185, 29–39.
3 Hickner, M. A.; Ghassemi, H.; Kim, Y. S.; Einsla, B. R.; Mc
Grath, J. E. Chem Rev 2004, 104, 4587–4612.
4 Mauritz, K. A.; Moore, R. B. Chem Rev 2004, 104, 4535–4586.
5 Kreuer, K. D.; Paddison, S. J.; Spohr, E.; Schuster, M. Chem
Rev 2004, 104, 4637–4678.
Crosslinking of Terpolymer VI with Decafluorobiphenyl
Used as Crosslinker
6 Li, Q.; He, R. H.; Jensen, J. O.; Bjerrum, N. J Chem Mater
2003, 15, 4896–4915.
Terpolymer VI (0.2829 g, 0.60 mmol) was dissolved in DMAc
(15 mL) after heating at 70 ꢀC for 17 h. K2CO3 (0.094 g,
0.68 mmol) was added and the reaction mixture was stirred
7 Wainright, J. S.; Wang, J. T.; Weng, D.; Savinell, R. F.; Litt,
M. J Electrochem Soc 1995, 142, L121–L123.
ꢀ
at 60 C for 1.5 h. Decafluorobiphenyl (0.0169 g, 0.05 mmol)
8 Ma, Y. L.; Wainright, J. S.; Litt, M.; Savinell, R. F. J Electro-
chem Soc 2004, 151, A8–A16.
dissolved in DMAc (0.5 mL) was added to the solution,
which was filtered and then poured onto a glass substrate.
9 Carollo, A.; Quartarone, E.; Tomasi, C.; Mustarelli, P.; Belotti,
F.; Magistris, A.; Maestroni, F.; Parachini, M.; Garlaschelli, L.;
Righetti, P. P. J Power Sources 2006, 160, 175–180.
ꢀ
The mixture was left covered for 16 h at 80 C in the oven.
After removing the cover, the membrane was obtained by
ꢀ
casting at 120 C. The membranes were dried for 3 d under
10 Li, Q.; Jensen, J. O.; Savinell, R. F.; Bjerrum, N. J Progr
Polym Sci 2009, 34, 449–477.
ꢀ
vacuum at 160 C.
11 Gourdoupi, N.; Andreopoulou, K.; Deimede, V.; Kallitsis, J.
K. Chem Mater 2003, 15, 5044–5050.
Doping Procedure
Membranes were immersed in 85% phosphoric acid at dif-
ferent temperatures for different doping times, to study the
doping behavior. After each immersion, the samples were
wiped with paper and the weight gain was calculated from
the sample’s weight before and after the immersion. The
doping level is defined as the weight percent of the acid
(85% H3PO4) per gram of each polymer.
12 Pefkianakis, E. K.; Deimede, V.; Daletou, M. K.; Gourdoupi,
N.; Kallitsis, J. K. Macromol Rapid Commun 2005, 26, 1724–
1728.
13 Kallitsis, J. K.; Geormezi, M.; Neophytides, S. G. Polym Int
2009, 58, 1226–1233.
14 Geormezi, M.; Chochos, C.; Gourdoupi, N.; Neophytides, S.
G.; Kallitsis, J. K. J Power Sources 2011, 196, 9382–9390.
15 Morfopoulou, C.; Andreopoulou, A. K.; Kallitsis, J. K.
J Polym Sci Part A: Polym Chem 2011, 49, 4325–4334.
CONCLUSIONS
16 Qingfeng, L.; Hjuler, H. A.; Hasiotis, C.; Kallitsis, J. K.; Kon-
toyannis, C. G.; Bjerrum, N. J. Electrochem Solid-State Lett
2002, 5, A125–A128.
Aromatic polyethers bearing polar pyridine units in the main
chain and side chain cross-linkable hydroxyl and propargyl
groups have been successfully prepared. The synthesized
polymers showed high molecular weights, high Tg values and
thermal stability, as well as high doping levels when doped
in phosphoric acid. Crosslinked membranes could be
obtained either by direct crosslinking of hydroxyl groups
with crosslinker or by further functionalization of hydroxyl
groups to the propargyl analog. Crosslinking of propargyl-
based copolymers by thermal treatment leads to the forma-
tion of a stable network. The obtained crosslinked mem-
branes showed very high glass transition temperatures up to
17 Avgouropoulos, G.; Papavasiliou, J.; Daletou, M. K.; Kallit-
sis, J. K.; Ioannides, T.; Neophytides, S. G. Appl Catal-B: Envi-
ron 2009, 90, 628–632.
18 Avgouropoulos, G.; Ioannides, T.; Kallitsis, J. K.; Neophyti-
des, S. G. Chem Eng J 2011, in press.
19 Kerres, J.; Ullrich, A.; Meier, F.; Haring, T. Solid State Ionics
¨
1999, 125, 243–249.
20 Kerres, J. A. Fuel Cells 2005, 5, 230–247.
21 Deimede, V.; Voyiatzis, G. A.; Kallitsis, J. K.; Qingfeng, L.;
Bjerrum, N. J. Macromolecules 2000, 33, 7609–7617.
ꢀ
22 Hofmann, M. A.; Ambler, C. M.; Maher, A. E.; Chalkova, E.;
Zhou, X. Y.; Lvov, S. N.; Allock, H. R. Macromolecules 2002, 35,
6490–6493.
325 C. The other method followed for covalent crosslinking
is the nucleophilic substitution of polyhalides, such as decaf-
luorobiphenyl with hydroxyl groups. In all cases, crosslinked
membranes showed improved mechanical in terms of glass
transition temperatures (Tgs) and thermal properties as well
as lower doping levels compared with the non crosslinked
ones. The obtained phosphoric acid doping levels although
lower compared with the non crosslinked ones are still in
the range where useful high temperature electrolytes can be
provided.
23 Ding, F. C.; Wang, S. J.; Xiao, M.; Li, X. H.; Meng, Y. Z.
J Power Sources 2007, 170, 20–27.
24 Ding, F. C.; Wang, S. J.; Xiao, M.; Meng, Y. Z. J Power Sour-
ces 2007, 164, 488–495.
25 Feng, S.; Shang, Y.; Xie, X.; Wang, Y.; Xu, J. J Membr Sci
2009, 335, 13–20.
26 Huang, Y.; Pan, Y.; Fu, J.; Huang, X.; Tang, X. J Appl Polym
Sci 2009, 113, 2353–2360.
Financial support of this work from the European Commission
through the program ‘‘Development of an Internal Reforming
Alcohol High Temperature PEM Fuel Cell Stack’’, IRAFC FCH-JU
245202 (2010–2012) is greatly acknowledged. The authors
thank Mr. G. Paloumbis MSc from Advent Technologies S. A. for
his help in monomer synthesis.
27 Papadimitriou, K. D.; Paloukis, F.; Neophytides, S. G.; Kallitsis,
J. K. Macromolecules 2011, 44, 4942–4951.
28 Jia, J.; Baker, G. L. J Polym Sci Part B: Polym Phys 1998,
36, 959–968.
29 Kelman, S. D.; Matteucci, S.; Bielawski, C. W.; Freeman, B.
D. Polymer 2007, 48, 6881–6892.
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 207–216
215

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
30 Oh, Y.-S.; Lee, H.-J.; Yoo, M.; Kim, H.-J.; Han, J.; Kim, T.-H.
47 Percec, V.; Clough, R. S.; Rinaldi, P. L.; Litman, V. E. Macro-
J Membr Sci 2008, 323, 309–315.
molecules 1991, 24, 5889–5892.
31 Yasuda, N.; Yamamoto, S.; Wada, Y.; Yanagida, S. J Polym
48 Percec, V.; Wang, J. H.; Clough, R. S. Makromol. Chem.,
Sci Part A: Polym Chem 2001, 39, 4196–4205.
Macromol. Symp. 1992, 54/55, 275–312.
32 Oh, Y.-S.; Lee, H.-J.; Yoo, M.; Kim, H.-J.; Han, J.; Kim, K.;
49 Percec, V.; Clough, R. S.; Grigoras, M.; Rinaldi, P. L.; Litman,
Hong, J. D.; Kim, T.-H. Chem Commun 2008, 2028–2030.
V. E. Macromolecules 1993, 26, 3650–3662.
33 Acierno, D.; Di Maio, L.; Iannelli, P.; Spadaro, G.; Valenza, A.
50 Percec, V.; Grigoras, M.; Clough, R. S.; Fanjul, J. J Polym
Polymer 2000, 41, 6647–6654.
Sci Part A: Polym Chem 1995, 33, 331–344.
34 Qiang, L.; Ma, Z.; Zheng, Z.; Yin, R.; Huang, W. Macromol.
51 Nakabayashi, K.; Higashihara, T.; Ueda, M. Macromolecules
Rapid Commun 2006, 27, 1779–1786.
2011, 44, 1603–1609.
35 Zhong, S.; Liu, C.; Na, H. J Membr Sci 2009, 326, 400–407.
52 Kakali, F.; Kallitsis, J. K.; Pakula, T.; Wegner, G. Macromole-
cules 1998, 31, 6190–6198.
36 Jeong, M.-H.; Lee, K.-S.; Lee, J.-S. J Membr Sci 2009, 337,
145–152.
53 Rhim, J. W.; Park, H. B.; Lee, C. S.; Jun, J. H.; Kim, D. S.;
Lee, Y. M. J Membr Sci 2004, 238, 143–151.
37 Jeong, M.-H.; Lee, K.-S.; Lee, J.-S. Macromolecules 2009,
42, 1652–1658.
54 Mikhailenko, S. D.; Wang, K.; Kaliaguine, S.; Xing, P.; Rob-
ertson, G. P.; Guiver, M. D. J Membr Sci 2004, 233, 93–99.
38 Kim, J.-P.; Lee, W.-Y.; Kang, J.-W.; Kwon, S.-K.; Kim, J.-J.;
Lee, J.-S. Macromolecules 2001, 34, 7817–7821.
55 Zhang, W.; Gogel, V.; Friedrich, K. A.; Kerres, J. J Power
Sources 2006, 155, 3–12.
39 Lee, K.-S.; Jeong, M.-H.; Lee, J.-P.; Lee, J.-S. Macromole-
cules 2009, 42, 584–590.
56 Lee, C. H.; Park, H. B.; Chung, Y. S.; Lee, Y. M.; Freeman, B.
D. Macromolecules 2006, 39, 755–764.
40 Lee, K.-S.; Lee, J.-S. Chem Mater 2006, 18, 4519–4525.
57 Heo, K. B.; Lee, H. J.; Kim, H. J.; Kim, B. S.; Lee, S. Y.; Cho,
E.; Oh, I. H.; Hong, S. A.; Lim, T. H. J Power Sources 2007, 172,
215–219.
41 Lee, H.-J.; Lee, M.-H.; Oh, M.-C.; Ahn, J.-H.; Han, S. G.
J Polym Sci Part A: Polym Chem 1999, 37, 2355–2361.
42 Lee, H.-J.; Lee, E.-M.; Lee, M.-H.; Oh, M.-C.; Ahn, J.-H.; Han,
S. G.; Kim, H. G. J Polym Sci Part A: Polym Chem 1998, 36,
2881–2887.
58 Zhang, C.; Guo, X.; Fang, J.; Xu, H.; Yuan, M.; Chen, B.
J Power Sources 2007, 170, 42–45.
59 Zhong, S.; Cui, X.; Cai, H.; Fu, T.; Zhao, C.; Na, H. J Power
43 Wang, K. Y.; Xiao, Y. C.; Chung, T.-S. Chem Eng Sci 2006,
Sources 2007, 164, 65–72.
61, 5807–5817.
60 Huang, W. Y.; Huang, S. Y. Macromolecules 2010, 43,
44 Li, Q.; Pan, C.; Jensen, J. O.; Noye, P.; Bjerrum, N. J. Chem
10355–10365.
Mater 2007, 19, 350–352.
61 Andreopoulou, A. K.; Kallitsis, J. K. Macromolecules 2002,
45 Liu, B.; Robertson, G. P.; Guiver, M. D.; Shi, Z.; Navessin, T.;
35, 5808–5815.
Holdcroft, S. Macromol Rapid Commun 2006, 27, 1411–1417.
46 Miyatake, K.; Hay, A. S. J Polym Sci Part A: Polym Chem
62 Andreopoulou, A. K.; Carbonnier, B.; Kallitsis, J. K.; Pakula,
2001, 39, 3770–3779.
T. Macromolecules 2004, 37, 3576–3587.
216
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 207–216