Polymer electrolyte membrane based on polyacrylate with phosphonic acid via long alkyl side chains

Article abstract of DOI:10.1039/c2ta00537a

A series of novel phosphonated polymers with phosphonic acid groups via different lengths of flexible pendant side chains was synthesized. The radical polymerization of the corresponding diethyl esters of acrylate monomers, followed by hydrolysis with trimethylsilyl bromide produced the expected polyacrylates. Among the three polymers, the cross-linked poly[6-(acryloyloxy) hexylphosphonic acid] (PAHPA) membrane prepared by the recombination of polymer radicals with benzoyl peroxide as a radical initiator showed excellent proton conductivity comparable to that of the Nafion 117 membrane in the range of 30-80% relative humidity (RH) at 80 °C, regardless of the significant low water uptake behavior. Furthermore, well-defined phase separation between the hydrophobic and hydrophilic domains was clearly observed by scanning transmission electron microscopy (STEM) and synchrotron X-ray scattering measurements. To the best of our knowledge, the cross-linked PAHPA membrane shows the best proton conductivity performance among phosphonated polymers already reported. The Royal Society of Chemistry 2013.

Full text of DOI:10.1039/c2ta00537a

Journal of
Materials Chemistry A
View Article Online
View Journal | View Issue
PAPER
Polymer electrolyte membrane based on polyacrylate
with phosphonic acid via long alkyl side chains†
Cite this: J. Mater. Chem. A, 2013, 1,
457
1
ab
a
a
c
Tomoya Higashihara, Namiko Fukuzaki, Youko Tamura, Yecheol Rho,
a
d
d
c
Kazuhiro Nakabayashi, Satoshi Nakazawa, Shigeaki Murata, Moonhor Ree*
and Mitsuru Ueda*a
A series of novel phosphonated polymers with phosphonic acid groups via different lengths of flexible
pendant side chains was synthesized. The radical polymerization of the corresponding diethyl esters of
acrylate monomers, followed by hydrolysis with trimethylsilyl bromide produced the expected
polyacrylates. Among the three polymers, the cross-linked poly[6-(acryloyloxy)hexylphosphonic acid]
(
PAHPA) membrane prepared by the recombination of polymer radicals with benzoyl peroxide as a
radical initiator showed excellent proton conductivity comparable to that of the Nafion
ꢀ
1
17 membrane in the range of 30–80% relative humidity (RH) at 80 C, regardless of the significant
low water uptake behavior. Furthermore, well-defined phase separation between the hydrophobic
Received 3rd October 2012
and hydrophilic domains was clearly observed by scanning transmission electron microscopy
Accepted 19th November 2012
(
STEM) and synchrotron X-ray scattering measurements. To the best of our knowledge, the cross-linked
DOI: 10.1039/c2ta00537a
PAHPA membrane shows the best proton conductivity performance among phosphonated polymers
already reported.
www.rsc.org/MaterialsA
ꢀ
function at high temperature (>100 C) and in nominally dry
1
Introduction
13–20
conditions, has been recently investigated.
In previous work, we focused on a high ion exchange capacity
Phosphonated polymers are of current interest as materials for
polymer electrolyte membranes (PEMs). Unlike sulfonated (IEC) value and prepared phosphonated poly(N-phenyl-
21
ꢁ
1
polymers which transport protons by water-assisted systems acrylamide) with an IEC value of 6.72 mequiv g (PDPAA).
1
,2
(the vehicle mechanism ), such as sulfonated per-
Although the cross-linked PDPAA membrane showed high
proton conductivity comparable to the Naon 112 membrane
under high RH due to the high IEC value, a drastic decrease in
3,4

uoropolymers (e.g., Naonꢀ and Flemionꢀ) and sulfonated
5–9
hydrocarbon polymers,
phosphonated polymers achieve
22
proton transportation from high self-dissociation and proton conductivity was clearly observed with decreasing RH.
hydrogen-bonding networks (the Grotthuss mechanism ) even That is, the proton conductivity behavior depended on not only
1,2
10–13
under anhydrous conditions.
Thus, the development of the hydrogen-bonding network but also on a water-assisted
PEMs based on phosphonated polymers, which can successfully system. The speculation that we obtained from these
phenomena is as follows: water-assisted proton transportation
like that with sulfonated polymers functioned successfully
under high RH in phosphonated polymers with high IEC values.
a
Department of Organic and Polymeric Materials, Graduate School of Science and
Engineering, Tokyo Institute of Technology, 2-12-1 O-okayama, Meguro-ku, Tokyo
1
52-8552, Japan. E-mail: ueda.m.ad@m.titech.ac.jp; Fax: +81-3-5734-2127; Tel: The rigid side chains, however, might hinder the formation of
+
81-3-5734-2127
the well-connected hydrogen-bonding networks and result in
decient proton conduction under low RH. Thus, in addition to
high IEC values, a exible polymer structure which can induce
the formation of hydrogen-bonding networks is considered to
be essential to achieve efficient proton transportation over a
b
PRESTO, Japan Science and Technology Agency (JST), 4-1-8 Honcho, Kawaguchi,
Saitama 332-0012, Japan. E-mail: thigashihara@polymer.titech.ac.jp; Fax: +81-3-
5
c
734-2126; Tel: +81-3-5734-2126
Department of Chemistry, Division of Advanced Materials Science, Pohang Accelerator
Laboratory, Center for Electro-Photo Behaviors in Advanced Molecular Systems,
Polymer Research Institute, and BK School of Molecular Science, Pohang University wide range of RH. Considering these factors, poly(-
ꢁ1
of Science and Technology (POSTECH), Pohang 790-784, Republic of Korea. E-mail: vinylphosphonic acid) (PVPA) (IEC ¼ 18.5 mequiv g ) is one of
ree@postech.ac.kr; Tel: +82-54-279-8143
the most promising candidates for PEMs due to the extremely
d
Higashifuji Technical Center, Toyota Motor Corporation, 1200 Mishuku, Susono,
Shizuoka 410-1193, Japan. E-mail: satoshi@nakazawa.tec.toyota.co.jp; Fax: +81-55-
23–26
high IEC values.
However, the proton conductivity of PVPA
was not as high as expected probably because of the limitation
of molecular motion. Although a cyclic oligosiloxane containing
phosphonic acid groups via a exible alkyl spacer was also
9
†
97-7120; Tel: +81-55-997-9020
Electronic supplementary information (ESI) available. See DOI:
0.1039/c2ta00537a
1
This journal is ª The Royal Society of Chemistry 2013
J. Mater. Chem. A, 2013, 1, 1457–1464 | 1457

View Article Online
Journal of Materials Chemistry A
designed, its proton conductivity (2.0 ꢂ 10 S cm ) was not 2.2 Thermal stability
Paper
ꢁ
3
ꢁ1
ꢀ
27
very high at 130 C under a RH of 37%.
The thermal stabilities of the cross-linked membranes were
evaluated by thermogravimetry analysis (Fig. S11†). A three-step
weight loss is observed from 150 to 250, from 250 to 300, and
above 300 C in the cross-linked membranes. Each weight loss
corresponds to the dehydration between the phosphonic acid
Here, we designed a novel series of phosphonated polymers,
which possessed phosphonic acid groups via exible pendant
side chains of 2, 4, and 6 methylene units, PAEPA, PABPA, and
PAHPA, toward high proton conductivity even under a low RH of
ꢀ
30%. The exible pendant side chains can hinder the formation
24,25
groups (the formation of the anhydrides),
the elimination of
of the anhydrides and give rise to promising performance due to
the high IEC values and hydrogen-bonding networks derived
from the exible polymer structures.
the phosphonic acid groups, and the decomposition of the
polymer backbone, respectively.
2.3 Oxidative stability
2
Results and discussion
The oxidative stabilities of the cross-linked membranes were
evaluated in Fenton's reagent at room temperature as an
accelerated test. The cross-linked membranes maintained the
shape of the membrane over 24 h. Membranes with high IEC
2
.1 Synthesis
PAEPA, PABPA, and PAHPA were synthesized by radical poly-
merization of diethyl 2-(acryloyloxy)ethylphosphonate (AEPN),
diethyl 4-(acryloyloxy)butylphosphonate (ABPN), and diethyl 6-
values generally show a weak oxidative stability against Fenton's
0
(acryloyloxy)hexylphosphonate (AHPN) using 2,2 -azobis(isobu-
28,29
reagent.
The cross-linked membranes, in contrast, showed
tyronitrile) (AIBN) as a radical initiator in toluene, followed by
hydrolysis of the phosphonic acid ester using trimethylsilyl
bromide (Scheme 1) (see ESI for details, Fig. S1–S9†). The
molecular weights of PAEPN, PABPN, PAHPN, were estimated to
high oxidative stability regardless of the high IEC values and the
aliphatic structure. One of the reasons for the high oxidative
stability is probably the quite low water uptake especially in the
case of PAHPA, as described below, by which the hydroxy radical
attack on the membrane could be signicantly suppressed.
be M
n
¼ 34 800 and M
w
¼ 89 200, M
n
¼ 43 000 and M
w
¼ 86 900,
M ¼ 40 000 and M ¼ 93 000, respectively, by size exclusion
n
w
chromatography. The structures of PAEPA, PABPA, and PAHPA
1
were characterized from H NMR (Fig. S3, S6 and S9†) and FT-IR 2.4 Hydrolytic stability
spectra. Aer the hydrolysis, the peaks corresponding to the
ethyl groups clearly disappeared in all H NMR spectra (Fig. S3,
The cross-linked membrane of PAHPA was also subjected to a
hydrolytic stability test at 80 C and 95% RH for 1 week. In the
1
ꢀ
S6 and S9†). All the polymers themselves are soluble in water due
to the high IEC values, and the cross-linked polymer membranes
were thus prepared by radical reaction with benzoyl peroxide as a
radical initiator (5 mol% for the polymers). The polymer radicals
formed by the a-hydrogen abstraction recombined to form cross-
linked polymers. The obtained membranes showed rubbery
properties and became insoluble in water, indicating that the
cross-linking reactions were successfully carried out to obtain
the cross-linked membranes. The occurrence of the cross-link-
FT-IR spectra (Fig. S12†), the absorbance peak intensity attrib-
uted to the carbonyl (C]O) stretching of the ester groups at
ꢁ1
1
731 cm remained unchanged, without appearance of a new
peak for the carbonyl (C]O) stretching of the carboxylic acid
ꢁ1
groups around 1700 cm aer testing. The result indicates that
the PAHPA membrane has a good stability to hydrolysis, prob-
ably because the ester groups are introduced apart from the
phosphonic acid groups via the long hydrophobic spacer,
repulsing the water molecules to be consumed for hydrolysis.
ing reaction at the CH or CH
2
group of a polyacrylate was further
conrmed by comparing the normalized absorbance peak
intensity attributed to carboxyl ester (C]O) groups at 1731 cm
ꢁ1
2.5 Water uptake and hydration number
ꢁ1
(
a) and that attributed to methine (C–H) groups at 2935 cm (b) Water uptake of the membranes is closely related to proton
of the polyacrylate in the FT-IR spectra, showing the decrease in conductivity. In sulfonated polymers, the IEC values of the
the peak intensity ratio, a/b, from 0.716 to 0.607 (Fig. S10†).
membranes generally dominate the water uptake of the
The IEC values of PAEPA, PABPA, and PAHPA determined by membranes, and high water uptake leads to efficient water-
30
titration with a 0.02 M NaOH aqueous solution were 5.44, 4.76, assisted proton transportation. In contrast, the high degree of
ꢁ
1
and 4.09 mequiv g , respectively, which were in good agreement hydrogen bonding and the lower acidity of the phosphonic acid
ꢁ1
with the theoretical IEC values of 5.55, 4.80, and 4.23 mequiv g
.
groups can limit the water uptake behavior of the membranes in
phosphonated polymers. The humidity dependence of water
uptake was investigated for the cross-linked polyacrylate and
ꢀ
Naon 117 membranes at 80 C (Fig. 1a). The slopes of the plots
for the water uptake depending on the RH are quite different
with the lengths of the side chains, becoming smaller in the
order of PAEPA, PABPA, and PAHPA. Especially, regardless of
ꢁ
1
the high IEC value (4.09 mequiv g ), the cross-linked PAHPA
membrane shows a lower water uptake than that of the Naon
ꢁ1
1
17 membrane (0.90 mequiv g ), leading to excellent dimen-
Scheme 1 Synthesis of PAHPA.
sional stability as well.
1
458 | J. Mater. Chem. A, 2013, 1, 1457–1464
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Paper
Journal of Materials Chemistry A
the whole range of RH (30–80%). It should be noted that the
proton conductivity increased in the order of PAEPA, PABPA,
and PAHPA, despite the IEC values decreasing from 5.44, 4.76,
ꢁ1
to 4.09 mequiv g . Even at 30% RH, the cross-linked PAHPA
ꢁ
3
membrane achieved a high proton conductivity of 8.9 ꢂ 10
S
ꢁ
1
cm . To the best of our knowledge, the cross-linked PAHPA
membrane shows the best performance of the proton conduc-
tivity among phosphonated polymers already reported.
The phosphonated polymer in our previous work (PDPAA)
showed
a drastic decrease in proton conductivity with
Fig. 1 Humidity dependence of (a) water uptake and (b) hydration number (l)
ꢀ
decreasing RH regardless of the high IEC value (6.72 mequiv
of Nafion 117, cross-linked PAEPA, PABPA and PAHPA membranes at 80 C.
ꢁ1
g
), which was concluded to result from the formation of the
decient hydrogen-bonding networks due to the rigid side
22
chains, preventing efficient proton transportation. In contrast,
the exible pendant side chains of PAHPA should contribute to
the formation of the hydrogen-bonding networks, which will
lead to excellent proton conductivity over the whole range of
RH. The signicantly low water uptake of the cross-linked
PAHPA membrane also supports the belief that the proton
transportation is derived from not only the water-assisted
feature but also the hydrogen-bonding network system.
Further, the hydration number (l), which is the number of
water molecules per acid group, is plotted against RH (Fig. 1b).
Compared to Naon 117 membrane, signicantly lower l values
were observed in all the cross-linked membranes over the whole
range of RH. Kaltbeitzel et al. previously investigated the water
sorption of PVPAs and noted that the water uptake strongly
ꢀ
decreased with decreasing RH (l < 1 at 30% RH and 80 C). They
also mentioned that such water uptake behavior resulted from
the dehydration between the phosphonic acid groups (the
23,24
formation of the anhydrides).
Wegner et al. demonstrated
2.7 Morphology
the formation of the anhydrides in PVPA by NMR spectroscopy
25
The cross-sectional morphology of the cross-linked PAHPA
membrane was investigated by STEM. In the STEM images
as well. Thus, the same phenomena can occur in the cross-
linked membranes, although the cross-linked structure can
affect the water uptake behavior as well.
(Fig. 3), the dark and bright regions were assigned to the
hydrophilic domains with phosphonic acid groups and the
hydrophobic domains, respectively, since the membrane was
stained with lead ions by ion exchange of the phosphonic acid
2
.6 Proton conductivity
The humidity dependence of proton conductivity was measured
for the cross-linked PAEPA, PABPA, PAHPA and Naon 117
membranes at 80 C (Fig. 2). Proton conductivity is generally
considered to play an important role in the performance of fuel
cells. Thus, higher levels of proton conductivity can provide
higher power densities in polymer electrolyte fuel cells (PEFCs).
As can be seen in Fig. 2, the slopes of the plots for proton
conductivity strongly depend on the length of side chain,
becoming smaller in the order of PAEPA, PABPA, PAHPA and
3 2
groups in 2 wt% Pb(OCOCH ) aqueous solution. As can be seen
in Fig. 3, a clear phase separation between the hydrophilic and
hydrophobic domains is observed. Interestingly, the inter-
distance of the cylinders is signicantly narrow (ꢃ2 nm) but
uniform and well-connected, which suggests that the exible
polymer structure successfully induced the self-arrangement of
each molecule to form a well-dened periodic nanostructure.
Such well-dened continuous hydrophilic domains (i.e., the
hydrogen-bonding networks) must contribute to efficient
proton transportation, and excellent proton conductivity
derived from the hydrogen-bonding networks was accom-
plished in the cross-linked PAHPA membrane.
ꢀ

nally the cross-linked PAHPA membrane shows excellent
proton conductivity comparable to the Naon 117 membrane in
Fig. 2 Humidity dependence of proton conductivity of cross-linked PAEPA,
Fig. 3 Cross-sectional STEM image of cross-linked PAHPA membrane, scale bar ¼
ꢀ
PABPA, PAHPA and Nafion 117 membranes at 80 C.
20 nm.
This journal is ª The Royal Society of Chemistry 2013
J. Mater. Chem. A, 2013, 1, 1457–1464 | 1459

View Article Online
Journal of Materials Chemistry A
Paper
ꢁ1
In order to gain more detailed morphology, synchrotron two sub-peaks (one peak centered at q ¼ 14.03 nm and the
ꢁ1
small and wide angle X-ray scattering (SAXS and WAXS) other centered at 17.45 nm ) by peak separation analysis using
measurements were further performed for the PAHPA lms a Gaussian function. Their d-spacings were determined to be
before and aer immersion in deionized water for 6 h (i.e., dry 0.45 nm and 0.36 nm, respectively. The d-spacing value of
and wet PAHPA lms) in order to obtain structural details and 0.45 nm is comparable to the mean interdistance of conven-
their changes with water swelling. The polymer lms were tional exible linear polymers such as polyethylene and poly-
observed to reveal featureless SAXS patterns before and aer propylene and, thus can be assigned as the mean interdistance
water swelling (data not shown). These results indicate that the between the bristles. In contrast, the d-spacing value of 0.36 nm
PAHPA chains in lms do not form any microstructure before is relatively shorter than the mean interdistances of conven-
and aer water swelling.
tional exible linear polymers and the bristles of the PAHPA
However, the polymer lms showed isotropic ring scattering polymer but larger than the interdistance (0.20 nm) of water
in the WAXS patterns (Fig. 4). The dry lm revealed several molecules displaying hydrogen bonding. Therefore, this
ꢁ
1
peaks positioned at q ¼ 3.21, 6.42 and 9.63 nm ; their relative d-spacing can be assigned as the mean interdistance between
scattering vector lengths from the beam center were 1, 2 and 3 the PAHPA polymers and water molecules. The PAHPA polymer
(
Fig. 4a and c). The results indicate that the scattering peak at was found to be hygroscopic because of the phosphonic acid
ꢁ
1
q ¼ 3.21 nm is the rst order peak and the others are the groups and thus its lm might contain some water molecules
second and third peaks of the rst one respectively. The d- which were absorbed when the lm was exposed to ambient air
spacing of the rst order peak was determined to be 1.96 nm. even though the air contact time was short.
This d-spacing value is smaller than twice the length (1.24 nm)
The wet PAHPA lms showed a WAXS pattern similar to that
of the fully stretched bristle including phosphonic acid esti- observed for the dry lms (Fig. 4b and c). However, the three
mated by molecular simulation with the Cerius2 soware peaks in the low q region were weakened in intensity, compared
package (Accelrys, San Diego, CA, USA). These results imply to those of the dry lms. Moreover, their positions were slightly
ꢁ
1
structural information on the dry PAHPA lm as follows. The shied toward the lower q region: 3.03, 6.07 and 9.10 nm
PAHPA chains have a cylinder-like structure centered at the (Fig. 4c). From these peaks, the mean interdistance of the
polymer backbone (i.e., helical conformation) rather than a helical polymer chains was determined to be 2.07 nm. This
sheet-like structure. Such molecular cylinders are laterally value is slightly larger than that of the dry lms. Another strong,
ꢁ
1
assembled together in a certain correlation length but without asymmetric peak additionally appeared around 20 nm . This
any regular packing order; such laterally assembled molecular peak was separated into two sub-peaks at q ¼ 14.03 (d-spacing ¼
ꢁ1
cylinders have a periodic interdistance of 1.96 nm, which is 0.45 nm) and 19.84 nm (d-spacing ¼ 0.32 nm). The rst sub-
consistent with that observed in the STEM image. The bristles peak corresponded to the mean interdistance of the bristles,
in the laterally assembled molecular cylinders are partially which is the same as that of the bristles in the dry lms. The
interdigitated via hydrogen bonding formations of the phos- second sub-peak originated from the mean interdistance
phonic acids at the bristle ends.
between the PAHPA polymers and water molecules, which is
In addition, another strong, broad peak was observed at q ¼ slightly shorter than that of the dry lms. These results collec-
ꢁ
1
1
4.03 nm . This broad scattering peak was found to consist of tively indicate that in the PAHPA lm the lateral interdistance of
the helical polymer chains was slightly increased by water
swelling.
3
Conclusions
A novel series of phosphonated polymers which had phos-
phonic acid groups via exible pendant side chains (PAEPA,
PABPA, and PAHPA) was synthesized by the radical polymeri-
zation of AEPN, ABPN and AHPN, respectively, followed by
hydrolysis with trimethylsilyl bromide. Among these three, the
cross-linked PAHPA membrane showed excellent proton
conductivity comparable to that of the Naon 117 membrane in
ꢀ
the range of 30–80% RH at 80 C, regardless of the signicant
low water uptake behavior. Furthermore, well-dened phase
separation between the hydrophilic and hydrophobic domains
was clearly observed in the cross-section of the cross-linked
PAHPA membrane by STEM. WAXS measurements revealed the
phase separated structures with d-spacing at 1.96 (dry state) and
Fig. 4 Representative two-dimensional (2D) WAXS patterns of (a) dry and (b)
wet PAHPA films. (c) One-dimensional (1D) scattering profiles extracted from the
2.07 nm (wet state), which correspond to the periodic mean
2
D scattering patterns in (a) and (b). In each 1D scattering profile, the scattering
interdistance of laterally assembled cylinder-like structures of
the cross-linked PAHPA. These results proved that the suffi-
ciently long exible pendant side chains (>6 methylene units)
peaks were separated WAXS patterns using a Gaussian function: the circles are
the measured data; the black dashed lines are the peaks separated by fitting with
the Gaussian function; the red solid line is the sum of the separated peaks.
1
460 | J. Mater. Chem. A, 2013, 1, 1457–1464
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Paper
Journal of Materials Chemistry A
induced the aggregation of the hydrophilic domains (i.e., the immersion in deionized water for 6 h. The scattering angle was
formation of the hydrogen-bonding networks) to achieve calibrated with pre-calibrated standards (sucrose, silver behen-
excellent proton transportation derived from the hydrogen- ate and polystyrene-b-polyethylene-b-polybutadiene-b-poly-
bonding networks. Consequently, the strategy in this work, styrene (SEBS) block copolymer). All scattering measurements
ꢀ
which is the introduction of the exible pendant side chain into were collected in 60 s and carried out at 25 C.
the phosphonated polymer, is truly promising to realize ideal
and efficient proton transportation derived from the hydrogen-
bonding networks, and PAHPA creates a new possibility for the
4.3 Synthesis
development of PEMs based on phosphonated polymers. Their Diethyl 2-(acryloyloxy)ethylphosphonate (AEPN). To a dry THF
actual fuel cell fabrication and evaluation are now under solution (11.0 mL) of DEHEP (2.00 g, 11.0 mmol) and triethyl-
investigation.
amine (1.84 mL, 13.2 mmol) was added acryloyl chloride
ꢀ
(1.07 mL, 13.2 mmol) dropwise at 0 C under a nitrogen atmo-
sphere. Then the mixture was stirred at room temperature for 1
h. Aer adding water, the mixture was extracted with diethyl
ether, washed with water and saturated NaHCO3 aqueous
4
Experimental
4.1 Materials
4
solution, and then dried over MgSO . Aer removing the
solvent, the obtained product was puried by column chroma-
Diethyl 2-hydroxyethylphosphonate (DEHEP), diethyl 4-bromo-
butylphosphonate (DEBBP) and diethyl 6-bromohex-
ylphosphonate (DEBHP) were prepared according to a previous
report. Tetrahydrofuran (THF) was reuxed over sodium
benzophenone for 12 h and then distilled. Triethylamine was
distilled from calcium hydride before use. Other solvents and
reagents were used without further purication.
tography (ethyl acetate : hexane ¼ 2 : 1) to obtain a colorless
1
31
liquid (1.79 g, 69%). H NMR (CDCl , d, ppm): 6.44 (1H, d), 6.12
3
(
1H, q), 5.86 (1H, d), 4.40 (2H, m), 4.13 (4H, m), 2.21 (2H, m),
1
3
1
5
1
3
.34 (6H, t). C NMR (CDCl , d, ppm): 166.1, 131.3, 128.5, 62.2,
ꢁ1
9.2, 31.2, 27.5, 25.7 16.8. FT-IR (NaCl, cm ): 2985, 2908, 1727,
635, 1446, 1411, 1265, 1191, 1025, 968, 809. Anal. calcd for
9 17 5
C H O P: C; 45.76%, H; 7.25%; found: C; 44.87%, H; 6.88%.
4
1
.2 Characterization
Poly[diethyl 2-(acryloyloxy)ethylphosphonate] (PAEPN).
13
0
H (300 MHz) and C (75 MHz) spectra were recorded with AEPN (0.14 g, 0.61 mmol), 2,2 -azobis(isobutyronitrile) (AIBN)
Bruker DPX300S spectrometers. Number- and weight-average (0.0040 g, 0.024 mmol), and dry toluene (0.25 mL) were placed
molecular weights (M
n
and M
w
) were measured by gel perme- in a glass tube using standard freeze–evacuate–thaw proce-
ꢀ
ation chromatography (GPC) on a Hitachi LC-7000 system dures. The mixture was stirred at 60 C for 6 h. Aer cooling to
equipped with polystyrene gel columns (TSKgel GMHHR-M) room temperature, the mixture was diluted with toluene and
eluted with N,N-dimethylformamide (DMF) containing 0.01 M then poured into hexane. The precipitate was collected and
LiBr at a ow rate of 1.0 mL min calibrated by standard poly- dried in vacuo at 50 C for 24 h to give PAEPN (0.14 g, 98%). H
styrene samples. Thermal analysis was performed on a Seiko NMR (MeOD, d, ppm): 4.47 (2H, broad), 4.34 (4H, m), 2.70–1.66
ꢁ
1
ꢀ
1
ꢁ1
EXSTAR 6000 TG/DTA 6300 thermal analyzer at a heating rate of (5H, broad). FT-IR (Si, cm ): 3451, 2985, 2935, 1735, 1446,
ꢀ
ꢁ1
10
C min for thermogravimetry (TG). For scanning trans- 1392, 1249, 1164, 1025, 964.
mission electron microscopy (STEM) observations, the Poly[4-(acryloyloxy)ethylphosphonic acid] (PAEPA). To a
membranes were stained with lead ions by ion exchange of the chloroform solution (5.0 mL) of PAEPN (0.24 g, 1.01 mmol) was
phosphonic acid groups in 2 wt% Pb(OCOCH ) aqueous solu- added trimethylsilyl bromide (0.78 mL, 6.0 mmol) dropwise at
3
2
ꢀ
ꢀ
tion, rinsed with deionised water for 3 days, and dried at room 0 C, and the mixture was stirred at 40 C for 24 h. Aer evap-
temperature for 1 day. The strained membranes were embedded oration, the residue was dissolved in methanol and the solution
in an epoxy resin, sectioned to 90 nm thickness with a micro- was stirred at room temperature for 6 h. Then methanol was
tome Ultracut UCT, and placed on copper grids. Images were evaporated and the obtained product was then dried in vacuo at
ꢀ
1
2
taken on a Hitachi H7100FA TEM with an accelerating voltage of 40 C for 24 h to obtain PAEPA (0.17 g, 94%). H NMR (D O, d,
1
ꢁ
1
00 kV. Synchrotron X-ray scattering measurements were carried ppm): 4.36 (2H, broad), 2.54–1.50 (5H, m). FT-IR (Si, cm ):
out at the 3C, 4C and 9A beamlines of the Pohang Accelerator 2965, 2287, 1727, 1454, 1164, 1010.
Laboratory at the Pohang University of Science and Tech-
4-(Diethoxyphosphoryl)ethyl acetate (DEPEA). A DMF solu-
32–37
nology.
The light source from an in-vacuum undulator (1.8 m tion (29 mL) of DEBBP (7.85 g, 28.9 mmol) and sodium acetate
ꢀ
long; 20 mm period) of the storage ring was monochromatized (5.68 g, 69.3 mmol) was stirred at 110 C for 5 h under a nitrogen
using a liquid nitrogen cooled double Si(111)/Si(311) crystal atmosphere. Aer removing the solvent, water was added and
monochromator system, giving an X-ray beam of 0.1154 nm (for the mixture was extracted with diethyl ether, then washed with
3
wide angle scattering measurements) and 0.1240 nm wavelength water and saturated NaHCO aqueous solution, and then dried
(
for small angle scattering measurements). The beam size was over MgSO . A colorless liquid was obtained (3.17 g, 44%). The
4
5
0 mm ꢂ 300 mm. A two-dimensional (2D) charge-coupled obtained product was used for the next reaction without any
1
detector (CCD) (Rayonix SX165, USA) was employed. Scattering purication. H NMR (CDCl , d, ppm): 4.09 (6H, m), 2.05 (3H, s),
3
patterns were measured at sample-to-detector distances (SDD) of 1.84–1.52 (6H, broad), 1.33 (6H, t).
230 and 3388 mm. X-ray scattering measurements were per-
Diethyl 4-hydroxybutylphosphonate (DEHBP). A MeOH
formed in transmission mode for PAHPA lms before and aer solution (14 mL) of DEBHP (8.26 g, 32.7 mmol) and potassium
This journal is ª The Royal Society of Chemistry 2013
J. Mater. Chem. A, 2013, 1, 1457–1464 | 1461

View Article Online
Journal of Materials Chemistry A
Paper
hydroxide (0.21 g, 3.67 mmol) was stirred at room temperature 4.13–3.96 (4H, m), 3.61–3.55 (2H, t), 2.21 (1H, s), 1.74–1.48 (6H,
for 15 h under a nitrogen atmosphere. Aer removing the m), 1.48–1.33 (4H, m), 1.30–1.25 (6H, t).
solvent, water was added and the mixture was extracted with
Diethyl 6-(acryloyloxy)hexylphosphonate (AHPN). To a dry
dichloromethane and then dried over MgSO . Aer removing THF solution (8.0 mL) of DEHHP (1.88 g, 7.89 mmol) and tri-
4
the solvent, a colorless liquid was obtained (4.61 g, 67%). The ethylamine (1.3 mL, 9.47 mmol) was added acryloyl chloride
ꢀ
obtained product was used for the next reaction without any (0.76 mL, 9.47 mmol) dropwise at 0 C under a nitrogen atmo-
1
purication. H NMR (CDCl
2
3
, d, ppm): 4.10 (4H, m), 3.67 (2H, t), sphere. Then the mixture was stirred at room temperature for
1 h. Aer adding water, the mixture was extracted with diethyl
Diethyl 4-(acryloyloxy)butylphosphonate (ABPN). To a dry ether, washed with water and saturated NaHCO aqueous
.01–1.62 (6H, broad), 1.33 (6H, t).
3
THF solution (11.0 mL) of DEHBP (2.25 g, 10.69 mmol) and solution, and then dried over MgSO . Aer removing the
4
triethylamine (1.8 mL, 12.8 mmol) was added acryloyl chloride solvent, the obtained product was puried by column chroma-
ꢀ
(
1.0 mL, 12.8 mmol) dropwise at 0 C under a nitrogen tography (ethyl acetate : hexane ¼ 2 : 1) to obtain a colorless
1
atmosphere. Then the mixture was stirred at room tempera- liquid (1.36 g, 59%). H NMR (CDCl
3
, d, ppm): 6.38 (1H, d), 6.08
ture for 1 h. Aer adding water, the mixture was extracted with (1H, q), 5.79 (1H, d), 4.14–3.98 (6H, m), 1.75–1.54 (6H, m), 1.39–
1
3
diethyl ether, washed with water and saturated NaHCO
3
3
1.36 (6H, m), 1.29 (6H, t). C NMR (CDCl , d, ppm): 166.3,
aqueous solution, and then dried over MgSO . Aer removing 130.5, 128.5, 77.5, 77.1, 76.6, 64.4, 61.4, 61.3, 30.1, 28.3, 26.5,
4
ꢁ
1
the solvent, the obtained product was puried by column 25.5, 25.4, 24.6, 22.4, 22.3, 16.5, 16.4. FT-IR (NaCl, cm ): 2981,
chromatography (ethyl acetate : hexane ¼ 2 : 1) to obtain a 2935, 2912, 1724, 1635, 1246, 1030. Anal. calcd for C H O P: C;
1
3
25 6
1
colorless liquid (2.2 g, 80%). H NMR (CDCl , d, ppm): 6.41 53.42%, H; 8.62%; found: C; 52.26%, H; 8.44%.
3
(
(
1H, d), 6.12 (1H, q), 5.83 (1H, d), 4.22–4.02 (6H, m), 1.85–1.59
Poly[diethyl 6-(acryloyloxy)hexylphosphonate] (PAHPN).
1
3
0
6H, broad), 1.31 (6H, t). C NMR (CDCl
3
, d, ppm): 166.2, AHPN (0.18 g, 0.61 mmol), 2,2 -azobis(isobutyronitrile) (AIBN)
1
30.6, 128.5, 63.9, 61.6, 29.5, 26.4, 24.5, 19.3, 16.6. FT-IR (0.0040 g, 0.024 mmol), and dry toluene (0.25 mL) were placed
ꢁ
1
(NaCl, cm ): 2985, 1724, 1635, 1457, 1411, 1199, 1025. Anal. in a glass tube using standard freeze–evacuate–thaw proce-
ꢀ
calcd for C13
H; 7.98%.
H
25
O
6
P: C; 50.00%, H; 8.01%; found: C; 49.78%, dures. The mixture was stirred at 60 C for 24 h. Aer cooling to
room temperature, the mixture was diluted with toluene and
Poly[diethyl 4-(acryloyloxy)butylphosphonate] (PABPN). then poured into hexane. The precipitate was collected and
0
ꢀ
1
ABPN (0.16 g, 0.61 mmol), 2,2 -azobis(isobutyronitrile) (AIBN) dried in vacuo at 50 C for 24 h to give PAHPN (0.17 g, 96%). H
0.0040 g, 0.024 mmol), and dry toluene (0.25 mL) were placed NMR (CDCl , d, ppm): 4.15–3.98 (6H, m), 2.22–1.59 (9H, m),
(
3
ꢁ
1
in a glass tube using standard freeze–evacuate–thaw proce- 1.37–1.29 (10H, m). FT-IR (Si, cm ): 2981, 2935, 1732,
ꢀ
dures. The mixture was stirred at 60 C for 6 h. Aer cooling to 1242, 1030.
room temperature, the mixture was diluted with toluene and
Poly[6-(acryloyloxy)hexylphosphonic acid] (PAHPA). To a
then poured into hexane. The precipitate was collected and chloroform solution (5.0 mL) of PAHPN (0.28 g, 0.96 mmol) was
ꢀ
1
dried in vacuo at 50 C for 24 h to give PAHPN (0.15 g, 95%). H added trimethylsilyl bromide (0.75 mL, 5.7 mmol) dropwise at 5
ꢀ
ꢀ
NMR (CDCl , d, ppm): 4.11 (6H, broad), 2.45–2.22 (1H, broad),
.00–1.60 (8H, broad), 1.34 (6H, t). FT-IR (Si, cm ): 2985, 1731, ration, the residue was dissolved in methanol and the solution
C, and the mixture was stirred at 40 C for 24 h. Aer evapo-
3
ꢁ1
2
1
650, 1446, 1392, 1230, 1025.
was stirred at room temperature for 6 h. Then methanol was
ꢀ
Poly[4-(acryloyloxy)butylphosphonic acid] (PABPA). To a evaporated and the obtained product was dried in vacuo at 40 C
chloroform solution (3.0 mL) of PABPN (0.16 g, 0.62 mmol) was for 24 h to obtain PAHPA (0.2 g, 90%). H NMR (CD
1
3
OD, d,
ꢁ
1
added trimethylsilyl bromide (0.48 mL, 3.7 mmol) dropwise at ppm): 4.09 (2H, broad), 2.32–1.47 (13H, m). FT-IR (Si, cm ):
ꢀ
ꢀ
C, and the mixture was stirred at 40 C for 24 h. Aer evap- 2939, 2866, 2310, 1732, 1165.
0
oration, the residue was dissolved in methanol and the solution
was stirred at room temperature for 6 h. Then methanol was
ꢀ
4.4 Preparation of cross-linked membranes and ion
exchange capacity (IEC)
evaporated and the obtained product was dried in vacuo at 40 C
1
for 24 h to obtain PABPA (0.11 g, 90%). H NMR (CD OD, d,
3
ppm): 4.14 (2H, broad), 2.56–2.14 (1H, broad), 2.13–1.44 (8H, 2-Methoxyethanol solutions of PAEPA, PABPA and PAHPA (0.31
ꢁ
1
broad). FT-IR (Si, cm ): 2958, 2283, 1727, 1110, 991.
mmol) and benzoyl peroxide (BPO) (0.0037 g, 0.015 mmol) were
ꢀ
Diethyl 6-hydroxyhexylphosphonate (DEHHP). A DMF solu- cast onto Teon sheets. Drying at 80, 90, and 100 C for 1 h
tion (14 mL) of DEBHP (4.31 g, 14.3 mmol), sodium acetate (1.41 respectively (totally for 3 h) under a nitrogen atmosphere gave
g, 17.2 mmol), and tetrabutylammonium bromide (0.10 g, 0.32 cross-linked membranes. The IEC values of the cross-linked
ꢀ
mmol) was stirred at 120 C for 2 h under a nitrogen atmo- membranes were determined by titration with 0.02 M NaOH
sphere. Aer cooling to room temperature, NaOH aqueous aqueous solution.
solution (50 wt%, 20 mL) was added dropwise and the mixture
stirred at room temperature for 15 h. The mixture was extracted
4.5 Oxidative stability
with diethyl ether, washed with water, and then dried over
MgSO . Aer removing the solvent, a colorless liquid was The oxidative stabilities of the membranes were tested by
2 2
obtained (1.61 g, 47%). The obtained product was used for the immersing the membranes into Fenton's reagent (3% H O
4
1
3 4
next reaction without any purication. H NMR (CDCl , d, ppm): aqueous solution containing 20 ppm FeSO ) at room
1
462 | J. Mater. Chem. A, 2013, 1, 1457–1464
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Paper
Journal of Materials Chemistry A
temperature. The dissolution time of the membranes into the
reagent was used to evaluate their oxidative stability.
4 K. A. Mauritz and R. B. Moore, Chem. Rev., 2004, 104, 4535–
4586.
5
6
7
8
9
M. Rikukawa and K. Sanui, Prog. Polym. Sci., 2000, 25, 1463–
502.
A. H. Hickner, H. Ghassemi, Y. S. Kim, B. R. Einsla and
J. E. McGrath, Chem. Rev., 2004, 104, 4587–4611.
1
4
.6 Water uptake and dimensional change
The humidity dependence of water uptake was measured by
plating a membrane in a thermo-controlled humidity chamber
for 5 h. Then the membrane was taken out, and quickly weighed
on a microbalance. Water uptake was calculated from:
T. Higashihara, K. Matsumoto and M. Ueda, Polymer, 2009,
50, 5341–5357.
K. Goto, I. Rozhanskii, Y. Yamakawa, K. Otsuki and Y. Naito,
Polym. J., 2009, 41, 95–104.
T. J. Peckham and S. Holdcro, Adv. Mater., 2010, 22, 4667–
WU ¼ (W ꢁ W )/W ꢂ 100 wt%
s
d
d
4
690.
s d
where W and W are the weights of wet and dried membrane,
respectively. The average values of statistically collecting data
were obtained by repeated experiments.
1
0 M. Schuster, T. Rager, A. Noda, K. D. Kreuer and J. Maier,
Fuel Cells, 2005, 5, 355–365.
1 H. Steininger, M. Schuster, K. D. Kreuer, A. Kaltbeitzel,
B. Bing ¨o l, W. H. Meyer, S. Schauff, G. Brunklaus, J. Maier
and H. W. Spiess, Phys. Chem. Chem. Phys., 2007, 9, 1764–
1
4.7 Proton conductivity
1
773.
Proton conductivity in the plane direction of a membrane was
determined using an electrochemical impedance spectroscopy
technique over the frequency from 5 to 10 Hz using a chemical
impedance meter (Hioki 3532-80, Nihon Denkei Co., Ltd) in a
temperature and humidity chamber (IW 222 type, Yamato
Scientic Co., Ltd). A two-point-probe conductivity cell with two
1
1
2 J. Parvole and P. Jannasch, Advances in Fuel Cells, ed. T. S.
Zhao, K. D. Kreuer and T. V. Nguyen, Elsevier, New York,
5
2007, pp. 120–185.
3 J. Parvole and P. Jannasch, Macromolecules, 2008, 41, 3893–
3903.
platinum plate electrodes was fabricated. The cell was placed in 14 J. Parvole and P. Jannasch, J. Mater. Chem., 2008, 18, 5547–
a thermo-controlled humid chamber. Proton conductivity (s)
5556.
15 R. Perrin, M. Elomaa and P. Jannasch, Macromolecules, 2009,
was calculated from:
42, 5146–5154.
s ¼ d/(L
s
w
s
R)
16 K. D. Papadimitriou, A. K. Andreopoulou and J. K. Kallitsis,
J. Polym. Sci., Part A: Polym. Chem., 2010, 48, 2817–2827.
17 R. Tayouo, G. David, B. Am ´e duri, J. Rozi `e re and S. Rouald `e s,
Macromolecules, 2010, 43, 5269–5276.
where d is the distance between the two electrodes, L and w are
s
s
the thickness and width of the membrane, and R is the resis-
tance value measured. The R values were determined by
exporting the electrochemical impedance plotting line to X-
intercept. The average s values of statistically collecting data
were obtained by repeated experiments.
1
1
2
2
8 M. Ingratta, M. Elomaa and P. Jannasch, Polym. Chem., 2010,
, 739–746.
9 N. Y. Abu-Thabit, S. A. Ali and J. S. M. Zaidi, J. Membr. Sci.,
010, 360, 26–33.
0 T. Itoh, K. Hirai, M. Tamura, T. Uno, M. Kubo and Y. Aihara,
J. Solid State Electrochem., 2010, 14, 2127–2189.
1 In the calculation of the theoretical IEC value, dened as the
number of mequiv per g protonated membrane, only one of
the two protons of the phosphonic acid is considered. See:
J. Parvole and P. Jannasch, Macromolecules, 2008, 41, 3893–
3903.
1
2
Acknowledgements
This work is supported by the Toyota Motor Corporation. M. Ree
acknowledges the nancial supports from the National
Research Foundation (NRF) of Korea (Doyak Program 2011-
0
028678 and Center for Electro-Photo Behaviors in Advanced
Molecular Systems (2010-0001784)) and the Ministry of Educa- 22 N. Fukuzaki, K. Nakabayashi, S. Nakazawa, S. Murata,
tion, Science & Technology (MEST), Korea (BK21 Program and
T. Higashihara and M. Ueda, J. Polym. Sci., Part A: Polym.
World Class University Program (R31-2008-000-10059-0)). The
Chem., 2011, 49, 93–100.
synchrotron X-ray scattering measurements at the Pohang 23 B. Bing ¨o l, W. H. Meyer, M. Wagner and G. Wegner,
Accelerator Laboratory were supported by MEST, POSCO
Company and POSTECH Foundation.
Macromol. Rapid Commun., 2006, 27, 1719–1724.
24 A. Kaltbeitzel, S. Schauff, H. Steininger, B. Bingoel,
G. Brunklaus, W. H. Meyer and H. W. Spiess, Solid State
Ionics, 2007, 178, 469–474.
Notes and references
2
5 Y. J. Lee, B. Bing ¨o l, T. Murakhtina, D. Sebastiani,
W. H. Meyer, G. Wegner and H. W. Spiess, J. Phys. Chem.
B, 2007, 111, 9711–9721.
1
2
3
K. Kordesch and G. Simader, Fuel Cells and Their
Applications, VCH, Weinheim, Germany, 1996.
K. D. Kreuer, A. Rabenau and W. Weppner, Angew. Chem., 26 F. Jiang, H. Zhu, R. Graf, W. H. Meyer, H. W. Spiess and
982, 94, 224–225. G. Wegner, Macromolecules, 2010, 43, 3876–3881.
O. J. Savadoga, J. New Mater. Electrochem. Syst., 1998, 1, 27 H. Steininger, M. Schuster, K. D. Kreuer and J. Maier, Solid
7–66. State Ionics, 2006, 177, 2457–2462.
1
4
This journal is ª The Royal Society of Chemistry 2013
J. Mater. Chem. A, 2013, 1, 1457–1464 | 1463

View Article Online
Journal of Materials Chemistry A
Paper
2
2
3
3
8 Y. Yin, Q. Du, Y. Qin, Y. Zhou and K. Okamoto, J. Membr. Sci., 33 S. Jin, J. Yoon, K. Heo, H. W. Park, T. J. Shin, T. Chang
2
011, 367, 211–219.
9 H. Zhou, K. Miyatake and M. Watanabe, Fuel Cells, 2005, 5,
96–301.
and M. Ree, J. Appl. Crystallogr., 2007, 40,
950–958.
34 J. Yoon, S. Y. Yang, K. Heo, B. Lee, W. Joo, J. K. Kim and
M. Ree, J. Appl. Crystallogr., 2007, 40, 305–312.
35 B. Lee, T. J. Shin, S. W. Lee, J. Yoon, J. Kim and M. Ree,
Macromolecules, 2004, 37, 4174–4184.
2
0 T. J. Peckham, J. Schmeisser, M. Rodgers and S. Holdcro,
J. Mater. Chem., 2007, 17, 3255–3268.
1 Y. Catel, M. Degrange, L. L. Pluart, P. J. Madec, T. N. Pham
and L. Picton, J. Polym. Sci., Part A: Polym. Chem., 2008, 46, 36 B. Lee, T. J. Shin, S. W. Lee, J. Yoon, J. Kim, H. S. Youn,
7074–7090.
K. B. Lee and M. Ree, Polymer, 2003, 44, 2509–
32 J. Yoon, K. W. Kim, J. Kim, K. Heo, K. S. Jin, S. Jin, T. J. Shin,
2518.
B. Lee, Y. Rho, B. Ahn and M. Ree, Macromol. Res., 2008, 16, 37 J. Bolze, J. Kim, J. Y. Huang, S. Rah, H. S. Youn, B. Lee,
75–585. T. J. Shin and M. Ree, Macromol. Res., 2002, 10, 2–12.
5
1
464 | J. Mater. Chem. A, 2013, 1, 1457–1464
This journal is ª The Royal Society of Chemistry 2013