Structure and properties of polymer electrolyte membranes containing phosphonic acids for anhydrous fuel cells

Article abstract of DOI:10.1021/cm202064x

Recently, water-free, proton-conducting polymer electrolytes have been attracting attention because of their possible application as fuel cell membranes at intermediate temperatures (100 to 200 °C). Phosphonic acid groups are considered feasible anhydrous proton conducting moieties due to the high degree of proton self-dissociation arising from their intrinsic amphoteric character and high mobility of protonic charge carriers. In this work, we have synthesized and characterized model, phosphonic acid-functionalized proton-conducting polymers, poly(vinylbenzyloxy-alkyl-phosphonic acid)s, for the purpose of exploring the relationship between molecular design, nanostructure, and performance characteristics. These novel proton conducting materials were characterized for their thermal stability, nanostructure, and performance properties. Thermogravimetric analysis (TGA) indicates that the polymers are thermally stable up to 140 °C, where the condensation of phosphonic acid groups starts to occur. Results from small-angle X-ray scattering (SAXS) show a peak corresponding to a Bragg spacing of approximately 21-24 A, which is attributed to layerlike structure formation of the phosphonic acid containing conducting channels. The proton conductivity increases with temperature, reaching a value on the order of 3 × 10^-4 S/cm at 140 °C under nominally anhydrous conditions.

Full text of DOI:10.1021/cm202064x

Article
pubs.acs.org/cm
Structure and Properties of Polymer Electrolyte Membranes
Containing Phosphonic Acids for Anhydrous Fuel Cells
Sung-Il Lee,† Kyung-Hwan Yoon,† Myeongsoo Song,† Huagen Peng,‡ Kirt A. Page,*,‡
Christopher L. Soles,‡ and Do Y. Yoon*,†
†Department of Chemistry, Seoul National University, Seoul 151-747, Korea
‡Polymers Division, National Institute of Standards and Technology, Gaithersburg, Maryland 20899, United States
S
* Supporting Information
ABSTRACT: Recently, water-free, proton-conducting polymer electrolytes
have been attracting attention because of their possible application as fuel cell
membranes at intermediate temperatures (100 to 200 °C). Phosphonic acid
groups are considered feasible anhydrous proton conducting moieties due to
the high degree of proton self-dissociation arising from their intrinsic
amphoteric character and high mobility of protonic charge carriers. In this
work, we have synthesized and characterized model, phosphonic acid-
functionalized proton-conducting polymers, poly(vinylbenzyloxy-alkyl-phos-
phonic acid)s, for the purpose of exploring the relationship between molecular
design, nanostructure, and performance characteristics. These novel proton
conducting materials were characterized for their thermal stability, nanostruc-
ture, and performance properties. Thermogravimetric analysis (TGA) indicates
that the polymers are thermally stable up to 140 °C, where the condensation of
phosphonic acid groups starts to occur. Results from small-angle X-ray
scattering (SAXS) show a peak corresponding to a Bragg spacing of approximately 21−24 Å, which is attributed to layerlike
structure formation of the phosphonic acid containing conducting channels. The proton conductivity increases with temperature,
reaching a value on the order of 3 × 10−4 S/cm at 140 °C under nominally anhydrous conditions.
KEYWORDS: polymer electrolyte membrane, anhydrous fuel cell, phosphonic acid
(approximately 80 °C). Under such operating conditions one
encounters serious problems such as carbon monoxide
poisoning of the catalyst layer, complicated heat and water
management, and water/methanol crossover.4−6 The desired
optimal PEM material would need to exhibit a conductivity of
greater than 0.1 S/cm, with a water content on the order of
20% by volume at an operating temperature of 120 °C.7
One strategy to achieving these goals is to develop new
proton conducting materials capable of operating under low
humidity, or anhydrous conditions, and at intermediate
temperatures (100 to 200 °C).8−10 Phosphoric acid doped
polybenzimidazole (PBI/H3PO4) blends have been studied
extensively and exhibit relatively high proton conductivity (ca.
0.03 S/cm) at approximately 150 °C. The main disadvantage of
these blends is that the H3PO4 small molecules, while in excess
of the base sites, can diffuse out of the membrane over
time.11,12 A more recent approach is to attach heterocyclic
molecules such as imidazole, benzimidazole, and pyrazole to
oligomeric or polymeric systems via flexible spacers.9,13−16 This
is because heterocycles are ideal proton solvents and can be
INTRODUCTION
■
In recent years, polymer electrolyte (or proton exchange)
membrane fuel cells (PEMFCs) have been considered as a
promising energy conversion system for applications ranging
from small-scale power sources for portable electronic devices
to automotive power supplies because of their high current
densities, fast start capability, and the ability to construct
compact, lightweight cells.1,2 The polymer electrolyte mem-
brane (PEM), which is one of the key components in a
PEMFC, acts both as a separator to gas permeation from the
anode to cathode as well as a solid electrolyte proton
conductor. The most widely used PEM material has been the
polyperfluorosulfonic acid (poly-PFSA), Nafion produced by
DuPont, because of its excellent chemical and thermal stability
as well as its high proton conductivity. The sulfonic acid groups
facilitate proton conduction but require a minimal level of
hydration to have considerable conductivity (e.g., σ = 0.1 S/cm,
λ (moles of H2O/mol of SO3H) = 22).3 However, these
hydrated PEMs have significant shortcomings that begin to
limit performance at operating temperatures near 100 °C,
where the water content begins to drop, resulting in a
significant decrease in the proton conductivity. As a result,
operation temperature of the fuel cell is limited to a
temperature somewhat below the boiling point of water
Received: July 19, 2011
Revised: November 5, 2011
Published: November 9, 2011
© 2011 American Chemical Society
115
dx.doi.org/10.1021/cm202064x | Chem. Mater. 2012, 24, 115−122

Chemistry of Materials
Article
substituted for water because of their amphoteric character and
hydrogen bond formation similar to water molecules. In
addition, the higher boiling point of heterocycles compared to
water enables a PEM-based fuel cell to operate at higher
temperature. For these heterocyclic proton conductors, it is
more likely that proton transport is facilitated by structural
diffusion of the protogenic groups, also known as the Grotthuss
mechanism, under water-free condition.17,18 Schuster et al.19
compared the use of different protogenic groups at
intermediate temperatures for low humidity PEMFCs including
sulfonic acid groups, phosphonic acid groups and imidazole
groups as model compounds. They concluded that, at
intermediate temperatures (120 to 160 °C) and low humidity,
the phosphonic acid group is the most suitable proton
conductor. A high degree of self-dissociation, which results
from more pronounced intrinsic amphoteric character of
phosphonic acid, and a high mobility of the protonic charge
carrier were two key factors that were thought to contribute to
the observed higher proton conductivity. In addition,
phosphonic acid has comparatively better electrochemical and
thermo-oxidative stability and allows higher reaction rates for
hydrogen oxidation and oxygen reduction on Pt electrode
surfaces. These results reveal that phosphonic acid is very
attractive as a protogenic group in anhydrous proton
conducting polyelectrolytes. On the basis of such results,
immobilized phosphonic acid based systems were prepared by
tethering phosphonic acid groups to a polymer backbone. The
simplest case of a polymer containing phosphonic acid found in
the literature is poly(vinylphosphonic acid) (PVPA).20 The
proton conductivity of this material in the nominally dry state
was determined to be on the order of 1 × 10−4 to 1 × 10−3 S/
cm at approximately 150 °C. The proton conductivity is
attributed to the formation of interconnected, mobile
aggregates of phosphonic acid groups and the consequent
Grotthus mechanism.10 Additionally, copolymers of PVPA21,22
and blends with heterocycles23,24 were investigated along with
acid−base composite materials25,26 based on phosphate and
imidazole. Despite its ability to conduct protons, PVPA has
poor mechanical properties for use in a fuel cell in large part
due to its low glass transition temperature (Tg ≈ −23 °C).23 To
solve this problem, researchers have turned to incorporating
aromatic groups into the polymer structure to improve the
chemical and thermal stability of the PEM for practical
applications.27−31
In this work, a pair of model polystyrene derivatives
containing phosphonic acid side groups, with varying spacer
lengths, were prepared by the Michealis−Arbuzov reaction and
other synthetic methods. The polystyrene backbone was
selected due to its facile polymerization and better mechanical
properties including a significantly higher Tg as compared to
PVPA. The incorporation of flexible spacers will allow sufficient
mobility to the protogenic group to facilitate conductivity via
charge motion. Moreover, the surfactant-like structure of the
side chain would promote a micro or nano phase-separation of
phosphonic acid aggregates,32 which could serve as pathways
for proton conduction while the polystyrene-like backbone
provides the improved mechanical performance including a
high Tg. The molecular masses, thermal stability, structure−
property relationships, and proton conductivity of these
anhydrous fuel cell membrane materials were studied employ-
ing a variety of experimental techniques including aqueous gel
permeation chromatography (GPC), thermogravimetric anal-
ysis (TGA), small-angle X-ray scattering (SAXS), dynamic
mechanical analysis (DMA), and broadband dielectric
relaxation spectroscopy (DRS).
EXPERIMENTAL SECTION
■
Certain commercial equipment, instruments, or materials are identified
in this paper in order to specify the experimental procedure
adequately. Such identification is not intended to imply recommen-
dation or endorsement by the National Institute of Standards and
Technology, nor is it intended to imply that the materials or
equipment identified are necessarily the best available for the purpose.
Materials. 4-vinylbenzyl chloride, ethylene glycol, tri(ethylene)-
glycol, 1,6-hexanediol, sodium hydride, p-toluenesulfonyl chloride,
triethylamine (TEA), sodium iodide, triisopropyl phosphite, N,N-
dimethylformamide (DMF), acetone, 2,2′-azobis(isobutyramidine)
dihydrochloride (AIBA), hydroquinone were purchased from Sigma-
Aldrich, Inc. and bromotrimethylsilane from TCI Chemical Inc. All
reagents mentioned above were used without any further purification.
Dichloromethane, a reaction solvent, was distilled over calcium
hydride (CaH2). Spectra/Por dialysis membrane with molecular cutoff
value of 2000 was purchased from Spectrum Laboratories, Inc.
Measurements. 1H NMR and 13C NMR spectra were obtained
with a Bruker Fourier Transform DPX-300 MHz nuclear magnetic
resonance spectrometer at room temperature in CDCl3 and DMSO-
d6. Chemical shifts were measured in parts per million (δ value) from
tetramethylsilane as an internal standard.
The relative molecular weights of polymers were determined against
a poly(ethyleneglycol) standard calibration in an aqueous gel
permeation chromatography (GPC) system, comprised of Waters
515 pump, PL aquagel OH 30 column (column temperature 35 °C)
and Waters 2410 differential refractometer. A mixture of pH 9.0 buffer
(70 vol %) and methanol (30 vol %) was eluted at a flow rate of 1.0
mL/min. Into the GPC, 100 μL of sample solution with a
concentration of 3 mg/mL was injected.
Thermogravimetric analysis (TGA) was carried out with TGA Q50
(TA Instruments). The temperature was raised at a rate of 10 °C/min
from 30 to 700 °C with a constant nitrogen gas purge rate of 60 mL/
min. Glass transition temperatures were determined with a differential
scanning calorimeter (DSC Q1000, TA Instruments) at a heating rate
of 20 °C/min from −70 to 100 °C. Prior to measurements, samples
were kept in a vacuum oven at 60 °C for 2 days to completely remove
residual water or solvents.
Dynamic mechanical analysis (DMA) was carried out with DMA
2980 (TA Instruments), at a controlled force mode with a penetration
clamp. The polymer samples were prepared in shape of pressed pellets
(diameter, 12 mm) and dried out in a vacuum oven at 60 °C for 24 h.
A flat probe penetrated into the pellet with a static force of 0.5 N at a
heating rate of 1 °C/min from 20 to 120 °C.
Small-angle X-ray scattering (SAXS) measurements were performed
on a laboratory X-ray scattering instrument that utilizes pinhole
collimated CuKα radiation, using three consecutive pinholes with
diameters of 1000 μm, 400 μm, and 300 μm, decreasing in diameter as
the incident beam approaches the sample. Two-dimensional scattering
patterns from the fuel cell membrane were collected on an image plate
detector under vacuum to reduce the scattering from the air. The
sample-to-detector distance (SDD) is 21.3 cm, which provides a wave
vector (Q) range from 1 to 20 nm−1. Here, Q is expressed by the
equation, Q = (4π/λ)sin θ, where λ is the wavelength of CuKα, 0.154
nm at 8.04 keV, and 2θ is the scattering angle.
Ionic conductivities were measured under dry nitrogen atmosphere
using broadband dielectric spectrometer from NOVOCONTROL
GmbH, Germany, in the temperature range from −10 to 140 °C, with
temperature intervals of 10 °C over a frequency range from 10−2 Hz to
107 Hz. Temperature was controlled with the Novocontrol Quatro
Cryosystem with a set point precision of 0.1 °C. The measurements
were performed over several heating and cooling cycles to determine
the stability and reproducibility of the results. All samples for
conductivity measurements were obtained by preparing solvent-cast
films and then drying them under vacuum at 60 to 70 °C for 3 days to
remove residual solvents/water and to obtain the nominally dry
116
dx.doi.org/10.1021/cm202064x | Chem. Mater. 2012, 24, 115−122

Chemistry of Materials
Article
sample as previously reported.10 The films were sandwiched between
two gold-plated electrodes (diameter, 10 mm, 35 mm) using a Teflon
spacer (thickness, 129.5 μm).
2H), 3.71 (m, 2H), 2.13 (td, 2H); 13C NMR (300 MHz, CDCl3) δ
137.3, 137.0, 136.6, 128.2, 126.4, 114.1, 72.6, 64.2, 28.2, 26.4.
2-(4-Vinyl-benzyloxy)- hexan-1-ol (2b). To a stirred solution of 1,
6-hexandiol (35.4 g, 0.30 mol) in DMF (55 mL) was slowly added
sodium hydride (4.8 g, 0.12 mol, 60% dispersion in mineral oil) at 0
°C. After the solution was stirred over 1 h, 4-vinylbenzyl chloride (1,
14.1 mL, 0.10 mol) was added dropwise using a syringe at 0 °C and
vigorously stirred for 20 h at the room temperature. The same workup
and purification procedure described for 2a was carried out and 18.1 g
Synthesis of Monomers. 2-(4-Vinyl-benzyloxy)-ethanol
(2a). Ethylene glycol (16.7 mL, 0.30 mol) was added dropwise to a
stirred solution of sodium hydride (4.8 g, 0.12 mol, 60% dispersion in
mineral oil) in N,N-dimethylformamide (DMF) (50 mL) at 0 °C.
After the solution was stirred over 1 h, 4-vinylbenzyl chloride (1, 14.1
mL, 0.10 mol) was added dropwise to the suspension, using a syringe
at 0 °C and vigorously stirred at room temperature. After 20 h of
stirring, 300 mL of water was poured into the solution. Ethyl acetate
was used to extract organic products (3 × 300 mL). The organic layer
was collected and dried over anhydrous magnesium sulfate. The
solvent was removed by a rotary evaporator. The product was further
purified by a silica gel column chromatography. 11.0 g (62%) of 2a was
obtained as a colorless liquid. 1H NMR (300 MHz, CDCl3) δ 7.39 (d,
2H), 7.29 (d, 2H), 6.71 (dd, 1H), 5.74 (d, 1H), 5.24 (d, 1H), 4.53 (s,
2H), 3.73(t, 3H), 3.57 (t, 3H), 2.29 (s, 1H).
Toluene-4-sulfonic Acid 2-(4-vinyl-benzyloxy)-ethyl Ester (3a). p-
Toluenesulfonyl chloride (22.9 g, 0.12 mol) was dissolved in distilled
dichloromethane (80 mL) and 2a (18.2 g, 0.10 mol) was added to the
stirred solution dropwise at 0 °C. After stirring for 30 min,
triethylamine (16.7 mL, 0.12 mol) was added dropwise and the
solution was stirred at room temperature overnight. The product was
extracted with water and diethyl ether (3 × 200 mL). The organic
layer was collected and dried over anhydrous magnesium sulfate. The
solvent was evaporated by a rotary evaporator. The product was
further purified by a silica gel column chromatography. 11.0 g (33%)
of 3a was obtained as a colorless liquid. 1H NMR (300 MHz, CDCl3)
δ 7.79 (d, 2H), 7.37 (d, 2H), 7.30 (d, 2H), 7.21 (d, 2H), 6.71 (dd,
1H), 5.74 (d, 1H), 5.24 (d, 1H), 4.46 (s, 2H), 4.19 (t, 2H), 3.65 (t,
2H), 2.43 (s, 3H).
1
(77%) of 2b was obtained as a colorless liquid. H NMR (300 MHz,
CDCl3) δ 7.39 (d, 2H), 7.29 (d, 2H), 6.71 (dd, 1H), 5.74 (d, 1H),
5.23 (d, 1H), 4.48 (s, 2H), 3.62 (t, 2H), 3.46 (t, 2H), 1.68−1.51 (m,
4H), 1.49 (s, 1H), 1.45−1.30 (m, 4H).
Toluene-4-sulfonic Acid 6-(4-Vinyl-benzyloxy)-hexyl Ester (3b). p-
Toluenesulfonyl chloride (38.6 g, 0.20 mol) was dissolved in distilled
dichloromethane (120 mL) and cooled to 0 °C. 2b (36.0 g, 0.15 mol)
was added to the stirred solution dropwise at 0 °C. After stirring for 30
min, triethylamine (27.6 mL, 0.20 mol) was added dropwise. The
solution was stirred overnight at the room temperature. The same
workup and purification procedure described for 3a was applied and
1
57.2 g (95%) of 3b was obtained as a colorless liquid. H NMR (300
MHz, CDCl3) δ 7.78 (d, 2H), 7.38 (d, 2H), 7.32 (d, 2H), 7.27 (d,
2H), 6.71 (dd, 1H), 5.74 (d, 1H), 5.23 (d, 1H), 4.46 (s, 2H), 4.01 (t,
2H), 3.41 (t, 2H), 2.43 (s, 3H), 1.69−1.49 (m, 4H), 1.38−1.24 (m,
4H).
1-(6-Iodo-hexoxymethyl)-4-vinyl-benzene (4b). Sodium iodide
(49.5 g, 0.33 mol) was dissolved in 300 mL of anhydrous acetone.
3b (57.0 g, 0.15 mol) was added dropwise and the solution was
refluxed for 40 h. Acetone was removed by a rotary evaporator and the
mixture was extracted with water and diethyl ether (3 × 150 mL). The
same workup and purification procedure described for 4a was applied
1
to obtain 41.5 g (82%) of 4b as a yellowish liquid. H NMR (300
1-(2-Iodo-ethoxymethyl)-4-vinyl-benzene (4a). Sodium iodide
(27 g, 0.18 mol) was dissolved in 300 mL of anhydrous acetone. 3a
(30 g, 0.090 mol) was added dropwise and the mixture was heated to
reflux for 15 h. Acetone was removed by a rotary evaporator and the
mixture was extracted with water and diethyl ether (3 × 150 mL). The
organic layer was collected and dried over anhydrous magnesium
sulfate. The solvent was removed by a rotary evaporator. The product
was further purified by a silica gel column chromatography. 18.8 g
MHz, CDCl3) δ 7.39 (d, 2H), 7.29 (d, 2H), 6.71 (dd, 1H), 5.74 (d,
1H), 5.23 (d, 1H), 4.48 (s, 2H), 3.45 (t, 2H), 3.18 (t, 2H), 1.88−1.76
(m, 2H), 1.68−1.55 (m, 2H), 1.46−1.33 (m, 4H).
[6-(4-Vinyl-benzyloxy)-hexyl]-phosphonic Acid Diisopropyl Ester
(5b). In a round-bottomed flask, 4b (5.2 g, 0.015 mol), triisopropyl
phosphite (3.1 mL, 0.014 mol) and hydroquinone (0.0495 g, 0.45
mmol) were mixed. The reaction temperature was first raised to 90 °C
with vigorous stirring and gradually increased to 120 °C. The reaction
mixture was stirred for 14 h and then loaded directly on the top of a
silica gel column chromatography for purification. 3.7 g (71%) of 5b
was obtained as a colorless liquid. 1H NMR (300 MHz, CDCl3) δ 7.39
(d, 2H), 7.29 (d, 2H), 6.71 (dd, 1H), 5.74 (d, 1H), 5.23 (d, 1H), 4.68
(m, 2H), 4.48 (s, 2H), 3.45 (t, 2H), 1.74−1.50 (m, 6H), 1.44−1.35
(m, 4H), 1.30 (d, 12H); 13C NMR (300 MHz, CDCl3) δ 138.3, 137.0,
136.7, 127.9, 126.3, 113.8, 72.7, 70.3, 61.5, 61.4, 30.6, 30.4, 29.6, 26.7,
25.8, 24.8, 22.5, 22.4, 16.6 ( × 2), 16.5 ( × 2).
1
(91%) of 4a was obtained as a yellowish liquid. H NMR (300 MHz,
CDCl3) δ 7.40 (d, 2H), 7.31 (d, 2H), 6.71 (dd, 1H), 5.75 (d, 1H),
5.25 (d, 1H), 4.79 (s, 2H), 3.73(t, 2H), 3.27 (t, 2H).
[2-(4-Vinyl−benzyloxy)-ethyl]-phosphonic Acid Diisopropyl Ester
(5a). Triisopropyl phosphite (3.8 mL, 0.017 mol), 4a (5.3 g, 0.018
mol) and hydroquinone (0.0594 g, 0.53 mmol) was mixed in a round-
bottomed flask without solvent. The reaction temperature was first
raised to 90 °C with vigorous stirring and gradually increased to 120
°C. The reaction mixture was stirred for 14 h at 120 °C, then was
loaded directly on the top of a silica gel column chromatography for
[6-(4-Vinyl-benzyloxy)-hexyl]-phosphonic Acid (6b). A dried
round-bottomed flask was filled with 5 mL of distilled dichloro-
methane and 5b (3.1 g, 0.0081 mol) was added with stirring. After
achieving a homogeneous solution, bromotrimethylsilane (2.4 mL,
0.018 mol) was added dropwise for 10 min under nitrogen gas flow.
After the color of the solution was changed from colorless to yellowish,
the mixture was stirred for 5 h at room temperature. The same workup
and purification procedure described in 6a was applied to obtain 6b
(2.2 g, 93%) as a yellowish solid. 1H NMR (300 MHz, CDCl3) δ 9.51
(s, 2H), 7.36 (d, 2H), 7.26 (d, 2H), 6.68 (dd, 1H), 5.72 (d, 1H), 5.21
(d, 1H), 4.46 (s, 2H), 3.42 (t, 2H), 1.81−1.49 (m, 6H), 1.42−1.26 (m,
4H); 13C NMR (300 MHz, CDCl3) δ 138.0, 137.1, 136.7, 128.1,
126.4, 113.9, 72.7, 70.4, 30.4, 30.2, 29.5, 26.9, 25.7, 25.0, 22.2, 22.1.
Polymerization. Monomer 6a (1.31 g, 0.0054 mol) and 2,2′-
azobis(isobutyramidine) dihydrochloride (AIBA) (0.015 g, 0.054
mmol, 1 mol-%) were put together in a Schlenk tube and dissolved
in 6 mL (1 mol/L) of 1,4-dioxane/deionized water (v/v = 3/2). The
solution was degassed by freeze−pump−thaw procedures: the reaction
mixture was completely frozen in liquid nitrogen for 4 min, the tube
was evacuated for 1 min, and warmed to the room temperature. The
whole procedure was repeated over 3 times. The Schlenk tube was
charged with nitrogen gas and the solution was stirred at 70 °C for 14
1
purification. 4.4 g (81%) of 5a was obtained as a colorless liquid. H
NMR (300 MHz, CDCl3) δ 7.38 (d, 2H), 7.29 (d, 2H), 6.71 (dd, 1H),
5.76 (d, 1H), 5.23 (d, 2H), 4.69 (m, 2H), 4.50 (s, 2H), 3.72 (m, 2H),
2.12 (td, 2H), 1.29 (s, 12H); 13C NMR (300 MHz, CDCl3) δ 137.6,
137.2, 136.6, 128.0, 126.3, 114.0, 72.8, 64.3, 61.8, 61.7, 28.1, 26.2, 16.6
( × 2), 16.5 ( × 2).
[2-(4-Vinyl-benzyloxy)-ethyl]-phosphonic Acid (6a). A dried
round-bottomed flask was filled with 5 mL of distilled dichloro-
methane and 5a (3.4 g, 0.010 mol) was added with stirring.
Bromotrimethylsilane (3.0 mL, 0.023 mol) was added dropwise for
10 min under nitrogen gas flow. After the color of the solution was
changed from colorless to yellowish, the mixture was stirred for 5 h at
the room temperature. Most of the solvent and unreacted
bromotrimethylsilane was removed by a rotary evaporator and a
subsequent high vacuum-dry. Five milliliters of deionized water was
added and the mixture was stirred for 3 h. The water and residual
solvent were removed by a rotary evaporator and the product was
completely dried in a high vacuum. 6a (2.3 g, 95%) was yielded as a
yellowish solid. 1H NMR (300 MHz, CDCl3) δ 8.80 (s, 2H), 7.32 (d,
2H), 7.22 (d, 2H), 6.65 (dd, 1H), 5.69 (d, 2H), 5.20 (d, 2H), 4.42 (s,
117
dx.doi.org/10.1021/cm202064x | Chem. Mater. 2012, 24, 115−122

Chemistry of Materials
Article
Scheme 1. Synthesis of Styrene-Derivative Monomers Containing the Phosphonic Acid
Scheme 2. Polymerization of Styrene-Derivative Monomers Containing the Phosphonic Acid
h. The resulting product was diluted with a small amount of deionized
water and poured into a Spectra/Por dialysis membrane with a
molecular cutoff value of 2000. After the polymer was purified by a
dialysis method for 5 days, water and residual solvent were removed by
freeze-drying. Polymer 7a (denoted as PVBPA-Ethyl) was obtained as
a yellowish-white powder (0.86 g, 66%). Monomer 6b (1.28 g, 0.0043
mol) and AIBA (0.012 g, 0.043 mmol) were put together in a Schlenk
tube and dissolved in 6 mL of 1,4-dioxane/deionized water (v/v = 3/
2). The polymerization was performed by the same procedure as
described above. Polymer 7b (denoted as PVBPA-Hexyl) was
obtained as a yellowish-white powder (0.81 g, 63%).
mixture without hydroquinone gradually became viscous and
finally yielded a gel-like product because of the polymerization
of the vinyl groups. After purification by a silica-gel column
chromatography, the yield of this phosphonation-step without
hydroquinone was as low as 30%. To minimize the thermal
polymerization, 0.03 mol % of hydroquinone was added as an
inhibitor. As a result, the phosphonation reaction proceeded
quite well at 120 °C with the yield of over 70%. Hydrolysis of
the phosphonate group was accomplished under mild condition
by using bromotrimethylsilane. Generally, phosphonate groups
can be hydrolyzed to phosphonic acids in an aqueous solution
of strong acids such as HCl, HBr under reflux. However, the
vinyl group in the monomer can be saturated under such highly
acidic conditions.34 In addition, highly basic condition is not
feasible either, because phosphonate cannot be fully hydrolyzed
and monoester of phosphonic acid can be obtained.35
Therefore, bromotrimethylsilane was used for mild hydrolysis
to avoid undesirable side products, and the phosphonate
compound was fully hydrolyzed to phosphonic acid at room
temperature.36,37
All polymers were prepared by free radical polymerization of
monomers 6a and 6b as shown in Scheme 2. Phosphonic acid
monomers 6a and 6b were polymerized in mixture of 1,4-
dioxane and deionized water with AIBA as a water-soluble
initiator. The resulting polymers were purified by dialysis in
deionized water medium using a separation-membrane with
approximately a Mn = 2000 g/mol cutoff. PVBPA-Ethyl was
quite soluble in water and the solution was transparent during
dialysis. The solubility of PVBPA-Hexyl in water was not as
good as indicated by the turbidity of the solution. The resulting
polymers, after removal of water by freeze-drying, were
RESULTS AND DISCUSSION
■
Synthesis of Monomers and Polymerization. Styrene
derivatives with phosphonic acid were prepared as described in
Scheme 1. Two alkyl chains with different numbers of carbon
units (ethyl and hexyl) were attached onto the styrene aromatic
ring by a simple substitution reaction of the diol compound and
4-vinylbenzyl chloride as the starting reagent using NaH. To
increase the yield of the monosubstituted diol compound in the
first reaction step, over 3 times excess amount of diol was used.
The hydroxyl group of the alkyl chain-end was converted to a
tosylate group because it is easily converted to an iodide group
by substitution. Alkyl iodide is the most susceptible halide to
Michaelis−Arbuzov reaction,33 which is a widely used method
to make carbon−phosphorus bonds. Normally, the primary
alkyl halide can be transformed to an alkyl phosphonate at the
temperature of 150 to 160 °C without any solvent or catalyst in
the Michaelis-Arbuzov reaction. But in the phosphonation-step
of these systems with vinyl group (preparation of 5a and 5b),
the reaction temperature should be controlled below 125 °C.
When the temperature was as high as 120−125 °C, a reaction
118
dx.doi.org/10.1021/cm202064x | Chem. Mater. 2012, 24, 115−122

Chemistry of Materials
Article
yellowish-white solids. The molecular masses of polymers were
analyzed by an aqueous GPC. The weight-averaged molecular
masses of PVBPA-Ethyl and PVBPA-Hexyl were approximately
25 000 and 22 000 g/mol, respectively, with a relatively narrow
polydispersity index of 1.4 for both samples. Molecular masses
are relative values against PEG standard. Measuring the
absolute molecular mass of PVBPA with GPC is a rather
complicated problem, since the molecular mass per a skeletal
bond of PVBPA monomer is much larger, by ca. 8 times, than
PEG monomer, and the hydrodynamic volume of PVBPA is
affected by the polyelectrolyte effects screened in part by the
buffer.20
°C/min and then maintaining the temperature for 5 min at
each step. This result is shown in Figure 1b for PVBPA-Ethyl,
where the steps in the mass loss are seen in temperature-range
between 100 and 200 °C. From 100 to 140 °C, the extent of
the mass loss is small, but above 150 °C the amount of mass
loss became quite significant. PVBPA-Hexyl also shows a
similar response. Our interpretation of these step-isothermal
TGA experiments is that these polymers are thermally stable up
to 140 °C against the self-condensation of phosphonic acid
groups.
Small-Angle X-ray Scattering Measurements. In order
to ascertain the formation of proton-conducting channels in the
PVBPA polymers, small-angle X-ray scattering experiments
were performed at room temperature on solvent-cast and dried
films. As shown in Figure 2, there are well-defined maxima in
Thermal Stability. The thermal stability of the polymers
was characterized by TGA under nitrogen at a heating rate of
10 °C/min. Figure 1a shows the temperature-dependent mass
Figure 2. Small-angle X-ray scattering (SAXS) profile measured at
room temperature. The Bragg spacing, ∼(21 1) Å, indicates that the
side chains form a bilayer lamellar structure.
the scattering intensity for both PVBPA-Ethyl and PVBPA-
Hexyl samples. The Qmax of these peaks correspond to a Bragg-
spacing of approximately (21
1) Å and (24
1) Å for
PVBPA-Ethyl and PVBPA-Hexyl, respectively. In addition, the
PVBPA-Hexyl sample shows a second order peak at Q2‑max that
is approximately twice the value of the first peak Q1‑max. A rough
estimate based on the molecular structure indicates that the
distance between the phosphonic acid moiety and the polymer
backbone for PVBPA-Ethyl is approximately 10 Å. This is about
one-half the value of the Bragg spacing determined from the
scattering maximum. These data are consistent with the notion
that the phosphonic acid-containing side-chains, through
aggregation of the acidic groups, form a bilayer or lamellar-
like proton-conducting channel structure. It is not surprising
considering that it is known that phosphonate-containing
surfactants readily form a number of nanostructured micellar
phases.38,39 This bilayer structure is quite different from the
proposed core-shell structure of Nafion.32,40
Thermal Relaxation Characteristic. DSC and DMA
measurements were carried out to investigate the thermo-
mechanical relaxation properties of these polymers, related to
the bilayer-type nanostructure in bulk films. In the DSC
thermograms shown in Figure 3a, the glass transition
temperature (Tg) was determined from the midpoint of the
step-change in the heat capacity. The Tgs for the different
materials are very similar, Tg ≈ 20 °C, with no apparent
Figure 1. Thermal mass loss characteristics investigated by
thermogravimetric analysis (TGA): (a) continuous heating at 10
°C/min and (b) step-isothermal heating for 5 min at each temperature
of 10 °C interval from 100 to 200 °C. The bottom portion shows the
temperature derivatives of mass loss and the inset is the expanded plot
of the 100 to 200 °C region.
loss of these materials. In both samples, the first mass-loss
process appears between 150 and 230 °C. This degradation is
consistent with the loss of water by self-condensation of the
phosphonic acid groups. The second mass-loss process occurs
between 230 and 330 °C and is attributed to the removal of the
side-chains attached to the polystyrene backbone by cleavage of
ether bonds. The polymer backbone starts to decompose above
330 °C, resulting in further mass loss. Such a thermal
degradation tendency is consistent with the previous result
obtained from poly(vinylbenzylphosphonic acid).31 To deter-
mine the onset temperature of self-condensation between
phosphonic acid groups, step-isothermal TGA experiments
were carried out between 100 and 200 °C, by increasing the
temperature from 100 °C in intervals of 10 °C with a rate of 10
119
dx.doi.org/10.1021/cm202064x | Chem. Mater. 2012, 24, 115−122

Chemistry of Materials
Article
dependent upon the side-chain length. This higher-temperature
softening is likely due to the molecular relaxations in the
polystyrene-like phase-separated domain.
Proton Conductivity Measurements. The conductivity
measurements of both PVBPA-Ethyl and PVBPA-Hexyl were
performed using broadband dielectric relaxation spectroscopy
(DRS) under dry nitrogen gas flow. The results of the
frequency dependent conductivity measurements are shown in
Figure 4 as a function of frequency at various temperatures.
Figure 3. (a) Thermal analysis of PVBPA-Ethyl and PVBPA-Hexyl
samples by DSC. (b) Thermal relaxation analysis by DMA
penetration-method under 0.5 N load for PVBPA-Ethyl and PVBPA-
Hexyl samples. The data for the reference sample of polystyrene with a
known Tg of 100 °C are shown in the inset. The Tg value was
determined at the intersect of two extension lines. The average error is
approximately 2 °C.
dependence on the side-chain length; the DSC Tgs of the two
materials differ by less than a few degrees. However, the probe
penetration measurements using the DMA instrument show a
very interesting thermal softening response in the region above
the DSC Tg as shown in Figure 3b. This particular technique
may be slightly more sensitive to the molecular motions within
the materials that do not give rise to detectable jump in the heat
capacity by DSC due to a very broad transition. Since the DMA
data start at 20 °C, essentially the midpoint of the DSC Tg, we
cannot quantitatively compare the DSC Tg to the softening
transition measured by DMA. However, there appears to be an
additional, higher softening temperature, at T ≈ 101 °C for the
PVBPA-Ethyl and T ≈ 74 °C for the PVBPA-Hexyl. Although
only a single, similar Tg was observed for both samples by DSC,
it is likely that the mechanical measurements are more sensitive
to subtle differences in the molecular motions that give rise to
bulk mechanical relaxations. It is not surprising that PVBPA-
Hexyl would have a slightly lower temperature for softening
than PVBPA-Ethyl because the increased flexibility of the hexyl
group is likely to facilitate molecular motions at lower
temperatures. The samples appear relatively glassy at room
temperature, so the transition measured by DSC could be due
to molecular motions within phosphonic acid aggregated
channels or network as described for other ion-containing
polymers. There does appear to be some softening of the
material near the calorimetric Tg, but significant softening does
not occur until much higher temperatures and is also
Figure 4. AC conductivity versus frequency of (a) PVBPA-Ethyl and
(b) PVBPA-Hexyl at various temperatures. The standard error for the
reported conductivity is on the order of 0.2% of the magnitude of the
conductivity.
Each sample is characterized by a frequency independent (DC)
conductivity plateau. These plateau regions develop at
frequencies below the segmental relaxation, and move to
higher frequencies as the segmental relaxation moves to higher
frequencies. There is a sharp drop in the conductivity on the
low frequency side of this plateau which is caused by electrode
polarization. The temperature-dependent direct current (DC)
conductivity of the samples was estimated from plateau values
in the AC conductivity in a frequency scan at each temperature,
ignoring the electrode polarization effects.21,23,41 The DC
conductivities are plotted as a function of inverse temperature
in Figure 5. The data shown include both samples over a
heating and cooling cycle. It appears that the two polymers
containing different side-chains have similar conductivity
behavior in the measured temperature regime. This may be
due to the fact that the hexyl system has a lower volume
fraction of acid groups, but forms a better organized acid
aggregate structure for charge transport, consistent with the
SAXS results in Figure 2.
120
dx.doi.org/10.1021/cm202064x | Chem. Mater. 2012, 24, 115−122

Chemistry of Materials
Article
conditions may be attributed to the highly aggregated proton-
conducting acid channels with mobile hydrogen bond network,
as evidenced by the existence of the nanophase separated
domains shown by SAXS results. It was previously reported in
the case of immobilized imidazoles that longer alkyl chain
tethered protogenic groups result in higher conductivity.13,47,48
This phenomenon was explained by chain flexibility which
enables less restriction for aggregation of protogenic group and
hydrogen bond formation,12,49 but no supporting structural
evidence was provided. Similarly, Steininger et al.50 investigated
cyclic siloxanes with tethered phosphonic acid groups via
flexible spacers, but no structural information on aggregation or
nanophase separation was shown.
CONCLUSIONS
■
In this paper we have reported the synthesis and character-
ization of phosphonic acid functionalized polystyrene deriva-
tives. Poly(vinylbenzyloxy-alkyl-phosphonic acid)s with two
different alkyl side chains (ethyl and hexyl) are successfully
synthesized by Michaelis−Arbuzov and other synthetic
methods. The synthesized polymers exhibit a narrow
polydispersity index of about 1.4 and molecular masses of
approximately Mw ≈ 24 000 g/mol by aqueous gel permeation
chromatography (GPC) with PEG standards and thermal
stability up to 140 °C. The phosphonic acid groups undergo
condensation with the loss of water above 140 °C. Both
polymers show nanophase-separated morphology with charac-
teristic domain spacing of ca. 2 nm, which is attributed to a
bilayer lamellar-type organization of the surfactant-like side
chains. The proton conductivities are found to be approx-
imately 2.6 × 10−4 S/cm at 140 °C under nominally anhydrous
condition by broadband dielectric relaxation spectroscopy. The
proton transport may occur through Grotthuss mechanism
(structure diffusion) which consists of aggregated protogenic
charge carriers and dynamic hydrogen bond formation. Further
optimized nanophase-separation morphology and molecular
dynamics of percolated phosphonic-acid channels would lead to
improved polymer electrolyte membranes for anhydrous fuel
cells with higher proton conductivity and better mechanical
properties.
Figure 5. Temperature-dependent proton conductivity of the
synthesized polymers determined by broadband dielectric relaxation
spectroscopy. The line represents an average fit of the VFT equation
to the data in the high-temperature range. The two vertical lines
denote the Tg estimated from To and the calorimetric Tg, respectively..
The four plots for the two polymers during heating and cooling are
hardly distinguishable and hence are treated as a single plot. The
average error associated with each point is approximately 10% of the
magnitude of the measured conductivity, σ.
We have performed a semiquantitative fit to the data with
Vogel−Fulcher−Tamman (VFT) equation:
⎡
σ = σoexp⎢
⎣
⎤
−B
⎥
⎦
T − T
o
where σo is the conductivity at infinite temperature (4.75 ×
10−3 S/cm), B is a constant with units of temperature (550K),
and To is the temperature below which ion motion should cease
(so-called Vogel−Fulcher temperature, usually ∼50 °C below
the normal glass transition temperature Tg42,43). From a
qualitative perspective, the VFT equation does a reasonable job
at parametrizing the data above the temperature range denoted
by the two vertical lines in Figure 5. As the temperature drops
below this region, the data follow an Arrhenius-type behavior.
The agreement of the high temperature portion with the VFT
equation indicates that the charge motion is coupled to the
motions of the polymer segments. This is likely facilitated
through a Grotthus hopping mechanism.44−46 A fit of the VFT
equation through all data can be forced, but the fitting
parameters then become nonphysical and are no longer useful.
From a more quantitative analysis, the Tg estimated from the To
(258 K) was found to be approximately 35 °C (the vertical line
on 3.25 [1000/K]), which is comparable to the Tg measured by
DSC (20 °C, the vertical line on 3.41 [1000/K]). Moreover,
the experimental data start to deviate from the VFT equation at
the temperatures below these two vertical lines. In conclusion,
it seems that the fitting parameters are consistent with the
proton conduction taking place in the phosphonic acid
aggregated channels, wherein significant segmental motions
occur above the Tg indicated by the DSC data.
ASSOCIATED CONTENT
■
S
* Supporting Information
1H NMR and 13C NMR spectra of all the synthesized
compounds. This material is available free of charge via the
Internet at http://pubs.acs.org
AUTHOR INFORMATION
Corresponding Author
*E-mail: dyyoon@snu.ac.kr (D.Y.Y.); kirt.page@nist.gov
(K.A.P.).
■
ACKNOWLEDGMENTS
■
We thank Dr. W. Meyer and Prof. G. Wegner for helpful
discussions and advice. This work was supported by the
Chemistry and Molecular Engineering Program of the Brain
Korea 21 Project, LG Chem., Ltd., and National Research
Foundation of Korea (R01-2008-000-11971-0).
Finally, it is important to note that the proton conductivity of
our samples under nominally dry conditions at 25 °C was
found to be approximately 3 × 10−8 S/cm, which is much
higher than the values from poly(vinylbenzylphosphonic acid)
(1 × 10−10 to 1 × 10−11 S/cm).30,31 The much higher proton
conductivities in these polymers under nominally dry
REFERENCES
■
(1) Barbir, F. PEM Fuel Cells: Theory and Practice; Elsevier: London,
2005.
121
dx.doi.org/10.1021/cm202064x | Chem. Mater. 2012, 24, 115−122

Chemistry of Materials
Article
(2) Winter, M.; Brodd, R. J. Chem. Rev. 2004, 104, 4245.
(3) Zawodzinski, J. T. A.; Derouin, C.; Radzinski, S.; Sherman, R. J.;
Smith, V. T.; Springer, T. E.; Gottesfeld, S. J. Electrochem. Soc. 1993,
140, 1041.
(4) Basu, S. Recent Trends in Fuel Cell Science and Technology;
Springer: New York, 2007.
(43) Ratner, M. A.; Nitzan, A. Faraday Discuss. Chem. Soc. 1989, 88,
19.
(44) Steinkam, K.; Schumacher, J. O.; Goldsmith, F.; Ohlberger, M.;
Ziegler, C. J. Fuel Cell Sci. Technol. 2008, 5, 011007−1−16.
(45) Lott, K. F.; Ghosh, B. D.; Ritchie, J. E. J. Electrochem. Soc. 2006,
153, A2044−A2048.
(46) Lott, K. F.; Ghosh, B. D.; Ritchie, J. E. Electrochem. Solid-State
Lett. 2005, 8, A513−A515.
(5) Smitha, B.; Sridhar, S.; Khan, A. A. J. Membr. Sci. 2005, 259, 10.
(6) Kreuer, K. D. J. Membr. Sci. 2001, 185, 29.
(47) Schuster, M.; Meyer, W. H.; Wegner, G.; Herz, H. G.; Ise, M.;
Kreuer, K. D.; Maier, J. Solid State Ionics 2001, 145, 85.
(48) Persson, J. C.; Jannasch, P. Chem. Mater. 2003, 15, 3044.
(49) Kreuer, K.-D. Chem. Phys. Chem. 2002, 3, 771.
(50) Steininger, H.; Schuster, M.; Kreuer, K. D.; Maier, J. Solid State
Ionics 2006, 177, 2457.
(7) Gubler, L.; Scherer, G. G. Desalination 2010, 250, 1034.
(8) Schuster, M. F. H.; Meyer, W. H. Annu. Rev. Mater. Res. 2003, 33,
233.
(9) Kreuer, K.-D.; Paddison, S. J.; Spohr, E.; Schuster, M. Chem. Rev.
2004, 104, 4637.
(10) Steininger, H.; Schuster, M.; Kreuer, K. D.; Kaltbeitzel, A.;
Bingol, B.; Meyer, W. H.; Schauff, S.; Brunklaus, G.; Maier, J.; Spiess,
H. W. Phys. Chem. Chem. Phys. 2007, 9, 1764.
(11) Wainright, J. S.; Wang, J. T.; Weng, D.; Savinell, R. F.; Litt, M. J.
Electrochem. Soc. 1995, 142, L121.
(12) Kerres, J. A. J. Membr. Sci. 2001, 185, 3.
(13) Schuster, M. F. H.; Meyer, W. H.; Schuster, M.; Kreuer, K. D.
Chem. Mater. 2004, 16, 329.
(14) Persson, J. C.; Jannasch, P. Macromolecules 2005, 38, 3283.
(15) Kreuer, K. D.; Fuchs, A.; Ise, M.; Spaeth, M.; Maier, J.
Electrochim. Acta 1998, 43, 1281.
(16) Herz, H. G.; Kreuer, K. D.; Maier, J.; Scharfenberger, G.;
Schuster, M. F. H.; Meyer, W. H. Electrochim. Acta 2003, 48, 2165.
(17) de Grotthuss, C. J. T. Ann. Chim. 1806, 58, 54.
(18) Agmon, N. Chem. Phys. Lett. 1995, 244, 456.
(19) Schuster, M.; Rager, T.; Noda, A.; Kreuer, K. D.; Maier, J. Fuel
Cells 2005, 5, 355.
(20) Bingol, B.; Meyer, W. H.; Wagner, M.; Wegner, G. Macromol.
Rapid Commun. 2006, 27, 1719.
(21) Bozkurt, A.; Meyer, W. H.; Gutmann, J.; Wegner, G. Solid State
Ionics 2003, 164, 169.
(22) Erdemi, H.; Bozkurt, A. Eur. Polym. J. 2004, 40, 1925.
(23) Sevil, F.; Bozkurt, A. J. Phys. Chem. Solids 2004, 65, 1659.
(24) Yamada, M.; Honma, I. Polymer 2005, 46, 2986.
(25) Yamada, M.; Honma, I. Angew. Chem., Int. Ed. 2004, 43, 3688.
(26) Yamada, M.; Honma, I. J. Phys. Chem. B 2004, 108, 5522.
(27) Rager, T.; Schuster, M.; Steininger, H.; Kreuer, K.-D. Adv.
Mater. 2007, 19, 3317.
(28) Parvole, J.; Jannasch, P. Macromolecules 2008, 41, 3893.
(29) Rusanov, A. L.; Kostoglodov, P. V.; Abadie, M. J. M.;
Voytekunas, V. Y.; Likhachev, D. Y. Fuel Cells II; Springer-Verlag:
Berlin, 2008; Vol. 216, p 125.
(30) Jiang, F.; Kaltbeitzel, A.; Meyer, W. H.; Pu, H.; Wegner, G.
Macromolecules 2008, 41, 3081.
(31) Jiang, F.; Kaltbeitzel, A.; Fassbender, B.; Brunklaus, G.; Pu, H.;
Meyer, W. H.; Spiess, H. W.; Wegner, G. Macromol. Chem. Phys. 2008,
209, 2494.
(32) Schmidt-Rohr, K.; Chen, Q. Nat. Mater. 2008, 7, 75.
(33) Bhattacharya, A. K.; Thyagarajan, G. Chem. Rev. 1981, 81, 415.
(34) Boutevin, B.; Hamoui, B.; Parisi, J.-P.; Amed
́
uri, B. Eur. Polym. J.
1996, 32, 159.
(35) Rabinowitz, R. J. Am. Chem. Soc. 1960, 82, 4564.
(36) Rabinowitz, R. J. Org. Chem. 1963, 28, 2975.
(37) Souzy, R.; Ameduri, B.; Boutevin, B.; Virieux, D. J. Fluorine
Chem. 2004, 125, 1317.
(38) Ma, T.-Y.; Zhang, X.-J.; Yuan, Z.-Y. J. Phys. Chem. C 2009, 113,
12854−12862.
(39) Malone, S. M.; Trabelsi, S.; Zhang, S.; Lee, T. R.; Schwartz, D.
K. J. Phys. Chem. B 2010, 114, 8616−8620.
(40) Haubold, H.-G.; Vad, Th.; Hiller, H. J. P. Electrochim. Acta 2001,
46, 1559.
(41) Bozkurt, A.; Ise, M.; Kreuer, K. D.; Meyer, W. H.; Wegner, G.
Solid State Ionics 1999, 125, 225.
(42) Schoenhals, A.; Kremer, F.; Hofmann, A.; Fischer, E. W.;
Schlosser, E. Phys. Rev. Lett. 1993, 70, 3459.
122
dx.doi.org/10.1021/cm202064x | Chem. Mater. 2012, 24, 115−122