Ion clustering in quaternary ammonium functionalized benzylmethyl containing poly(arylene ether ketone)s

Article abstract of DOI:10.1021/ma401620m

Quaternary ammonium functionalized poly(arylene ether ketone)s (QA-PAEKs) with mass-based ion exchange capacities (IEC_g) ranging from 1.12 to 2.88 mequiv g^-1 were synthesized via condensation polymerization of a newly designed highly benzylmethylated bisphenol, subsequent bromination of the benzylmethyl groups, and then quaternization with trimethylamine. The quaternary ammonium groups were densely and selectively anchored on the bis(3,5-dimethyl-4-hydroxyphenyl)-3,5-dimethylphenylmethane residues in the QA-PAEK backbone. ¹H and ¹³C nuclear magnetic resonance (NMR) spectroscopy, Fourier transform infrared spectroscopy (FTIR), and X-ray photoelectron spectroscopy (XPS) were used to confirm the chemical structure of the samples. The morphology, mechanical properties, thermal stabilities, water uptake, swelling ratio, and bromide (Br^-) conductivity of the QA-PAEK membranes were investigated. It was found that QA-PAEK samples had much higher Br^- conductivity than a randomly functionalized quaternary ammonium Radel (QA-Radel) membrane at similar IEC, which was attributed to the existence of distinct ion clusters in the QA-PAEK materials as evidenced by small-angle X-ray scattering (SAXS). The Br^- conductivity of the QA-PAEK membranes increased with increasing quaternary ammonium ion concentration, water uptake, temperature, and conducting volume fraction. The results indicated that ion clustering was important for enhanced ion conductivity, and QA-PAEKs are promising anion exchange membranes for use in energy and water treatment applications.

Full text of DOI:10.1021/ma401620m

Article
pubs.acs.org/Macromolecules
Ion Clustering in Quaternary Ammonium Functionalized
Benzylmethyl Containing Poly(arylene ether ketone)s
Dongyang Chen and Michael A. Hickner*
Department of Materials Science and Engineering, The Pennsylvania State University, University Park, Pennsylvania 16802, United
States
S
* Supporting Information
ABSTRACT: Quaternary ammonium functionalized poly-
(arylene ether ketone)s (QA-PAEKs) with mass-based ion
exchange capacities (IECg) ranging from 1.12 to 2.88 mequiv
g−1 were synthesized via condensation polymerization of a
newly designed highly benzylmethylated bisphenol, subse-
quent bromination of the benzylmethyl groups, and then
quaternization with trimethylamine. The quaternary ammo-
nium groups were densely and selectively anchored on the
bis(3,5-dimethyl-4-hydroxyphenyl)-3,5-dimethylphenylme-
thane residues in the QA-PAEK backbone. 1H and 13C nuclear
magnetic resonance (NMR) spectroscopy, Fourier transform
infrared spectroscopy (FTIR), and X-ray photoelectron
spectroscopy (XPS) were used to confirm the chemical structure of the samples. The morphology, mechanical properties,
thermal stabilities, water uptake, swelling ratio, and bromide (Br−) conductivity of the QA-PAEK membranes were investigated.
It was found that QA-PAEK samples had much higher Br− conductivity than a randomly functionalized quaternary ammonium
Radel (QA-Radel) membrane at similar IEC, which was attributed to the existence of distinct ion clusters in the QA-PAEK
materials as evidenced by small-angle X-ray scattering (SAXS). The Br− conductivity of the QA-PAEK membranes increased with
increasing quaternary ammonium ion concentration, water uptake, temperature, and conducting volume fraction. The results
indicated that ion clustering was important for enhanced ion conductivity, and QA-PAEKs are promising anion exchange
membranes for use in energy and water treatment applications.
Ion clustering is a key factor for enhanced ion conductivity in
PEMs. The importance of the existence of the ionic domain
morphology is a well-established finding from Nafion
membranes and has been validated by work on several aromatic
sulfonated polymers that display distinct ion clustering, or
hydrophilic−hydrophobic phase separation, as observed by
scattering and microscopy technuques.13,14 The hydrophilic
acidic groups in PEMs tend to segregate from the hydrophobic
polymer backbones during the membrane casting process. The
phase connectivity and domain size can be modulated by the
polymer backbone composition and chain topology.12,13 To
facilitate the formation of ionic domains, measures can be taken
to place the acid groups on polymer grafts or at the ends of
chains15 or clustered in certain segments of the polymer
chains16−19 including densely functionalized ionic monomers or
block copolymers. It has been found in many cases that the ion
conductivity of PEMs with obvious ion clustering is much
higher than that of the PEMs without ion clustering, owing to
the fact that ion clustering contributes to the formation of the
hydrophilic channels for the fast diffusion of water and
INTRODUCTION
■
Quaternary ammonium functionalized polymers are well-
known cationic polyelectrolytes that have found numerous
successful applications in the chemical, medical, and biological
fields.1,2 Membrane-based processes for water-insoluble cati-
onic polymers include fuel cells,3,4 flow batteries,5 water
electrolyzers,6,7 electrodialysis,8 and reverse electrodialysis.9 In
these membranes, positively charged quaternary ammonium
groups are tethered to the polymeric backbones while the
negatively charged counterions become mobile upon solvation.
Therefore, these materials are termed anion exchange
membranes (AEMs) where the anion motion can be used for
the transport of charge or the fixed cationic groups can be
employed for exclusion of mobile cations. Currently, knowledge
concerning the structure−morphology−performance relation-
ships of AEMs is much less developed compared to that of
sulfonated proton exchange membranes (PEMs) with fixed
anionic charges on the polymer backbone.10 Many types of
PEMs have been studied in depth over the past decade,
primarily for fuel cell applications, and insights into their
material properties, ionic domain structures, and their
interaction with water molecules have been widely re-
ported.11,12 However, the most promising strategies for
producing high-performance AEMs are not yet clear.
Received: August 1, 2013
Revised: November 11, 2013
Published: November 22, 2013
© 2013 American Chemical Society
9270
dx.doi.org/10.1021/ma401620m | Macromolecules 2013, 46, 9270−9278

Macromolecules
Article
conduction of ions through the membrane.12 Inspired by these
findings in PEMs, quaternary ammonium functionalized block
copolymers have been synthesized and enhanced ion
conductivity has been observed.20−26 However, only a few
such systems have been reported so far. Block copolymers are a
promising model system, but other easy-to-access methods of
creating phase separation in AEMs are desirable.27
powder was obtained as crude product, which was washed with
deionized water and then recrystallized from toluene−ethanol (volume
ratio 3:1) solutions twice. A new bisphenol monomer, bis(3,5-
dimethyl-4-hydroxyphenyl)-3,5-dimethylphenylmethane, was obtained
with ∼40% yield. The reaction route is depicted in Scheme 1.
Scheme 1. Synthesis of Bis(3,5-dimethyl-4-hydroxyphenyl)-
3,5-dimethylphenylmethane
Many existing AEM studies are focused on improving the
alkaline stability of cationic membranes by screening different
cation groups such as imidazolium,28 phosphonium,29
guanidinium,30 sulfonium,31 and pyridinium moieties.32 The
alkaline stability of AEMs is particularly important for alkaline
fuel cells where high pH, low relative humidity, and elevated
temperature are the target operating conditions.20 Many useful
concepts have been demonstrated in this space, but currently
no widespread breakthrough for stabilizing AEMs against
degradation has been established over the baseline of simple
alkyltrimethyl and benzyltrimethyl quaternary ammonium
groups.33−35 Therefore, most device demonstrations at low
pH conditions employ materials with quaternary ammonium
functionalization because of their ease of synthesis and
reasonable performance. A challenging issue for the currently
available AEMs is their low anion conductivities, primarily
because of the low mobility of anions as compared to the
mobility of protons in PEMs.36 Therefore, manipulating the
morphology of AEMs to increase their anion conductivity is
extremely important for next-generation materials.
The synthesis of quaternary ammonium functionalized AEMs
generally involves halomethylation of a polymer and subse-
quent reaction with the desired amine precursors.37−39 For the
halomethylation, chloromethylmethyl ether is the most popular
reagent, which is highly carcinogenic.39 Furthermore, only one
site on a benzene ring is functionalizable by chloromethyl-
methyl ether before side reactions and cross-linking occur.
Alternatively, N-bromosuccinimide is a much safer halomethy-
lation reagent than chloromethylmethyl ether and can impart
multiple functional groups onto a methylated benzene ring
depending on the number of benzylmethyl substituents.40,41 In
order to develop high performance AEMs with superior
morphology for anion conductivity from low-risk, scalable
synthetic steps, herein we report the synthesis of quaternary
ammonium functionalized benzylmethyl containing poly-
(arylene ether ketone)s (QA-PAEKs) using N-bromosuccini-
mide as halomethylation reagent. The quaternary ammonium
groups were densely anchored at the predesigned benzyl
positions on the monomer units for efficient phase separation.
Detailed physicochemical properties of QA-PAEKs were
measured and correlated to the ion conducting properties of
the membranes.
Polymer Synthesis. The synthesis and subsequent functionaliza-
tion of the polymers in this work are depicted in Scheme 2. First, the
above obtained monomer, bis(3,5-dimethyl-4-hydroxyphenyl)-3,5-
dimethylphenylmethane, was polymerized with bisphenol A and
4,4′-difluorobenzophenone to produce benzylmethyl containing poly-
(arylene ether ketone)s (PAEKs) with 10, 20, 30, 40, and 50 mol % of
the total bisphenol loading being bis(3,5-dimethyl-4-hydroxyphenyl)-
3,5-dimethylphenylmethane. Target stoichiometies of bisphenol to
activated dihalide were 1:1 mol:mol to ensure high molecular weight
step-growth copolymerization. A typical polymerization procedure
follows: to a 100 mL three necked round-bottom flask equipped with a
Dean−Stark trap, an argon inlet and outlet, a condenser, a magnetic
stirrer, 2.0850 g (15.0 mmol) of K2CO3, 30 mL of DMSO, 15 mL of
toluene, 1.8025 g (5.0 mmol) of bis(3,5-dimethyl-4-hydroxyphenyl)-
3,5-dimethylphenylmethane, 1.1415 g (5.0 mmol) of bisphenol A, and
2.1820 g (10.0 mmol) of 4,4′-difluorobenzophenone were charged.
The reaction was carried out at 145−150 °C for 3 h, and then the
toluene was distilled from the reaction mixture with the produced
water. The temperature was raised to 170 °C and maintained for 15 h.
Afterward, the viscous mixture was poured into water to form the
fibrous product PAEK-50, which was collected by filtration. The
number in the sample name represents the mole percentage of bis(3,5-
dimethyl-4-hydroxyphenyl)-3,5-dimethylphenylmethane to the total
bisphenol monomer content in the synthesis of the polymer. All the
synthesized PAEKs were obtained with ∼99% yield.
The PAEK samples were brominated using N-bromosuccinimide
(NBS) as the bromination agent. A typical procedure was given as
follows: to a 250 mL flask, 2.4 g (5.0 mmol) of PAEK-50, 100 mL of
dichloroethane, 2.7 g of N-bromosuccinimide (NBS) (15.0 mmol),
and 0.2 g of benzoyl peroxide (BPO) (0.8 mmol) were added. The
reaction was carried out at 80 °C for 5 h under argon. Upon cooling to
room temperature, the reaction mixture was poured into 500 mL of
methanol. Fibrous product was formed and collected by filtration. The
product (Br-PAEK-50) was washed with 100 mL of acetone three
times, dissolved in chloroform, precipitated in methanol, collected by
filtration, and dried under vacuum at 40 °C for 24 h. The prefix “Br” in
the sample name designates brominated precursor polymer before
quaternization.
After bromination, the polymers were quaternized by trimethyl-
amine in N,N′-dimethylacetamide (DMAc) solution using the
Menshutkin reaction at room temperature.33 In a typical procedure,
3.2 g (5 mmol) of brominated Br-PAEK-50 was dissolved in 25 mL of
DMAc in a glass vial. 1.8 g of trimethylamine solution (45 wt % in
water) was added, and the container was sealed. The mixture was
magnetically stirred at room temperature for 24 h, then poured into a
Petri dish, and dried at 70 °C under ambient pressure for 24 h. A solid
product (QA-PAEK-50) was obtained. The prefix “QA” designates
“quaternary ammonium” functionalized polymer.
Membrane Preparation. QA-PAEK samples were redissolved in
DMAc as ∼8 wt % solutions, cast onto leveled glass plates, and dried at
70 °C under ambient pressure for 24 h followed by vacuum drying at
100 °C for an additional 24 h. The dried membranes were peeled from
the casting substrate and immersed in deionized water for 24 h to
EXPERIMENTAL METHODS
■
Materials. Quaternary ammonium functionalized Radel (QA-
Radel) with an ion exchange capacity (IEC) of 2.36 mequiv g−1 was
synthesized according to a previous report.42 All other chemicals were
purchased from common commercial suppliers and used as received.
Bis(3,5-dimethyl-4-hydroxyphenyl)-3,5-dimethylphenylme-
thane Monomer Synthesis. 13.4 g (0.1 mol) of 3,5-dimethylben-
zaldehyde and 36.7 g (0.3 mol) of 2,6-dimethylphenol were dissolved
in 200 mL of toluene in a 250 mL three-neck round-bottom flask with
magnetic stirring at room temperature (25 °C) under the protection of
argon. 5 mL of sulfuric acid (98 wt %) was added dropwise to the
solution, and then the reaction temperature was increased to 70 °C
and held for 24 h. Afterward, the reaction was cooled to room
temperature, and the precipitate was collected by filtration. A yellowish
9271
dx.doi.org/10.1021/ma401620m | Macromolecules 2013, 46, 9270−9278

Macromolecules
Article
Scheme 2. Synthesis of PAEKs, Br-PAEKs, and QA-PAEKs
remove any remaining DMAc residue. The membranes were then
stored in fresh deionzied water until use.
and titrated with 0.1 M AgNO3 using K2CrO4 as a colorimetric
indicator. The IECg was calculated from
Characterization. 1H nuclear magnetic resonance (NMR) and 13
C
ΔVAgNO CAgNO
NMR spectra were recorded on a Bruker DRX-400 NMR instrument
and the chemical shifts were listed in parts per million (ppm)
downfield from tetramethylsilane (TMS). FTIR spectra were recorded
on a Bruker Vertex 70 FTIR spectrometer in transmission. The
thermal stability of the QA-PAEK samples was analyzed using a Q50
TGA (TA Instruments Corp.). The temperature was increased from
room temperature to 100 °C and held for 30 min to remove the
absorbed water from the sample and then increased to 650 °C at a
heating rate of 10 °C min−1 under a N2 atmosphere. Gel permeation
chromatography (GPC) analysis was performed on a Waters Breeze
system equipped with a Waters Styragel column, Waters 515 HPLC
pump, and Waters 2414 refractive index detector. Chloroform was
used as the elution solvent with the flow rate of 1 mL min−1. Nearly
monodisperse polystyrene standards were used for GPC calibration.
The mechanical properties of the wet membranes were measured on
an Instron 5866 universal testing machine at room temperature. Small-
angle X-ray scattering (SAXS) patterns were collected on a Rigaku
(formerly Molecular Metrology) instrument with a pinhole camera
with Osmic microfocus source and parallel beam optic. The
instrument was equipped with a Cu target (λ = 1.542 Å) and a
multiwire area detector. Measurements were carried out at room
temperature under vacuum. X-ray photoelectron spectroscopy (XPS)
was recorded on Kratos Analytical Axis Ultra XPS instrument with
charge corrected to C at 285 eV.
3
3
IECg =
md
(1)
where md is the mass of the dry membrane, ΔVAgNO is the consumed
3
volume of AgNO3 solution, and CAgNO is the concentration of AgNO3
3
solution.
The water uptake (WU) of the membranes was defined as weight
ratio of the absorbed water to that of the dry membrane as given by
⎛
⎜
⎝
⎞
⎟
⎠
mw − md
WU =
× 100%
md
(2)
where md and mw are the weight of membranes before and after water
absorption, respectively. The hydration numbers (λ) were calculated
from
⎛
⎝
⎞
⎠
⎛
⎜
⎝
⎞
⎟
⎠
mw − md
1000
M × IEC
H2O
⎜
⎜
⎟
⎟
λ =
md
(3)
where M
is the molecular weight of water. The water uptake at
H2O
different relative humidities (RHs) at 30 °C was measured by a TA
Instruments Q5000SA water vapor sorption microbalance.
The concentration (c) of quaternary ammonium ions in the wet
membranes was calculated from the titrated IECg and water uptake via
the equation
The mass-based ion exchange capacities (IECg) of the membranes
were calculated from the 1H NMR spectra and confirmed by titration.
For the 1H NMR method, the mole percent of the quaternary
ammonium groups was calculated from the integral ratio of the
methylene peak to the polymer backbone peaks. For titration, 0.2 g of
membrane was immersed in 50 mL of 0.2 M NaNO3 solution for 24 h
IECg
c =
1
WU
+
ρ
100ρ
(4)
polymer
water
9272
dx.doi.org/10.1021/ma401620m | Macromolecules 2013, 46, 9270−9278

Macromolecules
Article
where ρpolymer is the density of the membranes (membrane mass
divided by membrane volume) and ρwater is the density of water (1 g
cm−3).
confirmed its chemical structure. The benzylmethyl groups in
this monomer were designed as highly clustered reaction sites
for quaternary ammonium functionalization. Condensation
polymerization of this benzylmethyl-containing monomer
with bisphenol A and 4,4′-difluorobenzophenone was carried
out using nucleophilic aromatic step-growth copolymeriza-
tion,33,44 and high molecular weight benzylmethyl containing
polymers (PAEKs) were obtained as shown in Table 1.
Subsequent bromination of the polymer was conducted
successfully at the benzylmethyl positions using NBS, as
The ion conductivities of the membranes were measured by two-
probe electrochemical impedance spectroscopy (EIS) using a
Solartron 1260A frequency response analyzer as previously reported.43
Fully hydrated conductivity measurements were carried out in
deionized water. A temperature and relative humidity controlled
chamber (Espec SH-241) was employed for RH conductivity
measurements. The Br− conductivity (σBr−) was calculated from the
impedance plot according to the cross-sectional area available for ion
transport (A) and the distance between electrodes (d), σ = d/RA,
where the resistance, R, was derived from the intercept of the low
frequency complex impedance with the Re(Z′) axis. The Arrhenius
activation energy (Ea) for ion conduction was calculated from the
slope of ln σBr− versus 1/T through
1
evidenced by the new H NMR peak at 4.25−4.51 ppm in
1
Figure 2b compared to the H NMR of the unfunctionalized
Ea
RT
−
ln σBr = ln σ0 −
(5)
where σBr− is the ion conductivity, σ0 is the pre-exponential factor, R is
the gas constant, and T is the absolute temperature.
RESULTS AND DISCUSSION
■
Monomer and Polymer Synthesis. A new multi-
benzylmethyl containing bisphenol monomer, bis(3,5-dimeth-
yl-4-hydroxyphenyl)-3,5-dimethylphenylmethane, was synthe-
sized successfully from 3,5-dimethylbenzaldehyde and 2,6-
1
dimethylphenol. The H and 13C NMR spectra in Figure 1
Figure 2. 1H NMR spectra of (a) PAEK-20, (b) Br-PAEK-20, and (c)
QA-PAEK-20.
PAEK in Figure 2a. Around 70 mol % of the benzylmethyl
groups were brominated under our reaction conditions based
1
on the H NMR results. While we did find that increasing the
loading of NBS and BPO or increasing the reaction
temperature could lead to higher degrees of bromination,
reduced molecular weight polymer was obtained.40 Therefore,
we did not pursue materials with higher degrees of bromination
in this work. In Table 1, the molecular weight of PAEKs did not
change significantly after bromination, while the polydispersity
indices (PDI) were slightly higher for the brominated samples
compared to the starting material. Treating Br-PAEK samples
with trimethylamine converted the bromomethyl groups to
quaternary ammonium groups in quantitative yield, as
Figure 1. 1H NMR (upper) and 13C NMR (lower) spectra of bis(3,5-
dimethyl-4-hydroxyphenyl)-3,5-dimethylphenylmethane.
Table 1. GPC Results for the PAEK Samples before and after Bromination
before bromination
after bromination
sample
Mn (kg mol−1
)
Mw (kg mol−1
)
PDI
Mn (kg mol−1
)
Mw (kg mol−1
)
PDI
PAEK-10
PAEK-20
PAEK-30
PAEK-40
PAEK-50
34.4
34.7
34.4
51.3
45.2
65.4
74.0
62.5
98.8
95.0
1.9
2.4
1.9
1.9
2.1
32.3
36.1
35.7
31.3
41.0
63.6
94.3
79.0
65.7
94.3
2.1
2.6
2.3
2.1
2.3
9273
dx.doi.org/10.1021/ma401620m | Macromolecules 2013, 46, 9270−9278

Macromolecules
Article
Table 2. Properties of QA-PAEK Copolymers
1
density
IECg from H NMR
IECg from titration
c
water uptake
(wt %)
swelling ratio
(%)
Ea
sample
(g cm−3
)
(mequiv g−1
)
(mequiv g−1
)
(mmol cm−3
)
(kJ mol−1
)
QA-PAEK-10
QA-PAEK-20
QA-PAEK-30
QA-PAEK-40
QA-PAEK-50
1.01
1.14
1.20
1.26
1.32
1.13
1.80
2.33
2.65
2.92
1.12
1.79
2.23
2.49
2.88
1.01
1.75
2.11
2.31
2.46
8
14
23
29
41
3
8
27
26
26
26
26
13
20
33
1
evidenced by the H NMR spectrum in Figure 2c. QA-PAEK
materials were obtained with this simple final step.
To give more chemical structure information on the QA-
PAEK samples, the FTIR spectrum of QA-PAEK-30 was
measured, as shown in Figure S1 (Supporting Information).
The absorption peaks of the constituent benzene groups,
methyl groups, and ketone groups were all indexed in the
spectrum; however, the absorption peaks for the C−N bond in
the quaternary ammonium groups were not discernible, similar
to other reported quaternary ammonium containing polymer
systems.24,29 This absence of the C−N vibrational signature in
the QA moieties may be attributed to the low absorption
intensity of the C−N bond and the overlap of C−N peaks with
the absorption peaks of other functional groups in the polymer.
In order to detect the presence of the quaternary ammonium
groups in a straightforward way, XPS was employed. A clear N
1s signal at 399.8 eV was observed in the XPS spectrum (Figure
S2). The peaks for the counteranion, Br−, were also detected at
255.8, 181.8, and 68.8 eV, which corresponded to the Br 3s, 3p,
and 3d emissions, respectively. All these data supported the
successful synthesis of the target QA-PAEK samples with
backbone-tethered quaternary ammonium groups.
Figure 3. SAXS patterns of QA-PAEK samples and QA-Radel.
or length scale correlation peaks were observed, and their
intensities decreased with an increase in IECg. The interdomain
spacing distances (d) were found to be 8.2, 7.4, 6.4, 5.3, and 4.9
nm for QA-PAEK-10, QA-PAEK-20, QA-PAEK-30, QA-PAEK-
40, and QA-PAEK-50, respectively. This result demonstrated
that there were ion clusters in QA-PAEKs membranes, which
originated from the densely anchored quaternary ammonium
groups on a specific segment of the polymer backbone. The ion
clusters were more organized in low IECg QA-PAEK samples
than in high IECg QA-PAEK materials because of increased
phase mixing and shorter nonionic segment length in high IECg
QA-PAEKs.
The IECgs for QA-PAEKs are listed in Table 2. The
quaternary ammonium groups had methyl and methylene
1
protons in the H NMR spectra. The integral ratio of these
protons to the aromatic protons from the polymer backbone
was used to calculate the IECgs of the samples. It can be seen
that the IECg increased monotonically with an increasing
amount of bis(3,5-dimethyl-4-hydroxyphenyl)-3,5-dimethyl-
phenylmethane monomer residue in the polymers, as expected.
The titrated IECgs agreed well the IECgs from 1H NMR (Table
2). The concentration of ions, c in Table 2, is another
important parameter for the hydrated ionomers.45 It can be
seen that c increased with an increasing amount of bis(3,5-
dimethyl-4-hydroxyphenyl)-3,5-dimethylphenylmethane used
in the synthesis of the polymer. The rate of increase in c was
decreased as the IECg grew larger, suggesting that QA-PAEKs
had not yet reached but were approaching their highest ion
concentration before significant swelling set in an ion dilution
occurred.
Membrane Morphology. The quaternary ammonium
groups in QA-PAEK copolymers were anchored densely on
the bis(3,5-dimethyl-4-hydroxyphenyl)-3,5-dimethylphenylme-
thane monomer residues to facilitate the formation of ion
clusters. To validate this membrane morphology hypothesis
according to the polymer chemical structure, SAXS was
employed, and the results are shown in Figure 3. A QA-
Radel membrane with quaternary ammonium groups dispersed
randomly along the poly(arylene ether sulfone) backbone from
chloromethylation functionalization was chosen for compar-
ison. The SAXS pattern of the QA-Radel sample did not show
an ionomer peak, indicating that no ionic clustering was present
in the dry sample. For QA-PAEK membranes, obvious ionomer
Membrane Properties under Fully Hydrated Con-
ditions. The room temperature water uptake of QA-PAEK
samples increased with increasing IECg, ranging from 8 to 41
wt %, as shown in Table 2. Water is critical for ion pair
separation and ion conductivity in AEMs; however, higher
water uptake led to larger swelling ratios as shown in Table 2.
All of the reported membranes were flexible and mechanically
strong. The Young’s moduli of the QA-PAEK samples were
considerably higher than that of the N212 membrane (Table
S1). The yield strain and yield stress of QA-PAEK samples
decreased with increasing water uptake because of the
plasticization by water. Except for QA-PAEK-50, all of the
QA-PAEK samples had higher yield stresses than N212. The
elongation at break values for all aromatic-based QA-PAEK
samples were lower than for N212. Elongation at break is a
critical value for membranes as it is a measure of their resistance
to cracking under cyclic stress. The values for the QA-PAEK
samples were acceptable, and all membranes were not brittle in
our characterization measurements.
For anion conductivities, the Br− conductivity is generally
lower than the OH− and Cl− conductivities because of the low
dilute solution mobility of Br− and less water uptake in halide
form samples compared to bicarbonate and hydroxide form
samples.46 However, the conducting ion species in AEMs varies
across different applications so it is valuable to assess the anion
9274
dx.doi.org/10.1021/ma401620m | Macromolecules 2013, 46, 9270−9278

Macromolecules
Article
conductivity of AEMs with a number of different mobile ions.
We have observed previously that poly(arylene ether) back-
bones are susceptible to degradation under basic conditions.47
Therefore, only Br− conductivity was reported herein for
fundamental AEM studies. The Br− conductivities of QA-
PAEKs at room temperature are shown in Figure 4 along with
Figure 5. Br− conductivity of QA-PAEK samples as a function of
temperature.
promising in comparison to block anion exchange membranes,
given the fact that the diffusion coefficient of Br− (1.85 × 10−9
m2 s−1) is similar to the diffusion coefficient of Cl− (1.77 × 10−9
m2 s−1) at infinite dilution in water at 25 °C.50
Figure 4. Br− conductivity of QA-PAEK samples and QA-Radel at
room temperature in liquid water.
Membrane Properties under Partially Hydrated
Conditions. Partially hydrated conditions are extremely
important for investigating AEMs because in many cases the
AEMs are hydrated only through water vapor carried into the
cell, such as in gas-fed fuel cells. Also, surveying the range of
properties of AEMs as a function of water content provides
insights into how the hydration of the sample affects the ion
conductivity and morphology changes during the absorption of
water.51 The water uptake of the QA-PAEK samples as a
function of relative humidity (RH) is shown in Figure 6. The
the conductivity of QA-Radel, for comparison. The Br−
conductivity of the QA-PAEK samples increased gradually
with increasing IEC, while the Br− conductivity of the QA-
Radel sample was below the trend line for the QA-PAEK
materials. This observation was in good accordance with the
morphology results from SAXS. QA-Radel is a randomly
functionalized polymer that displays no evidence of nanophase
structure in SAXS. In this case for QA-Radel, the ionic groups
do not form obvious ion-rich domains to constitute the
hydrophilic channels for efficient Br− conduction, similar to the
case in some PEMs.48 However, QA-PAEK samples had
densely clustered ionic groups due to the highly methylbenzy-
lated monomer and thus showed obvious ionic clustering in the
membrane. The ion conduction in the QA-PAEK membranes
was therefore much greater than in the QA-Radel sample.
Increasing the temperature increased the water uptake,
swelling ratio, and Br− conductivity of the QA-PAEK samples.
As shown in Figures S3 and S4, the water uptake and swelling
ratio of the membranes increased gradually with increasing
temperature, and at any given temperature, the water uptake
and swelling ratio increased with increasing IECg. The Br−
conductivities of QA-PAEKs showed an approximate Arrhe-
nius-type temperature dependence promoted by the thermal
activation of water motion49 (Figure 5). The apparent
activation energy estimated from the slopes of the ln σBr− vs
1/T plots was ∼26 kJ mol−1 and was similar among all the
tested membranes (Table 2). The Br− conductivity for QA-
PAEK-50 increased from 11.8 to 58.9 mS cm−1 as the
temperature increased from 20 at 80 °C, the highest among
the QA-PAEK samples. Sudre et al.21 reported poly(styrene)-
based block anion exchange membranes with tethered
quaternary ammonium or imidazolium groups. The quaternary
ammonium sample SAM (27−21) had an IEC of ∼3.0 mequiv
g−1 and a Cl− conductivity of ∼4 mS cm−1 at 25 °C. The
imidazolium sample SIM (27−21) had an IEC of ∼2.1 mequiv
g−1 and a Cl− conductivity of 1.8 mS cm−1 at 25 °C. While the
water uptake values of these linear and non-cross-linked
samples were not reported, one can still estimate that the
conductivities of the QA-PAEK membranes should be
Figure 6. Water uptake of QA-PAEK samples as a function of relative
humidity.
water uptake increased gradually with increasing RH,
approaching the water uptake of the samples under fully
hydrated conditions. Samples with higher IECs absorbed more
water than the samples with lower IECs at any given RH. The
Br− conductivity of QA-PAEKs increased greatly with
increasing RH (Figure 7). A small increase in RH led to
significant increase in Br− conductivity. Therefore, water uptake
was critically important for ion conduction. Ye et al.52 reported
high conductivity anion exchange membrane with block
structures. Their sample, poly(MMA-b-MEBIm-Br-17.3)
(IECg = ∼1.4 mequiv g−1 without counteranion), had ∼17 wt
% water uptake and ∼1.0 mS cm−1 Br− conductivity at 90% RH
9275
dx.doi.org/10.1021/ma401620m | Macromolecules 2013, 46, 9270−9278

Macromolecules
Article
Figure 7. Br− conductivity of QA-PAEK membranes as a function of
relative humidity.
Figure 9. Schematic showing (left) low IEC, organized ionic domains,
and (right) high IEC, disordered domains.
at 30 °C. Our sample, QA-PAEK-30 (IECg = 2.23 mequiv g−1
with Br− as counteranion), had 16 wt % water uptake and 1.1
mS cm−1 of Br− conductivity at 90 RH at 30 °C. While the
IECgs were different, these two samples had similar Br−
conductivity at similar water uptake, demonstrating the
promising aspects of the QA-PAEK samples, which are still
random copolymers, albeit with designed clustered ionic
domains through the novel benzylmethylated monomer.
To understand the influence of water uptake on the Br−
conductivity of these materials, Br− conductivity was plotted
versus water uptake as shown in Figure 8. At fixed water uptake,
and high IEC QA-PAEK membranes are schematically
displayed in Figure 9. Low IEC membranes with well phase
separated ionic domains were observed in the SAXS data in
Figure 3. As the IEC increased, the SAXS peak became less
distinct and the ionic domains appear to be more disordered.
With their high ion content and disorganized domain structure,
the high IECg QA-PAEKs needed more water to promote Br−
conduction than the low IECg QA-PAEKs.
The hydration number, λ (moles of water per mole of
quaternary ammonium group), in this series of materials was
linked to Br− conductivity (Figure 10). The Br− conductivities
Figure 8. Br− conductivity of QA-PAEK samples as a function of water
uptake.
Figure 10. Br− conductivity of QA-PAEK samples as a function of
hydration number.
low IECg QA-PAEK membranes had marginally higher Br−
conductivity than the samples with high IECg. At fixed Br−
conductivity, low IECg QA-PAEK membranes required less
water uptake than the high IECg QA-PAEK samples. These
interesting results allude to ion conduction in highly hydrated
regions of the membranes with organized ionic domains.53
High IECg QA-PAEK samples had more ionic sites to bind
water and exhibited more phase mixing than low IECg QA-
PAEK membranes from the SAXS data in Figure 3. Thus, the
high IECg samples had more bound water solvating the ionic
groups at the same overall water content compared to the low
IECg QA-PAEK samples with organized ionic domains. The
data in Figure 8 demonstrate that adding more water to the
system does not necessarily increase conductivity unless that
water is loosely bound or free and available to participate in ion
conduction. The differences in morphology between low IEC
of QA-PAEK-30, QA-PAEK-40, and QA-PAEK-50 samples
exhibited a similar dependence on λ, suggesting that the
morphology of these samples was fully interconnected and
gains in conductivity were only achieved by increased
hydration. For QA-PAEK-10 and QA-PAEK-20, the low
concentration of quaternary ammonium groups in these
materials resulted in reduced Br− conductivity even when
they had similar λ to the other QA-PAEK samples.
Since ion conduction occurs over the entire length scale of
the membrane, on the order of micrometers, the influence of
the membrane conducting volume on the ion conductivity is of
great interest. The conducting volume should be composed of
the volumes of water and ionic groups in the membrane.
Stimulated by Kim and Pivovar’s work on calculating the water
9276
dx.doi.org/10.1021/ma401620m | Macromolecules 2013, 46, 9270−9278

Macromolecules
Article
volume fraction of the membranes by van der Waals volume
increments,54 we included the molar volume of the quaternary
ammonium groups with the water volume as the conducting
volume using the same method. The conducting volume
fraction (Xv) was obtained via the equation
control sample without obvious ion clustering. The Br−
conductivity of QA-PAEKs showed an approximate Arrhe-
nius-type temperature dependence under fully hydrated
conditions. Under partially hydrated conditions, it was found
that high IECg QA-PAEK samples required more water than
low IECg QA-PAEK membranes to achieved similar Br−
conductivity, perhaps due to better ionic domain morphology
in the low IECg membranes. The Br− conductivity of QA-
PAEK samples increased with increasing conducting volume
fraction (Xv) of the samples; however, Xv cannot be simply
compared in different QA-PAEK samples because of the
changes in ion hydration and water dynamics as the IEC and λ
of the samples vary. These results suggested that evaluating
different conducting volume character, the connectivity of the
ionic domains, and the water motion within the material is
critically important in understanding the ion conductivity of
AEMs.
18λ + VQA
Xv =
∑ niV + 18λ
i
where Vi is the contribution of the ith structural group which
appeared ni times per charge. VQA is the molar volume of
quaternary ammonium group with Br− as counteranion. The
molar volumes of all the structural groups were obtained from
ref 55.
The relationship between Br− conductivity and Xv for the
QA-PAEK membranes is shown in Figure 11. It can be seen
ASSOCIATED CONTENT
* Supporting Information
■
S
FTIR spectrum of QA-PAEK-30, XPS spectrum of QA-PAEK-
30, water uptake of QA-PAEK samples as a function of
temperature, swelling ratio of QA-PAEK samples as a function
of temperature, TGA curves of QA-PAEK membranes, and
tensile properties of QA-PAEK and Nafion N212 membranes.
This material is available free of charge via the Internet at
http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*Tel +1 814 867 1847; Fax +1 814 865 2917; e-mail hickner@
matse.psu.edu (M.A.H.).
■
Figure 11. Br− conductivity of QA-PAEK samples as a function of
conducting volume fraction.
Author Contributions
D.C. designed the new monomer, carried out the experiments,
and drafted the manuscript. M.A.H. contributed to the analysis
of the data and results, developed descriptions of the
structure−property relationships of these samples, and helped
to refine the manuscript language. All authors have given
approval to the final version of the manuscript.
that the Br− conductivity increased with increasing Xv for each
QA-PAEK sample, but the higher Xv in the high IECg QA-
PAEK samples did not guarantee their higher Br− conductivity
compared to the low IECg QA-PAEK samples. It seems that the
Xv cannot be simply compared in different QA-PAEK samples
as factors including water binding and dynamics need to be
taken into account. Xv is composed of both the volume
occupied by water and the volume occupied by ionic groups,
and the ratio of these volumes is different for different samples.
For a given water volume, there are free water molecules and
bound water molecules whose distributions are dependent on
the nature of the samples. Therefore, evaluating both ionic and
water conducting volumes, their connectivity, and the dynamics
within conducting domains is of critical importance in the
understanding of the ion conductivity of AEMs. More work on
these topics is needed as design, synthesis, analysis, and
deployment of AEMs continue.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The work was supported by the Office of Electricity (OE
Delivery & Energy Reliability (OE), U.S. Department of
Energy (DOE), under Contract DE-AC05-76RL01830. We
also acknowledge the support from the Southern Pennsylvania
Ben Franklin Commercialization Institute (Grant #001389-
002) and the Virginia S. and Philip L. Walker, Jr., Faculty
Fellowship to M.A.H. from the College of Earth and Mineral
Sciences at The Pennsylvania State University. Dr. Chen thanks
Ms. Melanie L. Disabb-Miller for the measurement of RH water
uptake and Dr. Nanwen Li for the measurement of SAXS.
CONCLUSIONS
■
Quaternary ammonium functionalized benzylmethyl containing
poly(arylene ether ketone)s (QA-PAEKs) with clustered ionic
groups and IECgs ranging from 1.12 to 2.88 mequiv g−1 were
synthesized successfully from a newly designed mononer for
high performance anion exchange membranes. The densely
functionalized quaternary ammonium groups in the samples
formed distinct ion clusters in solution-cast membranes, as
evidenced by SAXS measurements. A significant increase in the
ion conductivity was observed when comparing the QA-PAEK
membranes to a quaternary ammonium functionalized Radel
REFERENCES
■
(1) Gibbs, C. F.; Marvel, C. S. J. Am. Chem. Soc. 1935, 57, 1137−
1139.
(2) Buhler, D. R.; Thomas, R. C.; Christensen, B. E.; Wang, C. H. J.
Am. Chem. Soc. 1955, 77, 481−482.
(3) Couture, G.; Alaaeddine, A.; Boschet, F.; Ameduri, B. Prog. Polym.
Sci. 2011, 36, 1521−1557.
(4) Merle, G.; Wessling, M.; Nijmeijer, K. J. Membr. Sci. 2011, 377,
1−35.
9277
dx.doi.org/10.1021/ma401620m | Macromolecules 2013, 46, 9270−9278

Macromolecules
Article
(5) Chen, D.; Hickner, M. A.; Agar, E.; Kumbur, E. C. Electrochem.
Commun. 2013, 26, 37−40.
(6) Leng, Y.; Chen, G.; Mendoza, A. J.; Tighe, T. B.; Hickner, M. A.;
Wang, C. Y. J. Am. Chem. Soc. 2012, 134, 9054−9057.
(7) Xiao, L.; Zhang, S.; Pan, J.; Yang, C.; He, M.; Zhuang, L.; Lu, J.
Energy Environ. Sci. 2012, 5, 7869−7871.
(8) Sata, T. J. Membr. Sci. 2000, 167, 1−31.
(9) Guler, E.; Zhang, Y.; Saakes, M.; Nijmeijer. ChemSusChem 2012,
5, 2262−2270.
(10) Hickner, M. A. Mater. Today 2010, 13, 34−41.
(11) Kreuer, K. D. J. Membr. Sci. 2001, 185, 29−39.
(12) Peckham, T. J.; Holdcroft, S. Adv. Mater. 2010, 22, 4667−4690.
(13) Elabd, Y. A.; Hickner, M. A. Macromolecules 2011, 44, 1−11.
(14) Li, N.; Wang, C.; Lee, S. Y.; Park, C. H.; Lee, Y. M.; Guiver, M.
D. Angew. Chem. 2011, 123, 9324−9327.
(15) Matsumura, S.; Hlil, A. R.; Hay, A. S. J. Polym. Sci., Part A:
Polym. Chem. 2008, 46, 6365−6475.
(16) Yang, Y.; Shi, Z.; Holdcroft, S. Macromolecules 2004, 37, 1678−
1681.
(42) Chen, D.; Hickner, M. A.; Agar, E.; Kumbur, E. C. ACS Appl.
Mater. Interfaces 2013, DOI: 10.1021/am401858r.
(43) Fujimoto, C. H.; Hickner, M. A.; Cornelius, C. J.; Loy, D. A.
Macromolecules 2005, 38, 5010−5016.
(44) Chen, D.; Wang, S.; Xiao, M.; Meng, Y.; Hay, A. S. J. Mater.
Chem. 2011, 21, 12068−12077.
(45) Kim, Y. S.; Einsla, B.; Sankir, M.; Harrison, W.; Pivovar, B. S.
Polymer 2006, 47, 4026−4035.
(46) Lin, X.; Varcoe, J. R.; Poynton, S. D.; Liang, X.; Ong, A. L.; Ran,
J.; Li, Y.; Xu, T. J. Mater. Chem. A 2013, 1, 7262−7269.
(47) Nunez, S. A.; Hickner, M. A. ACS Macro Lett. 2013, 2, 49−52.
(48) Zhao, C.; Lin, H.; Li, X.; Na, H. Polym. Adv. Technol. 2011, 22,
2173−2181.
(49) Kornyshev, A. A.; Kuznetsov, A. M.; Spohr, E.; Ulstrup, J. J.
Phys. Chem. B 2003, 107, 3351−3366.
(50) Koneshan, S.; Rasaiah, J. C.; Lynden-Bell, R. M.; Lee, S. H. J.
Phys. Chem. B 1998, 102, 4193−4204.
(51) Varcoe, J. R. Phys. Chem. Chem. Phys. 2007, 9, 1479−1486.
(52) Ye, Y.; Sharick, S.; Davis, E. M.; Winey, K. I.; Elabd, Y. A. ACS
Macro Lett. 2013, 2, 575−580.
(17) Shin, C. K.; Maier, G.; Andreaus, B.; Scherer, G. G. J. Membr.
(53) Hickner, M. A.; Fujimoto, C. H.; Cornelius, C. J. Polymer 2006,
47, 4238−4244.
Sci. 2004, 245, 147−161.
(18) Lee, H. S.; Roy, A.; Lane, O.; Dunn, S.; McGrath, J. E. Polymer
2008, 49, 715−723.
(54) Kim, Y. S.; Pivovar, B. S. ECS Trans. 2009, 25, 1425−1431.
(55) Van Krevelen, D. W. Properties of Polymers, 3rd ed.; Elsevier:
New York, 1990; p 87.
(19) Chen, L.; Hallinan, D. T.; Elabd, Y. A.; Hillmyer, M. A.
Macromolecules 2009, 42, 6075−6085.
(20) Tanaka, M.; Fukasawa, K.; Nishino, E.; Yamaguchi, S.; Yamada,
K.; Tanaka, H.; Bae, B.; Miyatake, K.; Watanabe, M. J. Am. Chem. Soc.
2011, 133, 10646−10654.
(21) Sudre, G.; Inceoglu, S.; Cotanda, P.; Balsara, N. P. Macro-
molecules 2013, 46, 1519−1527.
(22) Lee, K. M.; Wycisk, R.; Litt, M.; Pintauro, P. N. J. Membr. Sci.
2011, 383, 254−261.
(23) MacKnight, W. J.; Kajiyama, T.; McKenna, L. Polym. Eng. Sci.
1968, 8, 267−271.
(24) Cheng, S.; Beyer, F. L.; Mather, B. D.; Moore, R. B.; Long, T. E.
Macromolecules 2011, 44, 6509−6517.
(25) Hemp, S. T.; Smith, A. E.; Bryson, J. M.; Allen, M. H.; Long, T.
E. Biomacromolecules 2012, 13, 2439−3445.
(26) Gao, R.; Wang, D.; Heflin, J. R.; Long, T. E. J. Mater. Chem.
2012, 22, 13473−13476.
(27) Li, N.; Leng, Y.; Hickner, M. A.; Wang, C. Y. J. Am. Chem. Soc.
2013, 135, 10124−10133.
(28) Qiu, B.; Lin, B.; Qiu, L.; Yan, F. J. Mater. Chem. 2012, 22, 1040−
1045.
(29) Gu, S.; Cai, R.; Luo, T.; Chen, Z.; Sun, M.; Liu, Y.; He, G.; Yan,
Y. Angew. Chem., Int. Ed. 2009, 48, 6499−6502.
(30) Wang, J.; Li, S.; Zhang, S. Macromolecules 2010, 43, 3890−3896.
(31) Dewi, E. L.; Oyaizu, K.; Nishide, H.; Tsuchida, E. J. Power
Sources 2003, 115, 149−152.
(32) Huang, A.; Xia, C.; Xiao, C.; Zhuang, L. J. Appl. Polym. Sci. 2006,
100, 2248−2251.
(33) Chen, D.; Hickner, M. A. ACS Appl. Mater. Interfaces 2012, 4,
5775−5781.
(34) Page, O. M. M.; Poynton, S. D.; Murphy, S.; Ong, A. L.;
Hillman, D. M.; Hancock, C. A.; Hale, M. G.; Apperley, D. C.; Varcoe,
J. R. RSC Adv. 2013, 3, 579−587.
(35) Deavin, O. I.; Murphy, S.; Ong, A. L.; Poynton, S. D.; Zeng, R.;
Herman, H.; Varcoe, J. R. Energy Environ. Sci. 2012, 5, 8584−8597.
(36) Chen, D.; Hickner, M. A.; Wang, S.; Pan, J.; Xiao, M.; Meng, Y.
Int. J. Hydrogen Energy 2012, 37, 16168−16176.
(37) Gu, S.; Skovgard, J.; Yan, Y. S. ChemSusChem 2012, 5, 843−848.
(38) Zeng, Q. H.; Liu, Q. L.; Broadwell, I.; Zhu, A. M.; Xiong, Y.; Tu,
X. P. J. Membr. Sci. 2010, 349, 237−243.
(39) Tanaka, M.; Koike, M.; Miyatake, K.; Watanabe, M. Polym.
Chem. 2011, 2, 99−106.
(40) Yan, J.; Hickner, M. A. Macromolecules 2010, 43, 2349−2356.
(41) Zhao, Z.; Wang, J.; Li, S.; Zhang, S. J. Power Sources 2011, 196,
4445−4450.
9278
dx.doi.org/10.1021/ma401620m | Macromolecules 2013, 46, 9270−9278