Side-chain sulfonated random and multiblock poly(ether sulfone)s for PEM applications

Article abstract of DOI:10.1016/j.reactfunctpolym.2011.04.006

Copoly(arylene ether sulfone)s from 4,4′-difluorodiphenyl sulfone, 4,4′-dihydroxydiphenyl sulfone bis-trimethylsilylether, and 2,5-diphenylhydroquinone bis-trimethylsilylether were obtained by nucleophilic displacement polycondensation with high molecular

Full text of DOI:10.1016/j.reactfunctpolym.2011.04.006

Reactive & Functional Polymers 71 (2011) 828–842
Contents lists available at ScienceDirect
Reactive & Functional Polymers
journal homepage: www.elsevier.com/locate/react
Side-chain sulfonated random and multiblock poly(ether sulfone)s
for PEM applications
Claus Vogel, Hartmut Komber, Annett Quetschke, Wladimir Butwilowski, Albrecht Pötschke,
⇑
Kornelia Schlenstedt, Jochen Meier-Haack
Leibniz Institute of Polymer Research Dresden, Hohe Strasse 6, 01069 Dresden, Germany
a r t i c l e i n f o
a b s t r a c t
Copoly(arylene ether sulfone)s from 4,40-difluorodiphenyl sulfone, 4,40-dihydroxydiphenyl sulfone bis-
trimethylsilylether, and 2,5-diphenylhydroquinone bis-trimethylsilylether were obtained by nucleophilic
displacement polycondensation with high molecular weights. While the random copolymers were
obtained with narrow molecular weight distributions, the molecular weight distributions of the multi-
block copolymers were much broader, probably indicating branching reactions. Random as well as mul-
tiblock copolymers showed a single glass transition temperature (Tg) at around 230 °C. Upon sulfonation
with concentrated sulfuric acid, the Tgs (from samples in the protonated form) were shifted to higher
temperatures. NMR spectra and the determined ion-exchange capacities, which were close to the theo-
retical values, indicated that the pendant phenyl rings of the 2,5-diphenylhydroquinone moieties in the
polymer backbone were sulfonated selectively. Membranes prepared from N-methyl-2-pyrrolidone solu-
tions were transparent and soft. The water uptake at room temperature increased from 30% to 80% with
increasing IEC. Samples with an IEC P1.8 mmol/g swell to a high extend (sMBC) or even dissolve in water
(sRC) at elevated temperatures. While the proton conductivities of the low IEC samples were lower than
or close to that of NafionÒ, the conductivities of the high IEC samples were superior to that of NafionÒ. In
general membranes from block copolymers showed similar water uptake and similar dimensional
changes but higher proton conductivities as compared to samples from random copolymers with similar
monomer composition and ion-exchange capacities.
Article history:
Received 17 December 2010
Received in revised form 19 April 2011
Accepted 23 April 2011
Available online 29 April 2011
Keywords:
Ion-exchange membranes
Poly(arylene ether sulfone)s
Side-chain sulfonation
PEM fuel cells
Ó 2011 Elsevier Ltd. All rights reserved.
1. Introduction
acid)s, as well as for the catalyst and on the other hand an opera-
tion temperature below 100 °C. However, an operation tempera-
Fuel cells are considered to be attractive and clean alternatives
to the widely employed internal combustion engines for power
production. Among the various types of fuel cells the low temper-
ature acidic polymer electrolyte membrane fuel cells (PEMFC) pos-
sess the broadest area of (possible) application, ranging from
power supply for small electronic devices over power source in
electrical cars to electrical and heat supply in household. One
key-component in PEMFC is the membrane which enables the
charge (proton) transport from the anode to the cathode side. Fur-
thermore the membrane should prevent the reactants from diffus-
ing to the opposite electrode. Despite intensive research activities
in the past decades, fuel cells are used to date only in niche-
markets. The reasons for not having reached the breakthrough in
mass-markets, are on one hand the costs for the membrane, the
standard is still NafionÒ or similar poly(perfluoroalkyl sulfonic
ture well above 100 °C would be preferred, because (a) the
catalytic activity is increased with increasing temperature, (b)
the catalysts are less susceptible to poisoning (faster desorption
of e.g. CO), both allowing the use of less amounts of catalysts,
and (c) water management is much easier due to the absence of li-
quid water [1]. Nowadays research in materials for ion-exchange
membranes is focused on sulfonated fully aromatic polymers such
as polyimides [2–5], poly(ether sulfone)s, polysulfones, poly(ether
ether ketone)s (for sulfonated poly(arylene ether)s see for example
the review [6] and references therein), or recently, poly(phenylene
sulfone)s [7–9]. The sulfonic acid group can be introduced into the
polymer chain either by the use of already sulfonated monomers or
by a post-sulfonation step. Regardless of the method employed for
the introduction of the sulfonic acid group, most often the polymer
main chain is sulfonated. As outlined e.g. by Iojoiu et al. [10], the
sulfonation step is the crucial step in the preparation of ion-
exchange membranes. Moreover sulfonation of the main chain
might have a negative impact on the mechanical as well as the
chemical properties of the polymer materials [11]. In order to cir-
⇑
Corresponding author. Tel.: +49 (0) 351 4658 519; fax: +49 (0) 351 4658 290.
E-mail address: mhaack@ipfdd.de (J. Meier-Haack).
1381-5148/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.reactfunctpolym.2011.04.006

C. Vogel et al. / Reactive & Functional Polymers 71 (2011) 828–842
829
cumvent these drawbacks, several research groups have prepared
side-chain sulfonated poly(arylene ether)s [12–20].
and calcium carbonate were purchased from Merck (Germany).
Potassium carbonate and hexamethyl disilazane (HMDS) were pur-
chased from Fluka (Germany) and decafluoro biphenyl (DFBP) from
FluoroChem (UK). Concentrated sulfuric acid (min. 96%) and tolu-
ene were obtained from Acros (Belgium). Aluminum trichloride
was purchased from ABCR (Germany).
The Jannasch group for example introduced sulfophenyl side-
groups directly to the polymer backbone by the reaction of lithiated
PSU with 2-sulfobenzoic acid cyclic anhydride [12,13]. In order to
enhance the sulfonation degree, the lithiated polymer was reacted
in a first step with 4-fluorobenzoyl chloride. In a second step the
fluorine moiety in the side-chain was converted by a nucleophilic
displacement reaction with highly sulfonated molecules like
2-naphthol-6,8-disulfonic acid or 8-hydroxy-1,3,6-pyrenetrisulfon-
ic acid [14]. Kim et al. [15], Liu et al. [16] and our group [17] reported
on side-chain sulfonated poly(arylene ether)s based on 2,5-dihy-
droxybiphenyl [15,17] and alkylated phenyl side-chains [16],
recently. Unlike the results reported in [15,16], we observed side-
chain as well as main-chain sulfonation when using 2,5-dihydroxy-
biphenyl as sulfonable monomeric unit. These different results can
be addressed to the electron withdrawing effects of monomers used
in Refs. [15,16] (decafluorobiphenyl, 2,6-difluorobenzonitrile, 4,40-
difluorobenzophenone) instead of 4,40-difluorodiphenyl sulfone
used in Refs. [17]. Random copoly(ether sulfone)s with 2,5-diphe-
nylhydroquinone in the polymer backbone showing enhanced ther-
mo-oxidative stabilities have been recently described by Liu et al.
[18]. On treatment with conc. sulfuric acid only sulfonation of the
side-chain was observed. Tian et al. reported on poly(ether sulfone)s
with highly sulfonated nanoclusters consisting of hexaphenylbiphe-
nol units in the main chain [19,20]. Moderate water-uptake and pro-
ton conductivities comparable to or even higher than NafionÒ were
reported. Promising results in single fuel cell tests were obtained
with the new membrane materials.
Vogel et al. reported on an unexpected high hydrolytic stability
of poly(styrene sulfonic acid) up to 200 °C in water [21] but
poly(styrene sulfonic acid) is not suitable for use in fuel cells due
to its low electrochemical and chemical stability. A fast degrada-
tion occurs during fuel cell operation on the cathode side due to
reactions at the tertiary carbon of the polymer backbone [22].
Due to the higher ion-exchange capacity that is needed for non-
fluorinated membrane materials compared to poly(perfluoroalkyl
sulfonic acid)s like NafionÒ to achieve a similar performance, these
membranes have a much higher water uptake combined with a
much lower dimensional stability. As shown by the McGrath
group, block copolymers have in this context superior properties
over random copolymers with similar chemical composition and
ion-exchange capacities [23–25].
All chemicals were used as received except 4,40-difluorodiphe-
nyl sulfone which was purified by vacuum distillation and NMP
which was distilled twice under reduced pressure from CaH2.
K2CO3 and CaCO3 were dried in vacuum at 150 °C for 24 h.
2.2. Characterization
Solution viscosities were measured in NMP with c = 2 g/L, using
an automated Ubbelohde viscosimeter thermostated at 25 °C.
Inherent viscosities were calculated from Eq. (1) where c is the
polymer concentration in g/dL and t and t0 are the running time
of the polymer solution and the pure solvent, respectively.
ꢀ
ꢁ
tꢀt0
ln 1 þ
t0
ginh
¼
ð1Þ
c
Differential scanning calorimetry (DSC) measurements were
carried out with a Netzsch DSC Phoenix 408 at a heating rate of
20 K/min in nitrogen atmosphere using aluminum pans. The sec-
ond heating scan was used to determine the Tg.
Detection of Tg of humidified samples was carried out in stain-
less steel high pressure pans at a heating rate of 5 K/min.
Mechanical properties of thin films were determined at room
temperature on a Zwick Z010 instrument. Samples were cut from
films into strips of 10 mm width and 100 mm length. The sample
length between clamps was set to 50 mm and a strain rate of
50 mm/min was applied. Before measurements samples were
equilibrated at 25 °C and 50% r.h. (dry samples) or were stored in
water at 25 °C (wet samples). The latter samples were taken from
the water just before the measurements.
NMR spectra were recorded on a Bruker DRX 500 spectrometer
operating at 500.13 MHz for 1H, 125.76 MHz for 13C and
470.59 MHz for 19F. 5 mm o.d. sample tubes were used. DMSO-
d6, chloroform-d1 or CDCl3/trifluoroacetic acid (TFA-d) mixtures
served as solvent and internal chemical shift reference (CHCl3: 1H
7.26 ppm, 13C 77.0 ppm; DMSO: 1H 2.50 ppm; 13C 39.6 ppm).
Hexafluorobenzene was used as external chemical shift reference
(ꢀ163 ppm) for 19F NMR spectra.
In this contribution we report on poly(ether sulfone)s based on
2,5-diphenyl hydroquinone, which are sulfonated at the side-chain
only. These polymers are mimicking on one hand poly(styrene sul-
fonic acid) (side chain sulfonation) but possessing on the other hand
a thermally, chemically and mechanically stable polymer backbone.
The degree of sulfonation and thus the ion-exchange capacities are
determined by the amount of 2,5-diphenylhydroquinone moieties
in the polymer backbone. Two different types of polymers, namely
random copolymers (RC) and multiblock copolymers (MBC), were
synthesized and their structure and composition was evaluated by
NMR spectroscopy. The effect of molecule architecture and ion-ex-
change capacity on thermal properties, water-uptake, dimensional
stability and proton conductivity will be discussed.
Molecular weights were obtained from GPC measurements on a
Knauer GPC equipped with Zorbax PSM Trimodal S columns and a
RI detector. A mixture of DMAc with 2 vol.% water and 3 g/L LiCl
was used as eluent. Poly(4-vinyl pyridine) samples served as stan-
dards for molecular weight calibration.
The ion-exchange capacities of sulfonated samples were deter-
mined by the titration method as described elsewhere [26].
The water uptake WU was measured by soaking a piece of dry
membrane in water for 24 h at 25 °C, 50 °C and 80 °C, respectively.
After removing the sample, excess water is wiped off gently with a
tissue, and the sample is weighed immediately to get the value
mwet. Subsequently the sample is dried in vacuum at 100 °C to con-
stant weight (mdry). The water uptake is calculated using Eq. (2)
and is given in percentage increase in weight, where mwet and mdry
are weights of the water-swollen and dry membrane sample,
respectively.
2. Experimental
2.1. Materials
mwet ꢀ mdry
WU ¼
ꢁ 100%
ð2Þ
mdry
4,40-Difluorodiphenyl sulfone (DFDPhS) was purchased from
FuMaTech GmbH (Germany). 4,40-dihydroxydiphenyl sulfone
(DHDPhS), 1,4-benzoquinone and sodium dithionite were obtained
from Aldrich (Germany). N-methyl-2-pyrrolidone (NMP), benzene
For evaluation of swelling properties, the dimensions (length,
width and thickness) of the membrane samples were determined
in the dry and wet state. The swelling is expressed in percentage

830
C. Vogel et al. / Reactive & Functional Polymers 71 (2011) 828–842
increase in dimensions (
in thickness).
D
L change in length and width;
D
d change
ether 23 g of a brownish solid remained. This powder was stirred
again with 100 mL diethyl ether. A white solid remained undis-
solved and was removed by filtration. The product was obtained
by removal of the diethyl ether and final drying in vacuum at
70 °C over night.
Yield: 20.4 g (78 mmol; 17%); melting point (DSC): 225.5 °C (Lit.
225 °C Ref. [27]). 1H NMR (DMSO-d6): d (ppm): 6.86 (s, 2H); 7.29 (t,
4H); 7.40 (t, 2H); 7.56 (d, 4H); 8.89 (s, 2H; OH).
The proton conductivities were measured in a temperature and
humidity controlled chamber using a four-point probe (FuMaTech
GmbH, Germany). The membrane conductivities (in-plane) were
calculated from cell geometry and membrane resistance (Eq. (3)),
which was obtained at the frequency that produced the minimum
imaginary response (phase angle close to zero). A Gamry Reference
600 potentiostat, operated over a frequency range from 1 Hz to
1 MHz, was used for the conductivity measurements.
2.4. Silylation of bisphenols
l
r
¼
ðS=cmÞ
ð3Þ
A 2L round-bottomed flask, equipped with a reflux condenser
and a drying tube (KOH), was charged with 1 mol of the respective
bisphenol, 1.1 mol of HMDS and 1 L toluene. The reaction mixture
was refluxed until the evolution of ammonia ceased. Toluene and
excess of HMDS were removed under reduced pressure and the
purified products were finally obtained in quantitative yield by
vacuum distillation (Table 1).
w ꢂ d ꢂ R
where l is the distance between electrodes (1 cm), w the width
(1.5 cm) and d the thickness of the sample and R the measured
resistance. The relative humidity (r.h.) in the sample compartment
was calculated from the saturated water vapor pressure in the
water compartment and in the sample compartment, at the ad-
justed temperatures and ambient pressure (1013 hPa), respectively
using the following equation:
2.5. Synthesis of random copolymers
SWVTsample
SWVTwater
r:h: ¼
ꢂ 100%
ð4Þ
10 mmol of DFDPhS, 10-x mmol of bis-TMS-DHDPhS and x
mmol of bis-TMS-DPhHQ were weighed into a 50 mL three-
necked flask equipped with a magnetic stirr bar, gas-inlet and
gas-outlet to which 30 mL of anhydrous and argon-saturated
NMP was added. After complete dissolution of the monomers a
1:5 mixture of anhydrous potassium carbonate and calcium car-
bonate (20 mmol with respect to carbonate) was added. The
reaction mixture was heated under stirring and argon purging
for 24 h to 175 °C and finally for 2 h at 190 °C. After cooling to
room temperature the viscous solution was diluted with 20 mL
of NMP and subsequently filtered for the removal of insoluble
material. The products were isolated by precipitation in metha-
nol and filtration. The solids were intensively washed with water
at rt and methanol at 50 °C. Finally the products were dried in
vacuum at 100 °C to constant weight.
where r.h. is the relative humidity in %, SWVTsample the saturated
water vapor pressure in the sample compartment at sample tem-
perature and SWVTwater the saturated water vapor pressure in the
water reservoir at water temperature. The values for the saturated
water vapor pressure were taken from Ref. [39].
2.3. Preparation of 2,5-diphenylhydroquinone
2,5-Diphenylhydroquinone (DPhHQ) was prepared by a slightly
modified procedure described by Shildneck and Adams [27]. A 1L
three-necked flask, equipped with a reflux condenser, a thermom-
eter and a 500 mL dropping funnel was charged with a suspension
of 150 g (1.125 mol) finely powdered aluminum chloride in 250 mL
dry benzene. A solution of 50 g (0.46 mol) 1,4-benzoquinone in
500 mL dry benzene was slowly added to the suspension via the
dropping funnel under stirring with a magnetic stirrer. The rate
of addition was regulated to maintain the temperature at 35–
40 °C. The stirring was continued overnight (17 h) while the tem-
perature was allowed to drop to room temperature. The Friedel–
Crafts-complex was decomposed by pouring the reaction mixture
slowly onto a mechanically stirred mixture of 1 kg ice and
125 mL of conc. hydrochloric acid. Stirring was continued until a
brown emulsion was formed. The benzene was removed by steam
distillation and the solid raw product was separated by filtration.
The filter cake was washed with hot water to obtain a green pow-
der (mixture of quinones, hydroquinones and quinhydrones). For
further work-up the product was dissolved in 500 mL diethyl
ether. Undissolved material was removed by filtration. The filtrate
was treated three times with a saturated sodium dithionite solu-
tion in a separating funnel for reduction of the quinones and quin-
hydrones. The ether layer was washed twice with water and finally
dried over sodium sulfate. After filtration and removal of diethyl
2.6. Synthesis of hydroxy endgroup functionalized oligomers
Two different types of end-group functionalized oligomers were
prepared in this study. One type was prepared from 4,40-difluorodi-
phenyl sulfone and 2,5-diphenyl hydroquinone bis-trimethylsilyl-
ether, which will be sulfonated later to introduce hydrophilicity
(oligo-DP-PES; diphenylated oligo-PES) and the other (oligo-PES)
from 4,40-difluorodiphenyl sulfone and 4,40-dihydroxydiphenyl sul-
fone bis-trimethylsilylether. The molecular weight and the end-
group functionality are predetermined by a stoichiometric offset
of the monomer feed ratios. In this study the molecular weight
(Mn) of the 2,5-diphenylhydroquinone containing oligomer was
set to approx. 10,000 g/mol. Molecular weights of the PES oligomers,
were designed to be approx. 5000 g/mol, 10,000 g/mol and 15,000 g/
mol, in order to make the ion-exchange capacities comparable to
those of the random copolymers. The molecular weights were deter-
mined by 1H NMR spectroscopy from the hydroxy-terminated olig-
omers using the integrals of the diphenylsulfone main chain protons
Table 1
Properties of silylated bisphenols.
Silylated bisphenol
Yield (%)
99
b.p.a/m.p.b
Literature
4,40-Dihydroxydiphenyl sulfone
230 °C 6 ꢁ 10ꢀ3 mbar
89 °C
90 °C [28]
2,5-Diphenyl hydroquinone
99
250 °C 2.5 x 10ꢀ3mbar
148.6 °C
–
a
Distillation via short-pass distillation: oil bath temperature.
The melting points were determined by DSC as maximum of the melting endotherm.
b

C. Vogel et al. / Reactive & Functional Polymers 71 (2011) 828–842
831
in ortho-position to the ether linkage (d = 7.27 ppm (oligo-PES,
2.7. Endcapping of hydroxy-terminated oligomers with decafluoro
biphenyl (DFBP)
solvent: DMSO-d6); d = 6.99 ppm (oligo-DP-PES, solvent: CDCl3/
TFA-d 4:1 v/v)) and the signal of the corresponding protons in the
ortho-position to the phenolic end-group (d = 6.92 ppm (oligo-
PES)) and in ortho-position to the terminal diphenyl hydroquinone
group (d = 6.94 ppm (oligo-DP-PES)), respectively. The synthetic
procedure was the same as for the random copolymers.
For end-capping the hydroxy-terminated oligomers, the meth-
od described by Lee et al. was employed [23]. 1 mmol of the
oligo-PES and 6 mmol of DFBP were weighed into a 100 mL
three-necked flask, equipped with a magnetic stirr bar, a gas-in-
O
O
TMS
a TMS
O
O
TMS + b TMS
O
S
O
+ (a + b)
F
S
F
O
O
NMP/CaCO /K CO
3
3
2
175°C 24h
O
S
O
S
O
S
O
O
O
O
O
O
O
x
a
b
RC-1: a=4; b=6
RC-2: a=5; b=5
RC-3: a=6; b=4
Scheme 1. Synthetic pathway to random copoly(arylene ether sufone)s by nucleophilic displacement polycondensation (TMS = trimethylsilyl); residual trimethylsilyl ether
groups are hydrolyzed during precipitation.
Scheme 2. Synthetic pathway to block copoly(arylene ether sufone)s by nucleophilic displacement polycondensation (TMS = trimethylsilyl); residual trimethylsilyl ether
groups are hydrolyzed during precipitation.

832
C. Vogel et al. / Reactive & Functional Polymers 71 (2011) 828–842
b' a'
a'' b''
O
S
A
B
HO
O
O
O
O
F
F
F
F
F
b''' a'''
a
b
O
S
F
O
F
F
F
a
b
a'''
b'''
B
A
a''
a'
b''
b'
8.4
8.2
8.0
7.8
7.6
7.4
7.2
7.0
6.8
δ
1H (ppm)
Fig. 1. 1H NMR spectra of hydroxy-terminated oligo-PES (A) before end-capping with decafluoro biphenyl and (B) after end-capping (solvent: DMSO-d6).
let and a gas-outlet. 50 mL of anhydrous and argon-saturated
NMP was added to the flask. After complete dissolution of the
reactants 4 mmol of anhydrous K2CO3 was added to the reaction
mixture and the temperature was raised to 100 °C under stirring
and argon purging for 24 h. After cooling to room temperature
the reaction mixture was filtered and the product was
coagulated in a 10-fold excess of methanol. The product was iso-
lated by filtration and washed intensively with water and
methanol. Finally the end-capped product was dried in
vacuum at 100 °C to constant weight. The molecular weight of
the end-capped oligomers was determined by NMR
spectroscopy.
2.9. Sulfonation of copolymers
The sulfonation of random as well as multiblock copolymers
was achieved by treatment with concentrated sulfuric acid (96–
98 wt.%) at room temperature. Approximately 3 g of the respective
polymer and 30 mL of conc. sulfuric acid were placed in a 50 mL
round-bottomed flask and the reaction mixture was stirred for
24 h. During the reaction time the polymer dissolves in the sulfo-
nating agent and the color changes from slightly yellow to brown-
ish-green. The sulfonated product was isolated by pouring the
reaction mixture into a large excess of isopropanol. The precipi-
tated product was filtered off and washed thoroughly with metha-
nol until neutral. The product was finally dried in vacuum at 110 °C
to constant weight.
2.8. Synthesis of multiblock copolymers
0.3 mmol of the end-capped oligo-PES and 0.3 mmol of the
2.10. Membrane preparation
respective oligo-DP-PES were weighed into
a 100 mL three-
necked round-bottomed flask, equipped with a magnetic stirr
bar, a gas-inlet and a gas-outlet. 50 mL of anhydrous and argon-
saturated NMP was added to the flask. After complete dissolution
of the reactants 2 mmol of anhydrous K2CO3 was added to the
reaction mixture and the temperature was raised to 100 °C under
stirring and argon purging for 24 h. After cooling to room temper-
ature the reaction mixture was filtered and the product was coag-
ulated in a 10-fold excess of methanol. The product was isolated
by filtration and washed intensively with water and methanol. Fi-
nally the product was dried in vacuum at 100 °C to constant
weight.
The sulfonated samples (in H+ form) were dissolved in NMP
(ꢃ15% wt./vol.) and stirred to give a homogeneous solution. Prior
to the casting process the solution was filtered through glass wool,
in order to remove undissolved matter. 3 mL of this solution was
poured in an aluminum frame (4 ꢁ 1.5 cm2) and dried at 80 °C in
vacuum for 24 h. The membrane samples were released by
immersing the frame into water. Remaining NMP was removed
by intensive washing with water and soaking the samples for at
least 24 h in 0.5 M sulfuric acid. The samples are washed thor-
oughly with DI water and finally stored in DI water at room tem-
perature until further characterization.

C. Vogel et al. / Reactive & Functional Polymers 71 (2011) 828–842
833
of the finally oligo-PES with decafluoro biphenyl. The complete con-
version of phenoxy endgroups was checked by 1H as well as 19F
NMR spectroscopy (Figs. 1 and 2). The 1H NMR spectrum of the hy-
droxyl-terminated oligo-PES (Fig. 1A) shows the expected signals of
the protons in ortho-position to the hydroxy end-group (b0) at
6.92 ppm, in meta-position to the hydroxy end-group (a0) at
7.77 ppm. The signals assigned to the corresponding main chain
protons (b and a) appear at 7.27 ppm (ortho to the ether group)
and at 7.99 ppm (ortho to the sulfone group). The signals of the pro-
tons attached to the second benzene ring of the terminal diphenyl-
sulfone moiety appear at 7.25 ppm (b00) and 7.92 ppm (a00). Upon
conversion of the hydroxy end-groups with decafluoro biphenyl,
the signals of the protons associated with the hydroxy end groups
disappear completely in the 1H NMR spectrum (Fig. 1B). The two
signals of lower intensity (b000 and a000) in this spectrum are assigned
to the protons in ortho- and meta-position to the newly formed
ether linkage with decafluoro biphenyl. The 19F spectrum (Fig. 2)
shows in addition to the signals of the nonafluoro biphenyl end-
group (assignment see figure) at least one signal (d0) appearing at
ꢀ138.3 ppm. From the 19F–19F 2D NMR spectrum (not shown) this
signal and a second one (e0), which is superimposed by the signal e,
are attributed to the formation of a symmetrical 4,40-substituted
decafluoro biphenyl structure (dimer formation). The dimer con-
tent is approx. 10% as estimated from the 19F NMR spectrum.
3. Results and discussion
3.1. Preparation and properties of random and multiblock copolymers
Random as well as multiblock copoly(arylene ether sulfone)s
from 4,40-difluorodiphenyl sulfone and different ratios of the bis-
trimethylsilylethers of 4,40-dihydroxydiphenyl sulfone and 2,5-di-
phenyl hydroquinone were successfully prepared by the so-called
‘‘silyl method’’ in solution using a 5:1 mixture of CaCO3 and
K2CO3 as base. It has been reported that the use of calcium or mag-
nesium salts in combination with small amounts of potassium car-
bonate facilitates the nucleophilic displacement polycondensation
[29–32]. This effect has been attributed to the removal of the highly
reactive fluoride ions, especially in the presence of trace amounts of
water, by the formation of insoluble earth alkali fluorides [30,31].
The synthetic pathways for the two different types of polymers pre-
pared in this study, random copolymers and multiblock copoly-
mers, are shown in Schemes 1 and 2. While the synthesis of the
random copolymers (RC-1–RC-3) is conducted by simply heating
of the respective amounts of monomers in NMP at 175 °C (Scheme
1), the synthesis of the block copolymers is more sophisticated
(Scheme 2). In this case the individual blocks with hydroxy end-
groups were prepared and characterized concerning their molecu-
lar weights (Mn) followed by conversion of the hydroxy end-groups
b
F
c
d
F
e
F
F
O
S
a
F
O
O
F
F
F
F
e'
F
d'
F
F
F
O
O
F
F
F
F
e, e'
b
c, d
a
d'
-140
-145
-150
19F (ppm)
-155
-160
δ
Fig. 2. 19F NMR spectrum of oligo-PES end-capped with decafluoro biphenyl (solvent: DMSO-d6).

834
C. Vogel et al. / Reactive & Functional Polymers 71 (2011) 828–842
In the next step the hydroxy-terminated oligomers are reacted
(a)
(b)
RC-2
sRC-2
a
b
with the decafluoro biphenyl-terminated oligomers to give the dif-
ferent multiblock copolymers (MBC-1–MBC-3). All polymers, ran-
dom as well as multi block copolymers, were obtained in almost
quantitative yield and with molecular weights up to 70.000
g/mol (Mn) (Table 2). The high molecular weights are expressed
by the high inherent viscosities (up to 0.93 dl/g) as well. However,
one feature is striking when taking a closer look at the polydisper-
sities (PD) of random and multiblock copolymers (Table 2). While
the random copolymers show relatively narrow PD in the range
from 2.3 to 2.9, the PD of the multiblock copolymers are in the
range from 6 to 8, although starting from oligomers with a PD of
around 2. Examples of gpc curves are given in Fig. 3 (random
copolymer RC-2) and Fig. 4 (multiblock copolymer MBC-2). The
broadening of the molecular weight distribution of multiblock
copolymers is explained by branching reactions at the decafluoro
biphenyl unit (Scheme 3). Some broadened 19F NMR signals of
low intensity were observed in the MBC spectra which might be
assigned to branching points. However, a conclusive proof of such
branching units is not straightforward because the intensity of
these signals is low and the substitution pattern of the branching
point result in a complex spectrum. However, the significant signal
broadening compared to the signals of the linear main chain sug-
gests a restricted mobility as expected for the branching units.
However, crosslinking was not observed since all multiblock
copolymers were soluble in dipolar aprotic solvents like NMP,
DMAc or DMSO. It should be noted, that the solubility of multi-
block copolymers is much better in amidic solvents like NMP or
DMAc than in DMSO, while the random copolymers dissolve read-
ily in all the above mentioned solvents.
On sulfonation with conc. sulfuric acid the molecular weights of
the polymers did not decrease as indicated by the viscosity data
and the gpc data (Table 3). On the contrary, the data suggest higher
molecular weights than the starting material. This effect is attrib-
uted firstly to the formation of aggregates due to the formation
of hydrogen bonds between the sulfonic acid groups of different
polymer chains. Such aggregation of ionic polymers has been re-
ported by several groups in the past [33–36]. Secondly, the solubil-
ities as well as interactions between the sulfonated polymer and
the solvent change, which leads to different chain conformations
(hydrodynamic radii), giving rise to the observed effect. However,
the molecular weight distribution keeps almost constant after
the sulfonation process, indicating that no degradation occurred
during the sulfonation step. The water uptake (WU) of sulfonated
random copolymers is increasing with increasing ion-exchange
capacity and increasing temperature. The values found in this
study are slightly higher but still comparable to those reported
by Liu et al. [18]. Even at 80 °C the samples sRC-2 and sRC-3 re-
tained their shape and showed moderate water uptake especially
when taking the relatively high IEC into account. Only the sample
1000
10000
100000
mol weight (g/mol)
1000000
1E7
Fig. 3. gpc curves of samples RC-2 (a, dashed line) and sRC-2 (b, solid line).
sRC-1, possessing an IEC of 2.05 mmol/g, dissolved at 80 °C. The
sulfonated multiblock copolymers showed a slightly higher water
uptake as compared to their random copolymers. This may result
from the fact that the sulfonic acid groups are distributed over
the whole polymer backbone in the case of the randon copolymers.
The non-sulfonated as well as the sulfonated sequences are much
shorter than in the blockcopolymers, leading to smaller respective
domains. These smaller domains prevent on one hand a high water
uptake due to a lower number of sulfonic acid groups. On the other
hand the resistive forces against swelling of the non-sulfonated do-
mains at high temperatures are also smaller than for the multi-
block copolymers. The sulfonic acid groups are incorporated as
highly sulfonated clusters (blocks) into the polymer chain of the
multiblock copolymers. Such clusters (blocks) might result in the
formation of larger sulfonated but also non-sulfonated domains
as compared to the random copolymers. These larger domains fa-
vor the water uptake while the larger non-sulfonated domains pre-
vent the sample from dissolving in water at high temperatures.
From this argumentation it is not surprising that the sample with
the highest IEC (sMBC-1) shows a high water uptake but does
not dissolve in water at 80 °C.
Upon water uptake the membrane materials show a dimen-
sional change. The in-plane swelling (length and width) is increas-
ing with increasing ion-exchange capacity and increasing
temperature as outlined in Table 4. In the case of the random
copolymers the determined values are in good agreement with
the values reported by Liu et al. [18]. All samples with low to med-
ium IEC showed good dimensional stability at temperatures up to
50 °C. There are no differences between the random and multi-
Table 2
Composition and properties of random copolymers (RC) and multiblock copolymers (MBC).
Sample
Feed ratio
DHDPhS
Product ratioa
DHDPhS
ginh (dl/g)
Mn (g/mol)
Mw (g/mol)
PD
Tg (°C)
DPhHQ
DPhHQ
RC-1
RC-2
RC-3
4
5
6
6
5
4
4
5
6
6.1
5.0
3.9
0.65
0.43
0.93
39,000
44,000
70,000
91,000
105,000
206,000
2.3
2.4
2.9
233
232
229
Mn oligo-DP-PESa,b(g/mol)
Mn oligo-PESa,c (g/mol)
MBC-1
MBC-2
MBC-3
9000
9200
9200
4600
10,600
14,500
1.08
1.15
1.64
44,000
31,800
55,000
285,000
244,000
316,000
6.5
7.7
5.8
224
–
233
a
b
c
Determined from 1H NMR spectra.
Composed of DPhHQ and DFDPhS.
Composed of DHDPhS and DFDPhS.

C. Vogel et al. / Reactive & Functional Polymers 71 (2011) 828–842
835
(a)
(b)
(c)
(d)
(e)
oligo-PES
oligo-DP-PES
oligo-DP-PES + decaF
MBC-2
b
c a
d
e
sMBC-2
1000
10000
100000
mol. weight (g/mol)
1000000
1E7
Fig. 4. gpc curves of samples MBC-2 (d, open symbol) and sMBC-2 (e, solid symbol) together with gpc curves of the starting oligomers (a, oligo-PES (dashed-dotted line); b,
oligo-DP-PES (dotted line); c, end-capped oligo-DP-PES (solid line)).
block copolymers within the error margin. Only the high IEC sam-
ples showed a much higher swelling (sMBC-1) or even dissolve in
water at 80 °C (sRC-1).
appearance of two Tgs in the dsc curves is not expected due to
the quite similar chemical composition of the two building blocks
of the polymer backbone.
The swelling in z-direction (perpendicular to the surface) shows
the same trend as the swelling in x–y-direction: the higher the IEC,
the higher the swelling. However, due to the smaller dimensions
Interestingly no Tg could be detected for the two sulfonated
multiblock copolymers sMBC-1 and sMBC-2 in the temperature
range up to 300 °C. Only the sulfonated multiblock copolymer with
the lowest IEC (sMBC-3) showed a single Tg at 255 °C. It is sup-
posed, that the high number of sulfonic acid groups in the samples
sMBC-1 and sMBC-2 leads to such a high number of hydrogen
bonds between the polymer chains, that the Tg is shifted to temper-
atures close to or even above 300 °C. Further studies on the ther-
mal properties of these materials will be performed and reported
in a forthcoming publication.
The thermal properties of the sulfonated materials change
dramatically in the hydrated state. The Tg of the high IEC ran-
dom copolymers (sRC-1and sRC-2) drops to ca. 110 °C at high
humidification levels and is thus in the same range as observed
for NafionÒ under fully hydrated conditions [38]. It is obvious
from Fig. 5, that from a certain water content on (ca. 15–
20 mol water per mol sulfonic acid group), only a minor or no
change in Tg occurs. This is attributed to the fact that the water
molecules are more or less tightly bound by hydrogen bonds to
the sulfonic acid groups hence disturbing the hydrogen bonds
between sulfonic acid groups. Thus this water acts as plasticizer
in these materials. Further adsorbed water shows no interaction
with the polymer and behaves like bulk water and has therefore
no influence on the thermal properties of the material. A more
detailed discussion on the effect of water on the thermal proper-
ties of ion-exchange materials will be published in a forthcoming
paper.
(lm instead of mm) and the determination procedure (micrometer
screw) a high deviation of the values is observed. Contrary to the
findings reported by Lee et al. [23] the random copolymers inves-
tigated in this study showed an ‘‘asymmetric’’ swelling, meaning
that the dimensional change in thickness is more pronounced than
the dimensional change in length and width. This effect is less pro-
nounced or even reversed for the multiblock copolymers. This
might be a result of the relatively short oligomers used for the
preparation of block copolymers, but this does not explain the
properties of the random copolymers. Furthermore, the choice of
solvent for the membrane preparation (NMP) might play a role,
which has a strong effect on the morphology and hence on the
membrane properties [37].
3.2. Thermal and mechanical properties
The non-sulfonated random copolymers possess a single glass
transition temperature (Tg) around 230 °C (Table 2). These values
are in good agreement with the values reported by Liu et al. [18].
On sulfonation the Tg of dry samples is shifted to temperatures
around 260 °C, due to the interactions (formation of hydrogen
bonds) between the sulfonic acid groups.
The glass transition temperatures of the oligomers used for the
preparation of the multiblock copolymers are 206 °C (4600 g/mol),
212 °C (10,600 g/mol) and 225 °C (14,500 g/mol) for the oligo-PES
blocks and 205 °C (9000 g/mol) for the oligo-DP-PES block. The
multiblock copolymers prepared from the oligomers show in the
non-sulfonated form only a single Tg, which is increasing with
increasing length of the oligo-PES block (224 °C (MBC-1); 227 °C
(MBC-2); 233 °C (MBC-3)). This result is not surprising since the
Tgs of the building blocks are quite similar. Furthermore, a phase
separation between the two different blocks and therefore the
Contrary to the perfluoroalkyl sulfonic acids, which show cold
drawing behavior, all sulfonated samples prepared in this study
showed a brittle fracture behavior on applied tensile stress in the
nominally dry state (conditioned at 25 °C, 50% r.h.). The Young’s
moduli are in the range of 1.5 to ca. 2.0 GPa (see Table 4). Perflu-
oralkylsulfonic acids like NafionÒ have a Young’s modulus of ca.
0.2 GPa under the same test conditions. On the other hand the
elongation at break is for the NafionÒ-type samples much higher

836
C. Vogel et al. / Reactive & Functional Polymers 71 (2011) 828–842
(A)
(B)
Scheme 3. Possible chain growth reaction during synthesis of multiblock copolymers: (A) linear chain growth and (B) branching.
Table 3
Composition and properties of sulfonated random copolymers (sRC) and multiblock copolymers (sMBC).
Sample
ginh (dl/g)
Mn (g/mol)
Mw (g/mol)
PD
Tg (°C)
IECtheo (mmol/g)
IECtitr. (mmol/g)
WU (%)
25 °C
50 °C
80 °C
sRC-1
sRC-2
sRC-3
sMBC-1
sMBC-2
sMBC-3
1.74
0.78
3.71
6.31
4.16
3.29
50,000
64,000
120,000
29,000
34,000
62,000
155,000
174,000
340,000
345,000
218,000
407,000
3.1
2.7
2.8
11.9
6.4
6.6
260
251
264
n.d.a
n.d.
255
2.12
1.82
1.50
2.24
1.73
1.49
2.05
1.71
1.45
1.95
1.53
1.39
69
48
22
80
51
38
97
57
26
114
65
47
dis.b
111
35
473
128
60
a
Not detected up to 300 °C.
Dissolved.
b
Table 4
Mechanical properties of sulfonated random and multiblock copolymers in nominally dry and humidified state.
Sample
IEC (mmol/g)
Dimensional change
D
L ꢁ
D
d (%)
Young’s modulus
Elongation at break
25 °C
50 °C
80 °C
Dry (GPa)
Wet (GPa)
Dry (%)
Wet (%)
a
sRC-1
sRC-2
sRC-3
sMBC-1
sMBC-2
sMBC-3
2.05
1.71
1.45
1.95
1.53
1.39
17 ꢁ 34
17 ꢁ 21
3 ꢁ 15
27 ꢁ 36
20 ꢁ 22
4 ꢁ 7
–
1.56
n.d.b
1.93
n.d.
1.98
1.82
0.40
n.d.
1.12
n.d.
0.88
1.25
5.9
n.d.
5.7
n.d.
3.3
6.0
6.1
n.d.
7.1
n.d.
6.0
10.0
33 ꢁ 36
6 ꢁ 17
14 ꢁ 22
17 ꢁ 16
8 ꢁ 20
22 ꢁ 32
18 ꢁ 15
10 ꢁ 14
88 ꢁ 73
34 ꢁ 31
13 ꢁ 25
a
Dissolved.
n.d. = not determined.
b
(>100%) than for the aromatic polymers (<10%). The plasticization
effect of water in the ion-exchange membranes is also reflected by
the mechanical properties of humidified samples. The Young’s
moduli of the random as well as multiblock copolymers drop sig-
nificantly on humidification and a clear trend is obvious. Samples
with high IEC possess a lower Young’s modulus than samples with
low IEC due to the higher water uptake. However, the Young’s
moduli of ion-exchange materials introduced in this study in the
wet state are still higher than that of NafionÒ-type materials.
Furthermore the type of respond on the tensile stress changes for
the aromatic polymers. While the dry samples show a brittle frac-
ture behavior, the wet ones showed a cold draw behavior like the
perfluoroalkyl sulfonic acids.
3.3. NMR spectroscopic characterization
The chemical composition of the various copolymers prepared
in this study was determined from 1H NMR spectra. Fig. 6 shows

C. Vogel et al. / Reactive & Functional Polymers 71 (2011) 828–842
837
the proton a in ortho-position to the sulfone groups in substructure
I, whereas the doublet at 7.75 ppm can be assigned to proton a000 of
substructure III. The nonsymmetric triad II with protons a0 and a00
results in two doublets of same intensity located between these
signals at 7.90 and 7.83 ppm. The corresponding signals b–b000
could be identified by 2D NMR (Fig. 7). The intensities of the sig-
nals assigned to the different substructures correspond very well
with the theoretical values from probability calculations, indicat-
ing the formation of real random copolymers. Overall, a very good
correlation between monomer ratio in the feed and monomer ratio
in the polymer backbone was found (Table 2). The 1H NMR spec-
trum of the multiblock copolymers (Fig. 8) is much simpler than
that of the random copolymers, because only the symmetric sub-
structures I (from the PES-block) and III (from the DP-PES-block)
of Scheme 4 can occur in this type of polymer neglecting the termi-
nal units of each block. As for the random copolymers, the compo-
sition determined by NMR spectroscopy fits very well with
composition calculated from the educts. With increasing relative
content of the still unsulfonated ‘‘hydrophilic’’ block, which is
poorly soluble in DMSO, a line broadening is observed. The 1H
NMR spectra of sulfonated random and multiblock copolymers
(Figs. 9 and 10) show the same changes, compared to the spectra
of the non-sulfonated samples. The signals d–f of the pendant phe-
nyl ring disappear and two doublets of same intensity appear at
7.46 and 7.56 ppm (Fig. 10). The signal pattern, intensities and
the down-field shift compared to signals d and e in the non-sulfo-
nated samples clearly indicate selective sulfonation of the pendant
phenyl rings in para-position. The signals of the diphenylsulfone
units remain unchanged and show the different pattern for RC
and MBC as discussed before. In the case of similar copolymers
having phenylhydroquinone units instead of 2,5-diphenyl hydro-
quinone units in the polymer backbone, the sulfonation of the pen-
280
260
240
220
200
180
160
140
120
100
80
sRC-1
sRC-2
sRC-3
Nafion®
sMBC-3
0
5
10 15 20 25 30 35 40 45 50 55
λ
(mmolH O/mmol
)
SO3H
2
Fig. 5. Glass transition temperature of humidified ion-exchange membranes as
function of hydration number. Data for Nafion were adapted from Ref.
[38].
the spectra of the non-sulfonated random copolymers together
with the signal assignment. The spectra show five well-separated
groups of signals appearing at 8.1–7.7, 7.45, 7.35, 7.3–7.15 and,
7.15–7.0 ppm. The assignment of the signals as outlined in Fig. 6
was deduced from 2D NMR spectra. As an example, the 2D spec-
trum of sample RC-3 is shown in Fig. 7. The splitting of the signals
a and b of the diphenylsulfone unit into four doublets, which is
more pronounced for the proton a, is assigned to the three different
diphenylsulfone-centered triads I–III which are formed in the poly-
mer backbone (Scheme 4). The doublet at 7.98 ppm is assigned to
f
e
d
c
a' b'
b
a
O
S
O
S
O
S
O
O
O
O
x
O
O
O
a
b
a'
d
c
a
e, f, b, b'
RC-1
RC-2
RC-3
8.2
8.0
7.8
7.6
7.4
7.2
7.0
6.8
δ 1H (ppm)
Fig. 6. 1H NMR spectra of random copoly(arylene ether sulfone)s (solvent: DMSO-d6).

838
C. Vogel et al. / Reactive & Functional Polymers 71 (2011) 828–842
C
7.0
7.1
7.2
7.3
7.4
7.5
7.6
7.7
7.8
7.9
8.0
8.00
7.75
7.50
7.25
7.00
F2 Chemical Shift (ppm)
Fig. 7. 2D-NMR spectrum (1H–1H COSY) of sample RC-3 (solvent: DMSO-d6).
b
a
a
b
O
S
O
S
O
S
Substructure I
Substructure II
O
O
O
O
O
O
O
O
O
O
b' a'
a'' b''
O
S
O
S
O
O
O
b''' a'''
a''' b'''
O
S
Substructure III
O
O
O
O
O
Scheme 4. Possible diphenylsulfone-centered triads occurring in the random copoly(arylene ether sulfone)s giving rise to the splitting of the proton signals of this unit.
dant phenyl ring as well as of the hydroquinone ring resulted in a
change of the splitting pattern of signal a [18]. Thus, within the
limits of the NMR method there is no indication that the sulfona-
tion occurs at other positions than the pendant phenyl rings. It
should be mentioned that the signal of proton c shows a splitting
in three signals due to pentad effects for the sRCs. Furthermore,
sulfonation results in an improved solubility of the hydrophilic
block and so the line broadening observed for MBCs is significantly

C. Vogel et al. / Reactive & Functional Polymers 71 (2011) 828–842
839
F
F
F
F
F
F
F
F
b''' a'''
b
a
O
S
O
S
O
O
O
O
O
O
O
O
c
n
m
d
e
f
a'''
d
c, b, f, e
b'''
a
MBC-1
MBC-2
MBC-3
8.2
8.0
7.8
7.6
7.4
7.2
7.0
6.8
δ
1H (ppm)
Fig. 8. 1H NMR spectra of multiblock copoly(arylene ether sulfone)s (solvent: DMSO-d6).
SO H
3
e
d
c
a' b'
b
a
O
S
O
S
O
S
O
O
O
O
x
O
O
O
a
b
HO S
3
a
e
d, c
b'
a'
b
sRC-1
sRC-2
sRC-3
8.2
8.0
7.8
7.6
7.4
7.2
7.0
δ
1H (ppm)
Fig. 9. 1H NMR spectra of sulfonated random copoly(arylene ether sulfone)s (solvent: DMSO-d6).

840
C. Vogel et al. / Reactive & Functional Polymers 71 (2011) 828–842
SO H
3
SO H
3
F
F
F
F
F
F
F
F
b''' a'''
b
a
O
S
O
S
O
O
O
O
O
O
O
O
c
n
m
d
e
HO S
3
HO S
3
a'''
e
d c
b
b'''
a
sMBC-1
sMBC-2
sMBC-2
8.2
8.0
7.8
7.6
7.4
7.2
7.0
6.8
δ
1H (ppm)
Fig. 10. 1H NMR spectra of sulfonated multiblock copoly(arylene ether sulfone)s (solvent: DMSO-d6).
C
7.0
7.1
7.2
7.3
7.4
7.5
7.6
7.7
7.8
7.9
8.0
8.00
7.75
7.50
7.25
7.00
F2 Chemical Shift (ppm)
Fig. 11. 2D NMR spectrum (1H–1H COSY) of sample sRC-3 (solvent: DMSO-d6).

C. Vogel et al. / Reactive & Functional Polymers 71 (2011) 828–842
841
reduced for sMBcs. The assignment of signals was again supported
by 2D NMR spectra recorded from random copolymers (Fig. 11).
Temperature (°C)
180 160 140 120 100 80
60
40
20
1000
100
10
1000
RH = 100%
pwater = 1atm
3.4. Conductivity measurements
100
10
The proton conductivities, measured in plane, of the mem-
branes were studied over a temperature range from 30 °C to
100 °C at 100% relative humidity (sample and water reservoir have
the same temperature) and above 100 °C at 1 atm water vapor
pressure, meaning that the relative humidity decreases with
increasing temperature (the relative humidity at temperatures
above 100 °C was calculated from the temperature of the water
reservoir (100 °C) and the sample temperature [39]). The proton
conductivities as function of temperature are displayed in Fig. 12
(random copolymers) and 13 (multiblock copolymers). The con-
ductivity curve of NafionÒ 115 is given in these plots as reference.
The conductivities of the random copolymers increase with
increasing ion-exchange capacity and increasing temperature.
Although the ion-exchange capacity of sRC-3 exceeds that of Naf-
ionÒ by at least 50%, the proton conductivity is much lower than
that of the reference. The samples sRC-1 and sRC-2 having IECs
of 2.05 and 1.71 mmol/g, respectively showed proton conductivi-
ties comparable to that of NafionÒ. At temperatures above 100 °C
and at reduced humidification the same trend for the conductivi-
ties is observed as for measurements at fully hydrated conditions,
the higher the IEC the higher the conductivities. However, the con-
ductivities decrease much stronger with increasing temperature
(decreasing humidity) than that of NafionÒ. The sample with the
lowest IEC (sRC-3) shows the lowest conductivity and the steepest
decrease. With increasing IEC the loss of conductivity with temper-
ature is less steep but still more pronounced than that of NafionÒ.
Although having comparable ion-exchange capacities, the multi-
block copolymers showed a completely different behavior than the
random copolymers. Even the samples with the lowest IEC (sMBC-
2, sMBC-3) possess at 100% r.h. a proton conductivity similar to Naf-
ionÒ, while the sample with highest IEC (sMBC-1) exceeds that of
NafionÒ by ca. 100% in the temperature range up to 100 °C under
fully hydrated conditions. As already observed for the random
copolymers, the loss of conductivity is much more severe compared
to NafionÒ. However, the inflection point is shifted to temperatures
above 100 °C. As an example, sMBC-1 (IEC: 1.95 mmol/g) shows a
higher conductivity than NafionÒ up to ca. 120 °C.
1
1
Sample
IEC (mmol/g)
®
Nafion117 0.88
0.1
0.1
sMBC-1
sMBC-2
sMBC-3
1.95
1.53
1,39
0.01
0.01
2.21 2.31 2.42 2.54 2.68 2.83 3.00 3.19 3.41
1000/T (K-1)
Fig. 13. Proton conductivities of sulfonated multiblock copolymers as function of
temperature and relative humidity (at T > 100 °C).
The differences in proton conductivities can be explained when
taking the water uptake (Table 3) of the samples into account. The
hydration numbers k of the random copolymers sRC-1–sRC-3 are
57.6, 31.2 and 16.5, respectively. In the same order as the k-value
increases an increase in the proton conductivity of these mem-
brane materials is observed. Due to the higher water uptake (at
20 °C) the hydration numbers of the multiblock copolymers are
substantially higher than those of the random copolymers at com-
parable ion-exchange capacities (Fig. 14), which leads to the ob-
served higher conductivities at fully hydrated conditions. The
loss of conductivity with increasing temperature might be a result
of a faster desorption of the water compared to NafionÒ, due to the
lower acid strength of aromatic sulfonic acids. A second reason for
the higher water uptake on one hand and the faster desorption of
water on the other hand may be due to differences in the morphol-
ogy of hydrocarbon-based membranes and perfluoroalkyl sulfonic
acids-based like NafionÒ. The latter shows a strong phase sepa-
rated morphology between the highly hydrophobic backbone and
the hydrophilic side chains. The differences in hydrophilic and
hydrophobic properties are not that pronounced in hydrocarbon-
based membranes, resulting in a less-ordered (phase-separated)
morphology, even in the multiblock copolymers.
Temperature (°C)
180 160 140 120 100 80
60
40
20
120
multiblock copolymers
1000
1000
100
10
random copolymers
RH = 100%
®
pwater = 1 atm
Nafion 117
90
60
30
0
100
10
1
1
Sample
IEC (mmol/g)
®
Nafion117 0.88
0.1
0.1
0.01
sRC-1
sRC-2
sRC-3
2.05
1.71
1.45
0.01
0.5
1.0
1.5
2.0
2.5
2.21 2.31 2.42 2.54 2.68 2.83 3.00 3.19 3.41
1000/T (
K-1
)
ion-exchange capacity (mmol SO3H/g polymer)
Fig. 12. Proton conductivities of sulfonated random copolymers as function of
temperature and relative humidity (at T > 100 °C).
Fig. 14. Hydration number k as function of ion-exchange capacity after storage in
water at 20 °C.

842
C. Vogel et al. / Reactive & Functional Polymers 71 (2011) 828–842
[7] M. Schuster, K. Kreuer, H.T. Andersen, J. Maier, Macromolecules 40 (2007) 598–
4. Conclusions
607.
[8] M. Schuster, C.C. de Araujo, V. Atanasov, H.T. Andersen, K. Kreuer, J. Maier,
Macromolecules 42 (2009) 3129–3137.
Random as well as multiblock copolymers were successfully
prepared from 4,40-difluorodiphenyl sulfone, bis-(4-hydroxy phe-
nyl) sulfone and 2,5-diphenyl hydroquinone by nucleophilic dis-
placement polycondensation reaction using the silyl-method.
While the random copolymers had a linear structure with low
PD values, it turned out that during preparation of the multiblock
copolymers branching reactions occurred at the decafluoro biphe-
nyl unit. Nonetheless the multiblock copolymers were still soluble
in dipolar aprotic solvents. Contrary to the findings with monoph-
enylated polymers (using phenyl hydroquinone instead of 2,5-di-
phenyl hydroquinone) reported earlier by our group, a selective
sulfonation of the side-chain was achieved with conc. sulfuric acid.
Membranes prepared from both types of polymers were transpar-
ent and flexible. Although having comparable ion-exchange capac-
ities, the sulfonated multiblock copolymers showed higher proton
conductivities than the sulfonated random copolymers and at high
IEC (P1.95 mmol/g) even higher than NafionÒ at 100% relative
humidity, which was attributed to the relatively high water uptake
of the membranes prepared from the multiblock copolymers. On
the other hand, membranes prepared from random copolymers
with an IEC P1.8 mmol/g tend to soften to a high extend or even
dissolve in water at elevated temperatures.
[9] C.C. de Araujo, K.D. Kreuer, M. Schuster, G. Portale, H. Mendil-Jakani, G. Gebel, J.
Maier, Phys. Chem. Chem. Phys. 11 (2009) 3305–3312.
[10] C. Iojoiu, M. Marechal, F. Chabert, J. Sanchez, Fuel Cells 5 (2005) 344–354.
[11] J.F. Blanco, Q.T. Nguyen, P. Schaetzel, J. Appl. Polym. Sci. 84 (2002) 2461–
2473.
[12] B. Lafitte, L.E. Karlsson, P. Jannasch, Macromol. Rapid Commun. 23 (2002) 896–
900.
[13] L.E. Karlsson, P. Jannasch, Electrochim. Acta 50 (2005) 1939–1946.
[14] P. Jannasch, B. Lafitte, Adv. Funct. Mater. 17 (2007) 2823–2834.
[15] D.S. Kim, G.P. Robertson, Y.S. Kim, M.D. Guiver, Macromolecules 42 (2009)
957–963.
[16] B. Liu, G.P. Robertson, D. Kim, M.D. Guiver, W. Hu, Z. Jiang, Macromolecules 40
(2007) 1934–1944.
[17] J. Meier-Haack, H. Komber, C. Vogel, W. Butwilowski, K. Schlenstedt, D.
Lehmann, Macromol. Symp. 254 (2007) 322–328.
[18] B. Liu, G.P. Robertson, D. Kim, X. Sun, Z. Jiang, M.D. Guiver, Polymer 51 (2010)
403–413.
[19] S. Tian, Y. Meng, A.S. Hay, J. Polym. Sci.: Part A: Polym. Chem. 47 (2009) 4762–
4773.
[20] S. Tian, Y. Meng, A.S. Hay, Macromolecules 42 (2009) 1153–1160.
[21] C. Vogel, J. Meier-Haack, A. Taeger, D. Lehmann, Fuel Cells 4 (2004) 320–327.
[22] J. Yu, B. Yi, D. Xing, F. Liu, Z. Shao, Y. Fu, H. Zhang, Phys. Chem. Chem. Phys. 5
(2003) 611–615.
[23] H. Lee, A. Roy, O. Lane, M. Lee, J.E. McGrath, J. Polym. Sci.: Part A: Polym. Chem.
48 (2010) 214–222.
[24] H. Lee, A. Roy, O. Lane, S. Dunn, J.E. McGrath, Polymer 49 (2008) 715–723.
[25] A. Roy, M.A. Hickner, X. Yu, Y. Li, T.E. Glass, J.E. McGrath, J. Polym. Sci. Part B:
Polym. Phys. 44 (2006) 2226–2239.
Due to their high proton-conductivity and good dimensional
stability at elevated temperatures sulfonated multiblock copoly-
mers turned out to be promising candidates for membrane materi-
als in PEM fuel cells operated at temperatures up to 120 °C.
[26] M.V. Fedkin, X. Zhou, M.A. Hofmann, E. Chalkova, J.A. Weston, H.R. Allcock, S.
Lvov, Mater. Lett. 52 (2002) 192–196.
[27] P.R. Shildneck, R. Adams, J. Am. Chem. Soc. 53 (1931) 2373–2379.
[28] L. Birkofer, O. Stuhl, J. Organomet. Chem. 177 (1979) C16–C18.
[29] H. Deckers, Verfahren zur Herstellung von aromatischen Polyäthern, DE
3342433, 1985.
[30] C.A. Martinez, A.S. Hay, J. Polym. Sci.: Part A: Polym. Chem. 35 (1997) 1781–
1798.
Acknowledgment
[31] C.A. Martinez, A.S. Hay, J. Polym. Sci.: Part A: Polym. Chem. 35 (1997) 2015–
2033.
[32] J. Ding, M. Day, G.P. Robertson, J. Roovers, Macromol. Chem. Phys. 205 (2004)
1070–1079.
The authors like to thank Mrs. Harnisch and Mrs. Treppe for
conducting the gpc measurements.
[33] Y. Yang, Z. Shi, S. Holdcroft, Macromolecules 37 (2004) 1678–1681.
[34] M. Ueda, H. Toyota, T. Ouchi, J. Sugiyama, K. Yonetake, T. Masuko, T. Teramoto,
J. Polym. Sci.: Part A: Polym. Chem. 31 (1993) 853–858.
[35] F. Wang, M. Hickner, Y.S. Kim, T.A. Zawodzinski, J.E. McGrath, J. Membr. Sci.
197 (2002) 231–242.
[36] F. Wang, M. Hickner, Q. Ji, W. Harrison, J. Mecham, T.A. Zawodzinski, J.E.
McGrath, Macromol. Symp. 175 (2001) 387–395.
[37] M. Lee, J.K. Park, H. Lee, O. Lane, R.B. Moore, J.E. McGrath, D.G. Baird, Polymer
50 (2009) 6129–6138.
[38] Y.S. Kim, L. Dong, M.A. Hickner, T.E. Glass, V. Webb, J.E. McGrath,
Macromolecules 36 (2003) 6281–6285.
[39] C. Synowietz, K. Schäfer (Eds.), Chemiker Kalender, third ed., Springer Verlag,
Berlin, Heidelberg, New York, Tokyo, 1984, p. 592.
References
[1] C. Wieser, Fuel Cells 4 (2004) 245–250.
[2] M. Cakir, S. Karatas, Y. Menceloglu, N. Kayaman-Apohan, A. Güngör, Macromol.
Chem. Phys. 209 (2008) 919–929.
[3] H. Bi, S. Chen, X. Chen, K. Chen, N. Endo, M. Higa, K. Okamoto, L. Wang,
Macromol. Rapid Commun. 30 (2009) 1852–1856.
[4] N. Li, J. Liu, Z. Cui, S. Zhang, W. Xing, Polymer 50 (2009) 4505–4511.
[5] Z. Hu, Y. Yin, K. Okamoto, Y. Moriyama, A. Morikawa, J. Membr. Sci. 329 (2009)
146–152.
[6] G. Maier, J. Meier-Haack, Adv. Polym. Sci. 216 (2008) 1–62.