Note: Descriptions are shown in the official language in which they were submitted.
CA 02292703 1999-12-20
1
TITLE OF THE INVENTION
COMPOSITE ELECTROLYTE MEMBRANES FOR FUEL CELLS AND
METHODS OF MAKING SAME
FIELD OF THE INVENTION
The present invention relates to composite electrolyte
membranes for fuel cells and methods of making same. More
specifically, the present invention is directed to proton-conducting
membranes for fuel cell applications. The present invention further
describes materials which reach high intrinsic proton conductivity, and are
suitable for use as electrolytic membranes in methanol fuel cells.
BACKGROUND OF THE INVENTION
The need for pollution control stimulated the
development of polymer electrolyte membrane fuel cells (PEMFC) and
attracted an increasing interest particularly for the automotive and
stationary power applications [1-3]. For example, Daimler Benz presented
in 1998 a fuel cell powered car, NECAR II with a total electric power close
to 50kW. A fuel cell is an almost ideal energy source yielding a very high
thermal efficiency and an essentially zero release of atmospheric
pollutants. In transport applications, the direct methanol fuel cell (DMFC)
is presently considered as the most appropriate and promising. Up to
now, only perfluorinated ionomers (PFI) membranes were considered to
meet the requirements of polymer electrolyte membrane (PEM) fuel cells,
namely, a high proton conductivity, a high stability in the cell operating
conditions and a high durability. Presently the commercial solid polymer
#11229.117/DEM.WPD
CA 02292703 1999-12-20
2
electrolyte material used in PEMFC is either perfluorinated Nafion (Du
Pont) or Nafion-like polymers [4] supplied by Dow, Asahi Glass (Flemion)
and Asahi Chemicals (Acipex). Unfortunately, these PEMFC limit large
scale application due to a number of drawbacks. First of all, these
ionomers are very expensive. For example, the manufacturer's price for
the NAFION membranes (Dupont de Nemours) scale exceeds
600US$/m2. Other membranes of this kind (DOW, RAI, ...) are still more
expensive (up to 2000 US$/m2). In fact, such membranes have been
used for a long time in H2 fuel cells for application where cost was not a
main criterion (e.g. spacecraft, submarines etc). In addition, a significant
drawback of these materials is their high permeability to methanol which
allows an easy transport of this fuel from the anode to the cathode. This
phenomon, also called methanol crossover, reduces significantly the cell
performance and must be diminished if not eliminated before DMFC can
be commercialized.
Currently the necessity to reduce the cost of PEM
stimulates the development of new proton conducting polymers. New
studies are also undertaken in order to rationalize the most efficient
combination of properties of the perfluorinated ionomer (PFI) polymer,
which make them efficient proton conductors, and develop new polymers
with similar properties by a less expensive chemistry. As a result PFI
Nafion has been extensively studied and tested in low temperature fuel
cell systems [5]. In this context, Ballard Advanced Materials' group for the
development of PEM membranes [6] recently developed a membrane
based on a trifluorostyrene monomer called BAM3G (Ballard Advanced
#11229.117/DEM.WPD
CA 02292703 1999-12-20
3
Material 3rd generation), which has demonstrated excellent performance
and longevity of several thousand hours of operation.
The remarkable properties of PFI polymers lie in the
combination of the high hydrophobicity of the fluorinated polymer
backbone and high hydrophilicity of the sulfonic acid branches. The
hydration of the PFI membrane is crucial for the performance of PEMFC
since proton conductivity decreases drastically with dehydration. For
instance, with Nafion, which loses water above 80°C, the conductivity
drops to very low values above this temperature.
One more limitation associated with Nafion type PFI
membranes [2,4) is the methanol cross-over when used in the direct
methanol fuel cell (DMFC). This results in a decreased fuel cell
performance due to depolarization of the oxygen reducing cathode. A
further drawback of the perfluorinated polymers is that they are not
environmentally friendly, a criteria that will be important when fuel cells
become mass-produced.
The above mentioned disadvantages of PFI membranes
induced many efforts to synthesize PEM based on hydrocarbon-type
polymers and brought about the emergence of partially fluorinated and
fluorine free ionomer membranes as alternatives to Nafion. Among them
the membranes based on aromatic polyether ether ketone (PEEK) were
shown to be of promise for fuel cell application [7-9], as they possess
good mechanical properties, thermal stability, toughness and some
conductivity, depending on sulphonation degree. Nevertheless, the proton
conductivity of PEEK or SPEEK has yet to reach a level sufficient to
enable an adequate performance in a fuel cell.
#11229.117JDEM. WPD
CA 02292703 1999-12-20
4
Efforts have thus been made to improve the proton
conductivity to composite membranes. For example, the addition of
solids such as zeolites or tin-mordenite was aimed at improving the
performance of the composite membranes. Unfortunately, their presence
in the membranes does not impart thereto a high enough proton
conductivity to make them useful as a solid electrolyte in polymer
electrolyte membrane fuel cells (Mikhailenko et al., 1997, Microporous
Mat. 11:37-44).
"Polymer Material for Electrolytic Membranes in Fuel
Cells" by Yen et al., U.S. 5,795,496 is one such example of SPEEK with
the aim of using it in fuel cells. The materials described in Yen et al. have
low methanol permeability but high proton conductivity, and made from
inexpensive, readily available materials. According to that invention,
proton conducting membranes are formed based on a sulfonic acid-
containing polymer. One preferred material is PEEK or PES. This
invention is said to overcome disadvantages associated with the high cost
of NAFION and with its methanol permeability problems which allows for
a substantial amount of fuel crossover across the membrane by using
materials which were inexpensive starting materials and which enhanced
protection against fuel crossover. In a particular embodiment, PEEK was
sulfonated with H2S04 to give H-SPEEK, a polymer which is soluble in an
organic solvent and water mixture. While the inventors found that sulfonic
acid increases the proton conducting performance of PEEK (the sulfonate
groups are responsible for proton conductivity), it degrades the physical
structure of the resulting membrane. Hence, the inventors developed a
trade-off between the amount of sulfonation and appropriate physical
#11229.117/DEM.WPD
CA 02292703 1999-12-20
structure by sulfonating one out of every three benzene rings. PES was
treated in an analogous manner. Yen et al. also teach methods of
modifying the morphology of the processed polymers to limit the transport
of methanol across the membrane (to reduce the free volume) by using
5 zeolites tin morderite or the like. Unfortunately, such solids do not impart
high enough proton conductivity to the composite membrane. However,
Yen et al. do not teach a composite electrolyte membrane which reaches
a high enough proton conductivity to be useful in PEM fuel cells.
During the last two decades solid electrolytes have
attracted substantial attention owing to both their great potential in several
electrochemical technologies, such as fuel cells, batteries and sensors,
and the academic interest in the phenomenon of fast ionic mobility in
solids. In spite of the fact, that a vast number of various proton
conductors have already been identified, the development of chemically
stable superionic conductors still remains one of the prime objectives
among the current directions of research in solid state electrochemistry
and materials science. Currently considerable efforts are being devoted
to proton conducting salts of oxo acids, including various hydrated and
anhydrous phosphates.
Heteropolyacids (HPAs) are known as the most
conductive solids among the inorganic solid electrolytes. Nevertheless,
the use of HPAs or other solid inorganic acids in polymers and their effect
on membranes for fuel cells, for example, has not been significantly
' addressed to show that they could enable the production of an effective,
high proton-conductive membrane for fuel cells. For example,
heteropolyacid/polyethersulfone membranes are described in Park et al.,
#11229.117/DEM.WPD
CA 02292703 1999-12-20
6
1996, Denki Kagaku 64:743-747. However, these membranes are more
than likely impossible to use commercially due to their lack of stability
under conditions of fuel cell use.
Boron phosphate (BP04), a compound commonly
synthesized from boric and phosphoric acids, has been widely used over
the last three decades as an acidic catalyst in a number of reactions,
including particularly dehydration reactions.
BP04 belongs to the class of orthophosphates in which
both P5+ and B3+ are tetrahedrally coordinated by oxygen.
Although the foregoing suggests that BP04 in the
presence of adsorbed water can possess interesting electrical properties,
there were no attempts to look upon boron phosphate as being a solid
proton conductor until the paper of Mikhailenko et al:, 1998 (J. Chem.
Soc., Faraday Trans. 94:1613-1618) which investigated the electrical
properties of BP04. It is taught therein that the conductivity of thermally
treated boron phosphate is of only one order of magnitude inferior to that
of such prominent solid electrolytes as hydrated heteropolyacids. At the
same time, the chemically durability of BP04 is far higher than that of
HPAs, which are 100% soluble in water. Therefore, boron phosphate can
be regarded as having some potential in electrochemical applications,
such as fuel cells, hydrogen gas sensors and humidity sensors.
There thus remains a need to provide composite
electrolyte membranes, which retain a proton conductivity which is high
enough to make it useful in polymer material for electrolytic membrane in
fuel cells PEMFC. There also remains a need to provide such composite
membranes which provide a proton conductivity which is comparable to
#11229.117/DEM.WPD
CA 02292703 1999-12-20
7
that of NafionTM and in addition, overcome disadvantages in the
membranes of the prior art such as pollution, low thermal efficiency and
substantial costs.
The present invention seeks to meet these and other
needs.
The present description refers to a number of
documents, the content of which is herein incorporated by reference in
their entirety.
SUMMARY OF THE INVENTION
The invention concerns a composite electrolyte
membrane comprising a filler material which contributes to the
enhancement of the protonic conductivity of the membrane. In a
particular embodiment of the present invention the composite electrolyte
membrane possesses a high enough proton conductivity to be an
alternative energy source for stationary and mobile applications. In a
particularly preferred embodiment of the present invention, the composite
electrolyte membrane possesses a high enough proton conductivity to be
used in methanol fuel cells.
The present invention further relates to composite
electrolyte membranes comprising a polymer matrix and a filler material
which contributes to the enhancement of the protonic conductivity of the
membrane. In a particular embodiment the filler material contributes
significantly more than the polymer matrix to the protonic conductivity of
the membrane. In another embodiment, both the polymer matrix and filler
material contribute significantly to protonic conductivity of the membrane.
#11229.117/DEM.WPD
CA 02292703 1999-12-20
8
In a preferred embodiment of the present invention, the filler material is
an inorganic solid acid. In a particularly preferred embodiment of the
present invention, the polymer matrix of the composite electrolyte
membrane is modified PEEK and the filler material is BP04. In a more
particularly preferred embodiment, the modified PEEK is sulfonated
PEEK (SPEEK).
The invention in addition relates to an environmental-
friendly composite electrolyte membrane comprising an environmental-
friendly polymer matrix and an environmental-friendly filler material which
contributes to the enhancement of the protonic conductivity of the
membrane. In a preferred embodiment, the environmental-friendly
polymer matrix is a modified PEEK and the environmental-friendly filler
is an inorganic solid acid. In a particularly preferred embodiment of the
present invention, the the environmental-friendly polymer matrix of the
composite electrolyte membrane is SPEEK and the environmental-
friendly filler material is BP04.
The invention further relates to composite electrolyte
membranes for fuel cells, which remains efficient at a temperature above
about 80°C, preferably above about 90°C, most preferably about
100°C,
and even more preferably between about 100°C and 120°C.
In addition, the invention relates to fuel cell membranes
comprising a solid electrolyte embedded in a polymer matrix, wherein the
solid electrolyte contributes to the protonic conductivity of the membrane.
The invention also relates to methods for preparing and
pretreating an inorganic solid acid, so as to embed it in a polymer matrix
for fuel cells.
#11229.1171DEM.WPD
CA 02292703 1999-12-20
9
In addition, the invention relates to methods for
embedding a solid inorganic acid in a polymer matrix.
Before the present invention, solid fillers in composite
membranes were not contributing sufFiciently to the protonic conductivity
of the membrane.
In accordance with the present invention, there is
therefore provided a method of increasing the proton conductivity of a
composite electrolyte membrane comprising an acidic polymer matrix; the
method comprising an imbedding of a proton conductivity effective
amount of an inorganic solid acid in the matrix.
In accordance with another aspect of the present
invention, there is provided a method of increasing a proton conductivity
and/or a stability of a proton conductivity at a temperature above about
80oC, in a polymer electrolyte membrane for fuel cell, comprising an
embedding into the polymer of a proton conductivity effective amount of
an inorganic solid acid.
While the composite electrolyte membranes of the
instant invention have been demonstrated in particular with SPEEK in
which BP04 or heteropolyacids (HPAs) have been imbedded, the
invention should not be so limited. Indeed, the present invention relates
to composite electrolyte membranes in which a polymer matrix is
embeded with a solid inorganic acid. Non-limiting examples of inorganic
acids which could be used in accordance with the present invention
i
include HPAs (also examplifled hereinbelow), including tungsophoric acid
(Mikhailenko et al., 1997, Solid State tonics 99: 281-286),
molybdophosphoric acid, and molybdosilicic acid; zirconium or titanium
#11229.117IDEM.WPD
CA 02292703 1999-12-20
oxophosphates and sulphates; grafted silica materials such as for
example ORMOSIL and ORMOCER ceramics, and mixtures of these
inorganic acids. Broadly, the skilled artisan, to which the present
invention pertains, will understand that any polymer matrix, and especially
5 any sulfonated and/or phosphorylated polymer matrix having the requisite
structural characteristic enabling the formation of a membrane having the
requisite protonic conductivity, following the teachings of the present
invention, can be used.
The skilled artisan to which the present invention
pertains will understand however, that the inorganic solid acid should be
chosen amongst the acids having a sufficient acidic strength to provide
the necessary protonic conductivity to the polymer and membrane. The
contribution of the inorganic solid acid to the total weight of the membrane
should be below about 80% and preferably below 70%.
The skilled artisan should also comprehend that the
present invention should not be limited to SPEEK. Indeed, other polymer
matrices can be used in accordance with the present invention. Non-
limiting examples thereof include polysulfones, polystyrenes, polyether
imides, polyphenylenes, poly alpha olefins, polycarbonates and mixtures
thereof. Similarly, the modification of these polymer matrices is not
limited to sulfonation since, for example, phosphorylation could also be
used.
In addition, it is herein demonstrated that an inorganic
solid acid, such as BP04 can be modified to enhance is protonic
conductivity capacity and/or the stability thereof. The possibility to
enhance the BP04 inertness towards water was assessed by the
#11229.117/DEM.WPD
CA 02292703 1999-12-20
11
introduction of aluminum as a stabilizing component. The resulting
AI-BP04 compound was used to monitor its conductivity by impedance
spectroscopy as a function of AI loading, calcination temperature and
water content. It was discovered that the partial replacement of boron
with aluminum in boron phosphate brings about an increase in inertness
of the solid towards water. In some cases it is achieved at the expense
of a decrease in its conductivity. However, samples with AI/B (5/95
calcined at T<600°C where found to possess the same conductivity as
pure BP04, in spite of their lesser solubility compared to boron
phosphate. This observation reinforces the view that a compromise
between high proton conductivity and high stability of boron phosphate is
feasible, and can be achieved, among other approaches, through
chemical modification of the imbedded solid.
In accordance with the present invention, it will be
understood that further modifications can be made to the composite
membrane. as will be seen below with PEI, the incorporation of a basic
compound into the acidic polymer matrix can improve its conductivity
and/or stability. Non-limiting examples of such basic compounds which
can be incorporated into the composite electrolyte membranes, in
accordance with the present invention include polyetherimide (PEI) and
polyoxadiazole. PEI is a porous solid which can retain water and release
it upon drying of the membrane, thereby maintaining a more efficient
conductivity of the membrane with time (especially at higher
tempreatures, which tend to accelerate the drying and hence the
diminishing of the conductivity of the membrane).
#11229.1171DEM.WPD
CA 02292703 1999-12-20
12
In addition, while solid fillers such as a zeolite material,
(e.g. mordenite; see for example USP 5,795,496), or silicium material do
not contribute to the protonic conductivity of the membrane, it should be
clear to the skilled artisan, that the composite electrolyte membranes of
the present invention could benefit from an addition of such fillers which
can reduce methanol crossover (the filler serving as a seive through
which the methanol molecule cannot pass) and/or stabilize the membrane
(e.g. by retaining water), thereby enhancing the preservation of the proton
conductivity of the membranes. Numerous methods to incorporate fillers
into the polymer matrix are known in the art (e.g. WO 96/29752, of Grot
et al., published September 26, 1996).
Always keeping in mind the need to maintain a structural
and chemical stability for the resulting composite membrane, such fillers
(basic fillers such as PEI or non-basic fillers) can be added to an amount
of about 10% or less, and preferably to about 5% or less, based on the
total weight of the membrane.
Properties which are preferred or required for direct
methanol fuel cells (PMFC) are as follows: a high proton conductivity of
at least about 5 x 10-Z S/cm in order to avoid Ohmic losses; a good
mechanical resistance of films of 100 Nm thickness; a low permeation of
reactants and products of the electrochemical combustion; a high
chemical and electrochemical stability in the cell operating conditions; and
a cost compatible with commercial requirements.
The terminology "possessing high enough protein
conductivity", "maintaining a high enough proton conductivity" and the
like, should be understood by the person skilled in the art, to be
#11229.117/DEM.WPD
CA 02292703 1999-12-20
13
dependent on the context of use of the composite electrolyte membranes
in accordance with the present invention. For example, when using such
membranes in a fuel cell, the conductivity of the composite membrane is
preferably above about 5 10-2 S/cm. More preferably, above about 2,5
10-2 S/cm and even more preferably equal to or above 10-' S/cm. Since
the conductivity of Nafion membranes often ranges between 2 10-2 and
2 10-2 S/cm. When referring to a conductivity which is comparable to that
of Nafion, the conductivity for the fuel cell composite membrane refers to
a conductivity which is in that range or better.
It should be understood by the skilled artisan that since
the structural and chemical stability of the membrane is critical to ensure
an efficient and stable conductivity of the membrane, that the actual
contribution of different materials in the composite membranes based on
the total weight thereof needs to be taken into account.
BRIEF DESCRIPTION OF THE DRAWINGS
Having thus generally described the invention, reference
will now be made to the accompanying drawings, showing by way of
illustration a preferred embodiment thereof, and in which:
Figure 1 a shows the ion exchange capacity of SPEEK
polymers as a function of time of sulfonation. Figure 1 b shows the
reaction of sulfonation of PEEK;
Figures 2a, 2b, 2c and 2d show the conductivity of
PEEK as a function of their degree of sulfation (D.S.) and as a function
of the embedding of BP04 thereinto. Figure 2e shows the NMR spectra
of differently sulfonated PEEK samples dissolved in DMSO-d6;
#11229.117/DEM.WPD
CA 02292703 1999-12-20
14
Figure 3 shows the calculated degree of sulfonation
based on the NMR spectra of SPEEK samples;
Figure 4 shows the conductivity of PEEK as a function
of its sulfonation;
Figure 5 shows the glass transition temperature of PEEK
as a function of its degree of sulfonation;
Figure 6 shows the thermal stability of sulfonated PEEK
(Fig. 6b) and pure PEEK (Fig. 6a);
Figure 7 shows the first weight loss peek of SPEEK as
a function of the degree of sulfonation;
Figure 8 shows the temperature onset of the first and
second weight losses of SPEEK;
Figure 9 shows the hydrophilicity of SPEEK as a
function of sulfonation;
Figure 10 shows a micrograph of the cryogenic fracture
of a composite SPEEK/HPA membrane;
Figure 11 shows the effect of the degree of sulfonation
on the conductivity of SPEEK membranes at room temperature;
Figure 12 shows the effect of temperature on the
conductivity in a series of SPEEK membranes having different degrees
of sulfonation;
Figure 13 shows the effect of solid HPA incorporation
and temperature on the conductivity of a SPEEK/HPA composite
membrane having a 70% degree of sulfonation;
#11229.1171DEM.WPD
CA 02292703 1999-12-20
Figure 14 shows the effect of solid HPA incorporation
and temperature on the conductivity of a SPEEK/HPA composite
membrane having a 74% degree of sulfonation;
Figure 15 shows the effect of solid HPA incorporation
5 and temperature on the conductivity of a SPEEK/HPA composite
membrane having an 80% degree of sulfonation;
Figure 16 shows the Arrhenius plots of conductivity for
the 70% SPEEK-based composite membranes-embedded HPA;
Figures 17a and 17b show the apparent solubility of
10 aluminum boron phosphate in water (1 wt% slurry stirred for 15 h at
ambient temperature) as a function of the calcination temperature;
Figure 18 shows the conductivity of different samples of
aluminum boron phosphate at ambient temperature as a function of water
content. The solids were calcinated at : (a) 400°C and (b)
1000°C;
15 Figure 19 shows the conductivity at ambient temperature
of different samples of aluminum boron phosphate containing ca 40% of
water as a function of AI/B ratio;
Figure 20 shows the conductivity of a SPEEK/PEI
composite membrane as a function of temperature; and
Figure 21 shows the conductivity of a composite
membrane of SPEEK/PEI doped with H3P04 and HCL as a function of
temperature.
Other objects, advantages and features of the present
invention will become more apparent upon reading of the following
non-restrictive description of preferred embodiments with reference to the
#11229.1171DEM.WPD
CA 02292703 1999-12-20
16
accompanying drawing which is exemplary and should not be interpreted
as limiting the scope of the present invention.
DESCRIPTION OF THE PREFERRED EMBODIMENT
The approach that was used to develop the membranes
of the present invention is based on the general principle of composite
materials, which is to combine the properties of each of at least two
components to reach a desired set of properties for the composite
material. Most of the known proton conductors are not appropriate for the
fabrication of membranes, being too fragile and of little mechanical
resistance as a film. Electrolyte membranes can however be prepared
with a polymer matrix. Professor E. Skou of the University of Odense
(Denmark) first proposed the use of a composite membrane in fuel cells.
The studies of his group were however essentially oriented toward the
use of zeolites as proton conductors/filler material. The best ones thev
identified were tin modified mordenites. These studies were stopped in
the mid nineties without reaching their objectives, mainly because the
membranes never reached a high enough proton conductivity to be
effectie for fuel cell use.
A systematic study of the electrical properties of different
filler materials led to the present invention: an electrolyte composite
membrane comprising an inorganic solid acid which significantly
contributes to the proton conductivity of the membrane. In accordance
with one embodiment of the present invention, the composite membrane
comprises BP04 as the solid inorganic acid. In the presence of water
wafers of BP04, the composite membranes have a conductivity higher
#11229.117/DEM.WPD
CA 02292703 1999-12-20
17
than 10-2 S/cm and depending on the conditions of its preparation it can
reach a high stability in aqueous media. The electrical properties of BP04
are influenced by the preparation and pretreatment conditions. In
accordance with a preferred embodiment of the proton conducting
membranes of the present invention, powdered BP04 is used and
synthesized at 120°C, with a particle size of 60 mesh and calcined at
400-500°C.
As an example of a polymer matrix which can be used
in the composite membrane in accordance with the present invention,
poly-aryl ether ether ketone (PEEK), a rigid and thermally stable
thermoplastic, was used. Its formula is as follows:
O
O O C
PEEK has a hydrophobic character and does not allow
its use as solid electrolyte in the presence of water. In order to make
PEEK more water compatible and to give it proton conduction properties,
PEEK was sulfonated in concentrated sulfuric acid (Figure 1 b). As for
essentially every polymer, when the level of sulfonation is raised, highly
sulfonated SPEEI< become for example partly soluble in methanol. The
level of sulfonation of SPEEK should thus preferably not exceed 70% so
that it retains its chemical stability. Even partially sulfonated SPEEK
reaches a rather high proton conductivity. Figure 1 a shows the variation
#11229.117/DEM.WPD
CA 02292703 1999-12-20
18
in SPEEK conductivity with increased sulfonation. It can be seen that the
conductivity is continuously increasing with time. Of note, PEEK
dissolved in H2S04 does not undergo molecular weight degradation even
for long periods of time. Also, in the absence of solid inorganic acid,
SPEEK samples have a rather high conductivity which depends strongly
on the duration of the sulfonation process. In particular, the sample
treated or 112 h showed a conductivity of the order to 3 x 10-2 S/cm which
remains reversible even once the sample has been heated to 120°C. For
the other samples (sulfonated for a shorter time) this conductivity is not
as stable and it drops irreversibly after heating to 100-120°C.
The composite PEM of the present invention, comprising
an inorganic solid acid and particularly BP04/SPEEK membranes are very
promising for the fabrication of fuel cells. These membranes may indeed
be utilized as an electrolyte separating the anode from the cathode and
they may be used over a temperature range of 100-120°C. Usually PEM
fuel cells work at temperatures below 80°C (i.e. those based on
Nafion).
The membranes of the present invention being stable at higher
temperature should thus be very stable in the working conditions of the
cell.
Figures 2a and 2b show the conductivity of PEEK,
sulfonated up to different degrees of sulfonation (D.S.) as a function of
temperature. From these figures, it is clear that for D.S. of 40%, the
introduction of a solid inorganic acid such as BP04 significantly increases
the conductivity of the resulting composite electrolyte membrane by ten
to two thousand fold, depending on the temperature, as compared to pure
SPEEK which only exhibits a poor conductivity below 10-5 S/cm.
#11229.117/DEM.WPD
CA 02292703 1999-12-20
19
Similarly, embedding of BP04 into SPEEK of 65 and 72% also proves the
proton conductivity of the resulting composite membrane as compared to
pure polymer. SPEEK sulfonated at 80% possesses a very high
conductivity. In such a case, conductivity of the composite membrane
only moderately increases by embedding an inorganic solid acid such as
BP04 and with a very high concentration thereof (70%). Furthermore, as
can be seen from Figure 2d, the embedding of BP04 only increases the
conductivity of the composite membrane, as compared to pure SPEEK,
only up to about 80°C. At higher temperatures, conductivity of the
composite electrolyte membrane was lower than that of pure SPEEK. It
is possible that a dehydration of BP04 hampers proton transfer and is
responsible for the lower conductivity of the SPEEK/BP04 composite
membrane. Furthermore, it should be mentioned that the mechanical
properties of SPEEK sulfonated to D.S. of 80% is too poor to be used
efficiently in a polymer electrolyte membrane (PEM) in fuel cells.
The present invention is illustrated in further detail by the
following non-limiting examples.
EXAMPLE 1
The composite electrolyte membranes comprising
SPEEK and BP04
Powdered BP04 was dispersed in a dimethyl acetamide
(DMA) solution of sulfonated PEEK and stirred for 24-48 hours. Methods
for sulfonating a polymer such as PEEK are known in the art. One
example thereof is described in Examples 2 and 11. After evaporation of
the solvent, the polymer/BP04 blend was spread over a glass plate and
#~ ~ z2s. ~ ~ ~roEM.wPo
CA 02292703 1999-12-20
dried for 12 hours at room temperature; for 8 hours at 40°C and for
another 12 hours at 120°C under vacuum, thereby eliminating any trance
of DMA. Before their conductivity was measured, the membranes were
stabilized by immersion for several hours in water, which increases their
5 proton conductivity. Of note, the conductivity of the composite
membranes is more stable thermally than the one of SPEEK membranes
(Fig. 2).
EXAMPLE 2
10 Preparation of polymers
Poly-aryl ether ether ketone (PEEK), the chemical
structure of which was shown in Figure 1 b, is a thermostable polymer with
an aromatic, non-fluorinated backbone, in which 1,4-disubstituted phenyl
groups are separated by ether (-O-) and carbonyl (-CO-) linkages. PEEK
15 can be functionalized by the sulfonation technique and the sulfonation
degree can be controlled by reaction time and temperature. The second
component of these composite membranes, heteropolyacids (HPAs), are
known as the most conductive solids among the inorganic solid
electrolytes at near-ambient temperature [4,10-12]. For instance, the
20 hydrate form of tungstophosphoric acid exhibits a room temperature
conductivity of 1.9x10-1 S/cm [11], and its sodium form has a conductivity
of about 10-2 S/cm [12]. The HPAs are soluble in polar solvents, where
they produce stable Keggin type anions such as PW120403-. The strong
acidity of HPAs is attributed to the large size of the polyanion yielding low
delocalized charge density.
#11229.117/DEM.WPD
CA 02292703 1999-12-20
21
One aim of the present invention was to assess the
contribution of hydrated HPAs to the proton conductivity of a hydrophilic
polymer matrix when appropriately embedded thereinto and to assess the
mechanical properties of the polymer film. The fillers exempl~ed herein
include tungstophosphoric acid (TPA), its sodium salt (Na-TPA) and
molybdophosphoric acid (MPA). The binder matrix was partially
sulfonated PEEK (SPEEK). The sulfonated form of PEEK was used in
order to provide the polymer matrix with some hydrophilicity. The
electrical and thermal properties of these new composite
heteropolyacid-SPEEK membranes show their potential as alternatives
to Nafion membrane in PEMFC.
In order to improve proton conductivity, composite
membranes were synthesized in order to incorporate solid
heteropolyacids (HPA) into partially sulphonated PEEK polymer matrices
(SPEEK).
PEEK used in this study was bought from Polysciences,
Inc, USA in the form of extrudates. For the modification of this polymer
concentrated H2S04(95-98%) was used as the sulfonating agent. PEEK
was dried in a vacuum oven at 100 oC overnight. Thereafter, 20 g of
polymer was dissolved in one liter of concentrated sulfuric acid and
vigorously stirred for the desired time ranging from 24 h to 112 h at room
temperature. Then the polymer solution was gradually precipitated into a
large excess of ice-cold water under continuous mechanical agitation.
The agitation was continued one hour further and then the polymer
suspension was left overnight for phase separation. The precipitate was
collected by filtration and, in order to remove any trapped acid, it was
#11229.117/DEM.WPD
CA 02292703 1999-12-20
22
washed several times with an excess of distilled water until the washing
water became neutral. The solid was then dried under vacuum for 8-10
h at 100 oC. The final product is the sulfonic acid form of PEEK, i.e.
SPEEK. It has a yellow or yellowish red color due to self protonation of
carbonyl groups. The SPEEK samples were not neutralized and were
stored in the free acid form (H-SPEEK).
The ion-exchange capacity (IEC) of sulfonated PEEK
polymers was determined by titration: 2-5 g of the SPEEK were placed in
1 M aqueous NaOH and kept for one day. The solution was then back
titrated with 1 M HCI using phenolphthalein as an indicator.
EXAMPLE 3
Heteropolycompounds
The solid HPAs used in the work were purchased from
Fluka Chemicals, and were used as received. The HPAs used were
tungstophosphoric acid, H3PW12040 29H20 (TPA), molybdo-phosphoric
acid, H3PMo12040 29H20 (MPA) and the disodium salt of
tungstophosphoric acid, Na2HPW12040 (Na-TPA).
EXAMPLE 4
Differential Scanning Calorimetry
DSC measurements were performed on SPEEK
samples as well as on SPEEK membranes using a Du Pont 910
Differential Scanning Calorimeter at a heating rate of 10oC/min under
nitrogen atmosphere. Indium (melting point 156.6 oC) was used as the
calibration standard. The DSC experiments were carried out in two steps.
#11229.1171DEM.WPD
CA 02292703 1999-12-20
23
Approximately 10-15 mg samples were dried by heating at 20oC/min to
200oC, then quenched in liquid nitrogen and then heated again to 250oC
at 10oC/min. The glass transition temperature, Tg was calculated at the
intersection of the tangents to the corresponding DSC curve.
EXAMPLE 5
Thermogravimetry
The TGA was carried out using a Du Pont 951 model
thermobalance controlled by a 2100 thermal analysis station. The
samples were first dried at 200oC to remove any moisture/solvent for 30
min, and then programmed from 90 to 900oC at 10oC/min under a
nitrogen atmosphere. Approximately 15-20 mg sample was used for each
run.
EXAMPLE 6
1 H NMR
The 1 H-NMR spectra were recorded on a Varian Unity
Inova spectrometer at a resonance frequency of 399.961 MHz. For each
analysis, a 2-5 wt% polymer solution was prepared in DMSO-d6 and TMS
was used as the internal standard. NMR data were acquired for 64. scans
at a temperature of 30 °C. The acquisition parameters were set to a
spectral window of 6000Hz (15ppm), a pulse angle of 55 degrees,
acquisition time of 2 s and relaxation delay of 1 s. Accurate integrations
of distinct aromatic signals were used to determine the degree of
sulfonation.
#11229.117/DEM.WPD
CA 02292703 1999-12-20
24
EXAMPLE 7
Membrane preparation
The SPEEK polymer was first dissolved (5-10 wt.%) in
dimethylacetamide (DMAc) and the appropriate mass of powdered HPA
was then added to the solution. The resulting polymer mixture was stirred
for 16-24 h, and, after evaporation of most of the solvent, cast onto a
glass plate using a casting knife. The cast membranes were first dried at
room temperature overnight and then at 60oC for 4-6 h and at 80-120 oC
overnight. The content of HPA in the dry product was 60 wt.%.
EXAMPLE 8
Electron Microscopy
The morphologies of the composite polymer membranes
were investigated using a scanning electron microscope (JSM-849,
JEOL). Specimens for the SEM were prepared by freezing the dry
membrane samples in liquid nitrogen and breaking them to produce a
cross-section. Fresh broken surfaces of the samples were vacuum
spray-coated with a thin layer of Au/Pd prior to viewing in SEM.
2.8. Water absorption of membranes
The water absorption of SPEEK membranes was determined from the
difference in weight (W) between the dry and the swollen membranes.
The membrane, cast from DMAc solution after drying, was weighed and
then soaked in water until the weight remained constant. It was then
taken out, wiped with blotting paper and weighed again. The percentage
#11229.117/DEM.WPD
CA 02292703 1999-12-20
of water absorbed was calculated with reference to the weight of the dry
specimen: (Wwet/Wdry-1)x100%.
EXAMPLE 9
5 Conductivity measurements
The proton conductivity of the polymer membrane
samples was measured by the AC impedance spectroscopy technique
over the frequency range of 1-107 Hz with oscillating voltage 50-500 mV,
using a system based on a Solarton 1260 gain phase analyzer. A sample
10 of the membrane with diameter 13 mm was placed in an open,
temperature controlled, parallel-plate electrode test cell, where it was
clamped between two blocking stainless steel electrodes with a
permanent pressure of about 3 kg/cm2. The main shortcoming of an open
cell, the specimen dehydration during the measurement, is compensated
15 by such advantages as the possibility to provide good electrode-specimen
contact (by applying sufficient thrust using an external load) and access
to a larger temperature range (typically up to 150oC). Moreover, the thin
specimen discs (100-500 ?m) soaked in water prior to the test, being
tightly compressed between blocking electrodes, can only lose water
20 through their edges, which is negligible at low temperature over time
scale of the experiment.
The conductivity a of the samples in transverse direction
was calculated from the impedance data, using the relation ( a = d/RS
where d and S are the thickness and face area of the sample
25 respectively, and R was derived from the low intersect of the high
frequency semi-circle on a complex impedance plane with the Re(Z) axis.
#11229.117/DEM.WPD
CA 02292703 1999-12-20
26
The impedance data were corrected for the contribution from the empty
and short-circuited cell.
EXAMPLE 11
Sulfonation of PEEK
There are different methods for sulfonation of PEEK.
Herein, sulfonation by concentrated H2S04 (95-98%) was selected
because it is a simple and safe reaction procedure. It allows to avoid
degradation and cross-linking reactions [13,14] which always accompany
sulfonation with 100% H2S04 or with chlorosulfonic acid. Cross-linking is
presumably the result of sulfone formation, which should be negligible in
H2S04 containing a few percent of water (95-98%), because H O
decomposes the aryl pyrosulfate intermediate, required for sulfone
formation [13].
The PEEK was sulfonated for different reaction times
ranging from 24 h to 112 h to produce polymers of various degrees of
sulfonation. It is known that the degree of sulfonation can be controlled
by changing reaction time, acid concentration and temperature, which can
provide a sulfonation range of 30 to 100% per repeat unit [12,13]. It has
been observed in [13] that PEEK dissolved in H2S04 does not undergo
molecular weight degradation even for long periods of time.
It is worth noting that SPEEK is heterogeneous with
respect to the degree of sulfonation of individual polymer molecule repeat
units. The extent of this heterogeneity is important to consider, as it might
influence SPEEK properties, such as crystallinity and solubility [13,15].
Heterogeneity is a consequence of the random nature of the sulfonation.
#11229.117/DEM.WPD
CA 02292703 1999-12-20
27
Moreover, this heterogeneity is aggravated by the slow kinetic of
dissolution of PEEK in sulfuric acid, which depends on the molecular
weight of the polymer chains and their accessibility (surface of granules
dissolves before bulk). Presumably, the difficulty in dissolving PEEK is
caused by the strong intercrystalline forces. The solubility of PEEK in
strong acids can be attributed to protonation of the ketone groups and in
some cases to chemical modification (e.g. sulfonation) of the polymer
[14,15]. Sulfonation is an electrophilic reaction, therefore, its
effectiveness
depends on the substituents present on the aromatic ring. With PEEK the
hydroquinone unit (between the ether bridges) can be sulfonated under
relatively mild conditions, being doubly activated towards electrophilic
reactions.
The physical and chemical properties of SPEEK depend
on the content of sulfonic groups and the nature of counter ions.
Sulfonation modifies the chemical character of PEEK, reduces the
crystallinity and, consequently affects solubility. For example, at
sulfonation degrees lower than 30% SPEEK is soluble only in strong
H2SO4. Above 30% sulfonation the SPEEK polymers are soluble in hot
dimethylformamide (DMF), dimethylacetamide (DMAc), ~ and
dimethylsulfoxide (DMSO); above 40% in the same solvents at room
temperature; above 70% they are soluble in methanol and at 100%
sulfonation in hot water [13,14].
The ion-exchange capacity (IEC) of SPEEK polymers
measured at room temperature is presented in Fig.1 as a function of time
of sulfonation. It can be seen that IEC is continuously increasing with
#11229.117/DEM.WPD
CA 02292703 1999-12-20
28
reaction time. Upon sulfonation, PEEK reached an IEC close to 2.0 meq/g
within 112 h.
EXAMPLE 12
Determination of the degree of sulfonation
The degree of sulfonation was determined quantitatively
by ' H NMR using a modification of the method described in [16~ for
SPEEK polymers. In the'H NMR spectra, the presence of a sulfonic acid
group causes a significant 0.25 ppm down-field shift of the hydrogen HE
compared with H~, HD in the hydroquinone ring. The nomenclature of the
aromatic protons for the sulfonated PEEK repeat unit is given below:
O HA Hg H~ Hp HB~ Hp~ O
C
O O C
HA HB HE So3H HB. HA.
The presence of each sulfonic acid group S03H will
result in a distinct signal for protons at the E position. Knowing the
intensity of this 'H signal, one can estimate the HE content which is
equivalent to the S03H group content. The 1 H NMR signal for the S03H
group is less easy to record directly as this proton is labile. The ratio
between the peak area of the distinct HE signal, (AHD, and the integrated
peak area of the signals, corresponding to the other aromatic hydrogens
(AHA,~.,B,B.,c,D), is expressed as:
n / (12-2n) = AHE / ~A H,q,A',B,B',C,D ~ (0 s n ~ 1)
#11229.1171DEM.WPD
CA 02292703 1999-12-20
29
where n is the number of HE hydrogens per repeat unit. An estimate of the
degree of sulfonation x is obtained as : x = n x 100%.
The 'H NMR spectra of different sulfonated PEEK
samples dissolved in DMSO-d6 are shown in Figure 2e. Since
non-sulfonated PEEK is insoluble in any solvent except in strong acids,
its 'H NMR spectra could not be not recorded. In fact in the low
sulfonated SPEEK, several repeat units are unsubstituted and the HC
and HD of the hydroquinone aromatic ring of PEEK repeat units in the
SPEEK polymer appear as a characteristic singlet at 7.25 ppm. The
intensity of this singlet decreases as the degree of sulfonation increases.
Similarly, two new signals increasing in intensity with higher degree of
sulfonation (D.S.) can be clearly observed. They are associated with the
hydroquinone ring of the SPEEK repeat unit. HC is a doublet of doublets
at 7.20 ppm and the two HB' protons are a doublet at 7.00 ppm, shifted
upfield by proximity of the S03H group. The multiplicity of HC doublet of
doublets, is caused by its four bond coupling with HE (2.5 Hz) and its
three bond coupling with HD (8.8 Hz). The intensities of both H~ and the
two HB. increase proportionally with the intensity of HE.
The D.S. calculated from the 1 H NMR spectra of SPEEK
samples is plotted in Fig.3. These data are also reported in Table 1
together with the sulfur content, calculated from them, and with results of
elemental analysis for sulfur.
#11229.117/DEM.WPD
CA 02292703 1999-12-20
30 ~-
Table 1 : Sulfur content of different sulfonated SPEEK samples
Sample Degree of sulfonationSulfur content(wt.%)
(D.S.)
by NMR, mol% Experimental Calculated
(elemental from D.S.
analysis)
PEEK 0.0 0.0 0.0
SP 1 48 3.90 4.3
SP2 51 4.45 4.63
SP3 56 5.15 5.06
SP4 68 6.60 6.08
SPS 72 7.0 6.4
CA 02292703 1999-12-20
31
As can be seen the sulfur percentage determined
experimentally and the one, calculated from the estimated values of D.S.,
are in reasonably good agreement. The ion exchange capacity (IEC) and
the sulfur content expressed in mmol/g, given in Figure 4, also agree
rather well with each other. Thus, as the sulfonation proceeds with time
from 24 to 112 hours, ionomers of PEEK, having degrees of sulfonation
ranging from 40-80 mol% are obtained. Using this sulfonation method,
even at room temperature the dissolution/reaction of the PEEK granules
during a short period of 24h produced a SPEEK, having 0.4 sulfonic
groups per repeat unit.
EXAMPLE 13
Glass Transition Temperature: DSC Studies
The values of the glass transition temperature, Tg were
determined by the DSC technique at a heating rate of 10 oC/min. It is of
interest to compare how the D.S. affects Tg of sulfonated PEEK. Fig. 5
displays SPEEK Tg values as function of D.S. As discussed above, it is
important to use well purified and carefully dried samples to obtain well
defined and reproducible Tg values. For these particular experiments,
samples were first dried to 200 oC at a heating rate of 20 oC/min to
remove any solvent and water. It was observed that samples were not
discolored or degraded. Special care was exercised at this step as even
a small quantity of water or solvent acts as a plasticizer and could lead to
depressed Tg values. After quenching with liquid nitrogen, the DSC
experiments were repeated at a heating rate of 10oC/min to obtain the Tg
value.
#11229.117/DEM.WPD
CA 02292703 1999-12-20
32
From Figure 5, it can be seen that Tg increases
monotonically with the degree of sulfonation. Comparing the glass
transition results of the sulfonated PEEK samples, it was observed that
the introduction of -S03H groups into the PEEK polymer increased the
Tg by as much as 65 oC for the sample having 80% degree of sulfonation
relative to Tg of 150 oC for non sulfonated PEEK. These results are in
complete agreement with reference [15j for PEEK and similar behavior
has also been observed for sulfonated polysulfone [17-20). The increase
in Tg with the degree of sulfonation results from the increased
intermolecular interaction by hydrogen bonding of S03H groups (ionomer
effect). Increased molecular bulkiness may also contribute [19]. It is likely
that these intermolecular forces hinder the internal rotations compared to
unsulfonated PEEK. Interestingly similar conclusions have been reached
from the observations that sulfonation increases the molecular
dimensions of PEEK in solution [13].
EXAMPLE 14
Thermal Stability: TGA studies
The thermal stability of sulfonated and pure PEEK was
investigated by TGA. The samples were pretreated at 200oC and then
the experiment was run from 90 to 900 oC at a heating rate of 10 oC /min
under flowing nitrogen. Examples of TGA and DTG curves of PEEK and
sulfonated PEEK (SPEEK) are shown in Fig.6. Since PEEK is a high
temperature resistant polymer, the onset of significant weight loss for this
polymer starts at about 520 oC. This weight loss may be due to the main
chain decomposition of the polymer. The degradation takes place through
#11229.117/DEM.WPD
CA 02292703 1999-12-20
33
the pyrolysis of PEEK with formation of phenols and benzene. The major
pyrolysis product of PEEK, phenol, is produced in the early chain scission
reaction involving ether rather than carbonyl linkages [21).
From Figure 6, it can be seen that two weight loss steps
are observed for sulfonated PEEK which is reflected by two broad peaks
in the DTG curve in two separate temperature ranges in contrast to the
one weight loss peak for non-sulfonated PEEK. The first weight loss peak
in SPEEK is believed to be due to the splitting-off of sulfonic acid groups.
A similar observation was made for sulfonated polysulfones [18,19). The
maximum peak temperature of the second peak for all SPEEK samples
shifted to 560 oC from 600 oC for PEEK. The maximum temperature for
the peak, corresponding to the sulfonic acid decomposition, was found to
be 360oC in the TGA/DTG for all the sulfonated samples. From the first
peak, the weight loss corresponding to the sulfonic acid decomposition
was determined (Table 2) and plotted against the degree of sulfonation,
as shown in Figure 7.
#11229.117/DEM.WPD
CA 02292703 1999-12-20
34
Table 2: Measured and calculated weight loss of different SPEEK samples due to
the
splitting-off of sulfonic acid groups
Sample. Degree of sulfonationWeight loss Weight loss
by NMR, mol% by TGA, wt.% calculated, wt.%
SN1 40 13.0 10.0
SP2 51 15.0 12.2
SN2 65 16.5 15.3
SN3' 70 17 16.4
SN4' 74 19.5 17.0
SNS 76 19.0 17.4
SN6" 80 20.0 18.2
Polymers, used for preparation of the composite HPA/SPEEK membranes
CA 02292703 1999-12-20
The theoretical weight loss calculated assuming that the
splitting-off of a sulfonic acid group releases one S03 molecule, is given
in Figure 7. As can be seen, there is a rather close agreement between
the weight loss calculated theoretically and the one determined from TGA
5 experiments for the sulfonic acid decomposition. This indicates that the
sulfonic acid introduced after sulfonation is lost in this thermal
desulfonation step. The onset temperature of SPEEK, Tonset given in
Figure 8, indicates that SPEEK membranes are thermally stable up to
approximately 300 oC and that this temperature is only marginally
10 affected by an increase in the degree of sulfonation up to 80%. From the
thermal analysis results it can be concluded that the thermal stability of
the SPEEK membranes is good and that SPEEK will be stable for fuel cell
applications both at low and medium temperatures.
15 EXAMPLE 15
Water absorption of SPEEK membranes
The amount of water absorbed in sulfonated PEEK
membranes was determined in two ways as described in the experimental
section. The dry SPEEK membranes were immersed in H20 at room
20 temperature for 24 h until there was no further weight gain, then wiped
with blotting paper and weighed. In another experiment the wet
membranes were vacuum dried at 100-120 oC and their water loss
determined, then they were again soaked in water and weighed after
blotting. The vacuum dried membranes when immersed in water for
25 hydration regained the initial weight of the wet membranes within 4-6
hours.
#11229.117lDEM.WPD
CA 02292703 1999-12-20
36
The main purpose of sulfonating aromatic PEEK is to
enhance acidity and hydrophilicity as it is known that the presence of
water facilitates proton transfer and increases the conductivity of solid
electrolytes: The enhancement of hydrophilicity by sulfonation of PEEK
polymer can be followed by water absorption of SPEEK membranes as
a function of the degree of sulfonation. In Fig. 9 the mass of water
absorbed by SPEEK membranes expressed in wt.% of the dry membrane
is plotted against the number of sulfonic acid groups. The water uptake
of SPEEK membranes increased with increasing sulfonation level and it
reached 120 wt% for a SPEEK membrane of 80% D.S. (containing 0.8
S03H group per repeat unit). These results show that the water
absorption of SPEEK membranes increased linearly up to a D.S. of 65%
and very rapidly above 70%. In the highly sulfonated SPEEK membrane
the density of S03H group is high and there may be a possibility of
clustering (or agglomeration). Clustered ionomers absorb more water,
therefore a large water uptake may be suggestive of the presence of
ion-rich regions [14]. These results are important as proton conduction
depends on the presence of water in the polymer structure.
A similar trend was observed for the water content of
Ballard's BAM3G membranes [6,7j, which showed an increase in water
content with equivalent weight (EW) decreasing from 920 to 375, where
the EW is expressed. as g/mol-S03. The water uptake in BAM3G
membranes increased exponentially below 410 EW to more than 250%.
So in fact a decreased EW corresponds to a higher D.S. which brings
about an increase in water content. This is similar to our results where the
water content increased linearly with an increase in sulfonic group density
#11229.117/DEM.WPD
CA 02292703 1999-12-20
37
up to a D.S. of ca.65% and then increased exponentially. The water
content of our SPEEK membranes is higher than in Nafion117 (34%),
Dow membranes (56%), NASTA membranes (48%) /7/ and polyimide
membranes (30%) [22]. This high water adsorption capacity of SPEEK
membranes compared to others is also a characteristic of BAM3G
membranes. Consequently, both BAM and SPEEK membranes should
be equally efficient in DMFC water management.
EXAMPLE 16
Scanning Electron Microscopy
The morphology of the composite SPEEK/HPA
membranes has been studied by scanning electron microscopy. The
SEM micrograph of the cryogenic fracture of a composite membrane is
presented in Fig.10. It can be seen from this micrograph that the solid
HPA is well mixed with SPEEK and shows no agglomeration after
membrane preparation. High magnification micrographs of these
membranes indicate that HPA particles are homogeneously distributed
within the SPEEK polymer.
EXAMPLE 17
Conductivity studies
Membranes were prepared from the SPEEK polymer,
characterized in Examples 12 to 16, by casting from a solution of polymer
in dimethylacetamide. The various membranes obtained had a thickness
ranging from 300-500 (m and a degree of sulfonation ranging from
40-80%. The proton conductivity of these differently sulfonated SPEEK
#11229.1171DEM.WPD
CA 02292703 1999-12-20
38
membranes was measured in the transverse mode as described in the
experimental section. Prior to conductivity measurements all the
membrane samples were soaked in water for hydration. The effect of the
degree of sulfonation on the conductivity of SPEEK membranes at room
temperature is shown in Fig. 11. The conductivity was found to increase
with the degree of sulfonation and reached a value of 8x10-3 S/cm for the
membrane with 80% degree of sulfonation. With the increase in the
number of sulfonic acid groups in the SPEEK polymers, there is an
enhancement in their solubility in organic solvents e.g. in
dimethylacetamide and their crystallinity decreases as previously reported
in ref.[13]. As a consequence, the polymer becomes more hydrophilic and
absorbs more water. These water molecules facilitate proton transport,
since proton transfer essentially involves hydrated species, such as
H30+, H5O2+ etc. Hence, the sulfonation raises the conductivity of the
PEEK not only by increasing the number of protonated sites (S03H), but
also through formation of water mediated pathways for protons. Proton
conductivity depends also on such factors as density and distribution of
sulfonic acid sites, degree of dissociation of sulfonic acid functions and
some others.
The dependence of the conductivity on temperature for
a series of SPEEK membranes at different D.S. is presented in Fig. 12.
As can be seen, the conductivity depended substantially on temperature
and this dependence is strongly affected by the sulfonation level. The
membrane sulfonated at 40% did not show any improvement in
conductivity with temperature, which on the contrary even decreased with
temperature rise. The conductivity of the membrane sulfonated at 48%
#11229.117/DEM.WPD
CA 02292703 1999-12-20
39
increased up to 85 o C and then dropped sharply to very low values. As
to membranes sulfonated at 70 and 74%, they showed a continuous
increase in conductivity and a gradual decrease beginning only above
100 oC. The conductivity of the membrane sulfonated at 80% behaved
differently with temperature than the others. It increased slowly in the
temperature range of 20-50 oC, showed a fast rise in the 50-100 oC
range and then gradually increased again up to 145oC.
The temperature dependence of SPEEK conductivity
indicates the presence of two competing trends, one of which enhances
the conductivity and the other which reduces it. As ionic conductivity of
electrolytes is in general thermally stimulated, it is natural to expect a
rise
in proton conductivity with temperature. The decay in the conductivity
curves above a certain temperature, which is observed (Fig.12) for all but
one SPEEK samples, suggests that there is dehydration of the polymers
during the measurements. One can see that this dehydration starts at
room temperature for 40%SPEEK, and then shifts towards higher
temperatures as D.S. increases. This indicates, that it is not only the
water uptake which depends on the sulfonic group content (Fig.9), but
also the capacity of the membrane to retain water. The sample with 40%
D.S. for instance, loses water so fast that dehydration suppresses any
conductivity rise. With an increase in D.S. the conductivity curves show
a less steep decline, and for 80%SPEEK water retention is so high that
dehydration induces only a decrease in the conductivity rising rate up to
145oC. Since the low sulfonated membranes were not found to maintain
high conductivity at temperatures higher than 90 oC, the SPEEK
#11229.1171DEM.WPD
CA 02292703 1999-12-20
polymers sulfonated at 70-80% were selected as matrix for the
preparation of composite membranes with heteropolyacids.
EXAMPLE 18
5 Preparation of the SPEEKIHPA composite electrolyte
memrane and properties thereof
Composite SPEEK/HPA membranes were prepared by
incorporation of 60 wt% HPA (TPA, MPA and Na-TPA) into three SPEEK
matrices with 70, 74 and 80% degree of sulfonation. The thickness of
10 these composite membranes varied from 150 to 300 ~m (Table 3).
#11229.117/DEM.WPD
CA 02292703 1999-12-20
41
Table 3: Conductivity and water absorption of the composite membranes
Thickness
Membrane Membrane Type of Water-uptake,Conductivity,
d wet membrane, S/cm
esignation pxrl wt.% 25 C 100 C
SN3 Pure SPEEK 70 300 48 1.3x10'3 6.7x10'3
HPA 1 SPEEK 70+ TPA 170 64 3. Sx 1'.7x 10'2
10'3
HPA2 SPEEK 70+ Na-TPA375 143 2.9x 10'31. Sx 10'2
HPA3 SPEEK70+MPA 200 94 3.1 x 1.1 x 10'2
10'3
SN4 Pure SPEEK 74 300 62 2.1x10-3 9.5x10-3
HPA4 SPEEK74+ TPA 160 190 5.1 x 2x 10-2
10-3
HPAS SPEEK74+NaTPA 300 160 4.5x10-3 1.6x10'2
HPA6 SPEEK 74 + MPA 200 146 3.6x10'3 1.2x10'2
SN6 Pure SPEEK 80 500 120 8.0x10-3 7.6x10-2
HPA7 SPEEK80+TPA 300 600 2x10-2 9.5x10'2
HPA8 SPEEK80+NaTPA 200 400 1.4x10-2 5.8x10'2
HPA9 SPEEK 80 + MpA 160 320 1.2x10'2 3.0x10-2
CA 02292703 1999-12-20
42
The conductivity of the composite membranes was
measured in the transverse direction at temperatures ranging from 20 to
150 oC. The results of conductivity measurements, showing the effects
of solid HPA incorporation and temperature on the conductivity of these
composite membranes, are given in Fig.13-15. They are presented for
three different sets of composite membranes, containing the various solid
HPA into each of the three SPEEK matrices. Incorporation of HPAs
increased the membrane conductivity compared to the pure membrane,
except in the case of the 80% sulfonated PEEK for which only the
TPA/SPEEK membrane showed enhanced conductivity at temperatures
higher than 80 oC. In all cases the introduction of the HPA's improves the
high temperature stability of the conductivity. The conductivity of all the
composite membranes behaved in a similar fashion: the TPA based
membranes showed a higher conductivity than Na-TPA and MPA
composites. It appears that TPA, being a stronger acid, yields
systematically a higher proton conductivity increase as well as a better
water retention at high temperature.
These results show that by incorporation of HPA into
SPEEK the conductivity was raised up to values, close to that of pure
HPA. Hydrated wafers of TPA 13 mm in diameter and 0.5-2 mm thick at
ambient temperature exhibited a high conductivity of 1.9x10-1 S/cm
[11,12]. The highest conductivity of 1.5x10-1 S/cm was obtained for
composite membrane containing TPA into 80% sulfonated SPEEK at 120
oC. These results clearly show that it is possible to increase significantly
the proton conductivity of SPEEK matrices by incorporation of a solid
inorganic acid such as solid HPAs.
#~ ~ 22s. > > ~roenn.wP~
CA 02292703 1999-12-20
43
The water uptake at room temperature and the
conductivity of some of these membranes along with their thickness are
given in Table 3. The water uptake of the composite membranes
increased upon incorporation of hydrophilic solid HPA into SPEEK
matrices. It can be noted that the pure SPEEK membrane at 80%
sulfonation shows higher conductivity above 100 oC than its composite
membrane containing MPA or Na-TPA (Fig.15). The pure SPEEK 80
membrane has a thickness of 500 ,um which is three times higher than the
MPA composite membrane (160 ,um) and more than twice the thickness
(200,um) of the Na-TPA containing membrane. The moisture retention
and dehydration rate can be affected by the thickness of the membranes.
The thicker membranes will thus have higher moisture retention and lower
dehydration rate upon heating, than the thinner membranes. At the same
time the positive effect of HPA introduction can be clearly seen at
temperatures below 60oC for MPA and below 80oC for Na-TPA. This
difference can also be attributed to the different membrane thickness.
Comparing these results with studies by Park et al [23]
for sulfonated polysulfone membranes containing solid TPA, where the
conductivity of 1.6x10-3 S/cm was registered, the proton conductivity
displayed by HPA-SPEEK membranes proved to be significantly higher.
In a work of Tazi & Savadogo [24) a conductivity of 9x10-2 S/cm was
reported for H2S04 acidified Nafion-silicotungstic based membranes
having thickness of 200-300?m. Despite the fact that only fluorine-free
materials were used, the membranes reported in this work exhibited
conductivities of the same order of magnitude. Thus, the membranes of
the present invention show conductivities which are comparable to those
#11229.1171OEM.WPD
CA 02292703 1999-12-20
44
of Nafion and Nafion-derivatives but provide the advantage of being
fluorine-free and environmentally friendly.
From the results of water absorption it can be noted that
the water content of the SPEEK membranes increases with D.S. At the
same time Tg also increases, reflecting an increased intermolecular
association through the polar ionic sites. The increasing association of the
PEEK repeat unit with the anionic counter charge(-S03-) immobilized on
the polymer backbone of a neighboring chain suggests a less effective
separation of the aqueous phase compared to Nafion. This increased
association is not however sufficient to upset the effect of the increased
water content, which brings about an increase in conductivity of SPEEK
membranes. The proton conductivity in aromatic membranes depends
much more on the amount of water as second phase than in the case of
Nafion [25]. The higher water content of these membranes helps the
migration of proton, which involves such species as H30+ and H205+,
needed for conduction.
The composite membranes containing HPAs display
higher Tg than the pure SPEEK membranes. For composite membranes
containing TPA and MPA in 70% sulfonated SPEEK for example the
glass transition temperature increased from 208 oC to 215 oC. The glass
transition of the composite membrane not only on the sulfonation level but
also on the type of solid acid HPA. This increase in Tg of the composite
membranes may be due to a reduction in chain mobility most probably
caused by the interaction of the solid acid with polar groups of the SPEEK
polymer chain [17]. Such interactions are also consistent with the SEM
micrographs of the composite membranes, which show solid HPA
#11229.117IDEM.WPD
CA 02292703 1999-12-20
homogeneously mixed with the SPEEK polymer phase (Fig.10). These
pictures suggest that the conductivity of these composite membranes is
favored by the uniform spatial distribution of HPA particles which
minimizes interparticle distances. In addition SPEEK, being a conductive
5 polymer, is acting as an effective binder for HPA particles contributing to
the conduction process.
The water content of these composite membranes was
also found to increase over the pure SPEEK membranes. The water
uptake (Table 3) of composite membranes based on 70% sulfonated
10 SPEEK increases from 48% to 143%, for composite membranes based
on 74% sulfonated SPEEK increased from 62% to 190%, and the water
uptake of the composite membranes based on 80% SPEEK increased to
high values from 120 to 600%.
It is worth noting however, that the increased water
15 uptake associated with the incorporation of HPAs to the polymer matrix
is only one factor affecting the membrane conductivity. Other factors,
including the polymer intrinsic conductivity, strength, density and softness
of the solid acid sites, the solid phase loading, particle size and spatial
distribution, the aqueous phase dispersion, all affect the dependence of
20 conductivity on the water content. In the case of Nafion membranes which
are described as a nanoporous inert "sponge" for water of hydration, this
water shows little interaction with the polymer chain and forms hydration
shells around the sulfonic acid groups [25]. In HPA/SPEEK membranes
the aqueous phase is more continuous. This leads to higher water uptake
25 (swelling) and contributes to a greater extent to proton conductivity.
#11229.1171DEM.WPD
CA 02292703 1999-12-20
46
Examples of Arrhenius plots of conductivity for the 70%
SPEEK based composite membranes are given in Fig.16. For all
composite membranes reported here, the activation energy for proton
conduction was similar and close to 15 kJ/mol. This result suggests a
common charge transfer mechanism. The relatively low measured values
of activation energy are suggestive of a liquid like mechanism most
probably based on the Grotthus reorientational proton transfer scheme
[4J.
Of note, the conductivity of composite membranes was
not affected by storage in water at room temperature for nine months.
SPEEK polymer is, therefore, acting as an effective polymer matrix for
solid HPA. Taken together, the results presented herein show that
composite membranes containing solid HPA are stable for the long term,
with no change in their conductivity. These composite membranes
possess good stable mechanical strength and flexibility except the ones
based on SPEEK sulfonated at 80% which become weak because of
excessive swelling. The best mechanical properties were found with the
70 % sulfonated SPEEK based composites.
Conclusion
In this study a series of composite membranes have
been prepared by incorporation of tungstophosphoric acid, its disodium
salt and molybdophosphoric acid into partially sulfonated PEEK polymer.
These membranes exhibited a rather high conductivity of 10-2 S/cm at
ambient temperature and up to a maximum of about 10-1 S/cm above
100oC. The DSC studies of these membranes for the glass transition
temperature showed an increase in its value due to the incorporation of
#11229.117/DEM.WPD
CA 02292703 1999-12-20
47
both sulfonic acid groups and the solid HPA into the PEEK polymer. The
increase in Tg of the composite membranes compared with the Tg of
pure SPEEK suggests an intermolecular interaction between SPEEK and
HPAs. These membranes are thermally stable up to (275 oC,
mechanically strong and flexible. They preserve the high conductivity
during storage in water for several months. These membranes are easy.
to prepare, are much less expensive than the commercial perfluorinated
membranes and are fluorine-free. The high proton conductivity combined
with their long term stability qualify the HPA/SPEEK composite
membranes to be considered for use in PEM fuel cells as alternatives to
Nafion membranes.
EXAMPLE 19
The proton conductivity (a) of aluminum boron
phosphate was measured using impedance spectroscopy. It was found
to be inversely related to the AI/B ratio and calcination temperature, and
drastically increased with adsorbed water content. The water assisted
conductivity of low aluminum AI-BP04 was registered to be approaching
as high as 10-2 S cm-1. At the same time partial replacement of B with
AI in boron phosphate substantially enhanced its stability in aqueous
surrounding. It appears that the AI addition suppresses the formation of
completely hydrolyzed species and thereby increases inertness of solid
towards water.
Samples of AI-BP04 were prepared from boehmite,
orthophosphoric and boric acids. First the appropriate amount of
boehmite was dissolved in hot H3P04 and then H3B03 was added to the
#11229.117/DEM.WPD
CA 02292703 1999-12-20
48
solution. It was stirred continuously in a crystallizing pan at 120oC in the
air until a thick mass was formed, which was then kept without stirring at
the same temperature for 6 hours. The obtained solids were calcined for
12 h in air at 400-1000oC, ground and a powder of <60 mesh grade was
separated by sieving. A stoichiometric BP04 was prepared in a similar
manner to be used as a reference throughout the work. The list of the
prepared AI containing samples is reported in Table 4, along with the
results of the X-ray diffraction analysis.
#11229.117/DEM.WPD
CA 02292703 1999-12-20
49
Table. Phase composition
of Al-BP04
samples
(B+Al)/P Al/B Sample X-ray diffraction
DesignationPhase compositionI,~ /
IB~'
2:98 2BAP100 BPOa -
B2O3
: SBAP 100 B pO4 -
95
Al(P03)3 0.32
1.0 10:90 lOBAP100 BPO4 -
~p04 0.21
Al(P03)3 1.0
20:80 20BAP100 BPO4 -
A1P04 0.73
Al(P03)3 2.71
5:85 SBAP90 BP04 -
A1P04 0.09
Al(P03)3 0.52
10:80 IOBAP90 BPO4 -
0.9
A1P04 0.24
Al(P03)3 1.86
20 20BAP90 B pO4 -
:
70
A1P04 1.61
_ Al(P03)3 3.0
30:60 30BAP90 Bpp4 _
A1P04 6.36
1.2 30:90 30BAP120 BP04 _
A1P04 3,24
~'I x - Maui peak aCeaS Of: B2O3 (Id ~ 3.2()~ ~4 (~=4.08) and f~(~3)3 ~= 5.48)-
As I B~ the peak of 4% in the BP04 pattern (d = 3.07~r) is taken (marked
with * in Fig.l).
CA 02292703 1999-12-20
Powder X-ray diffraction spectra, of the pelletized
samples were recorded with a Philips diffractometer using Cu-K?
irradiation with a step size of 0.02 degrees and a count time of 1.2 s per
point.
5 Water adsorption experiments were carried out using an
AD-2 microbalance (Perkin-Elmer) equipped with a
temperaturelatmosphere controlled cell. About 0.2 g of solid was dried at
350oC following which some water was introduced at room temperature
into the cell in order to ensure 100% relative humidity (RH). A constant
10 weight was usually reached over a period of 12-24 h.
The details of MAS-NMR measurements are described
in Mikhailenko et al. 1998 (supral). The 11 B, 31 P, 27AI and 1 H MAS
NMR spectra have been acquired at room temperature using a
Bruker-300 spectrometer with 4 mm zirconia rotors spun at 11 kHz.
15 Chemical shifts were respectively referred to BF30(C2H5)2, 85%H3P04,
aqueous solution of AI(N03)3 and tetramethylsilane (TMS), used as
external standards.
Dehydration of the samples of AI-BP04 has been
carried out prior to MAS-NMR measurements directly in the rotor at
20 200oC in vacuum for 4 hours. In order to obtain re-hydrated samples,
those were exposed to saturated water vapor (100%RH) in a desiccator
at room temperature overnight.
The impedance spectroscopy measurements were
performed at ambient temperature over the frequency range 1 to107 Hz
25 with oscillating voltage 100mV, using SI 1260 impedance/gain-phase
#11229.117/DEM.WPD
CA 02292703 1999-12-20
51
analyzer (Solartron). The impedance data were corrected for contribution
from the empty and short-circuited cell.
The high affinity of AI-BP04 for water was a barrier to
the application of a conventional method [27] of measuring the
conductivity of compressed pellets at controlled RH. Storage of pellets
of some AI-BP04 samples at 100%RH yielded a soft, difficult to handle
mass. Therefore, we used the following procedure to measure the
conductivity of powders. Measured amounts of solid (about 0.5g) and
water were mixed together on a balance and immediately placed in a
glass tube, where it was compressed between faces of two brass pistons
used as electrodes with a pressure of 10 kg/cm2. The tube inner diameter
was 6 mm and the length of a sample was about 10 mm. Each sample
was weighed together with the glass tube before and after the impedance
test in order to detect any decrease in water content. Since the duration
of acquisition of an impedance spectrum did not exceed three minutes,
the water content did not change much and the error in the estimation of
water content was within ca 20%. The procedure of impedance
measurement was repeated twice or three times and average values of
conductivity as function of the water/solid weight ratio were calculated
from the data obtained.
Phase composition and solubility
The results of X-ray difFraction analysis for some of the
samples studied herein showed that the diffraction pattern of
stoichiometric BP04 used as reference contains only peaks, attributable
to tetrahedral BP04 (JCPDS pattern 34-132). These peaks were also the
#11229.117lDEM. WPD
CA 02292703 1999-12-20
52
dominant component in the X-ray diffraction patterns of all AI-BP04
samples.
The XRD pattern of 2BAP100 displayed also the main
lines of B203 corresponding to d=6.08, d=3.21 and d=2.92 of JCPDS
reference pattern 6-297. The other AI-BP04 samples evidenced the
presence of aluminum phosphate AIP04 and ultraphosphate AI(P03)3
(JCPDS reference diffraction patterns 11-500 and 13-266 respectively).
The phase composition of the AI-BP04 samples is presented in Table 4.
The peak intensity ratios for aluminum phosphate main lines referred to
BP04 reflection at d=3.07 (a peak in the BP04 pattern with I=4% are also
shown in Table 4. Not claiming this approach as quantitative, it
nevertheless gives an idea of the order of magnitude for the phase ratios
in the solids.
It is evident from Table 4 that as AI/B ratio rises, first
ultraphosphate AI(P03)3 and then phosphate AIP04 phases appear and
their contents gradually increase. At high AI/P ratio (sample 30BAP90)
and when the (B+AI)/P stoichiometry is shifted to the (B+AI) side (sample
30BAP120) ultraphosphate was not registered and the phase composition
comprised essentially BP04 and AIP04.
The impact of AI in addition to changes in the phase
composition of the specimens, showed an alteration of the structure of
BP04 itself as assessed by variations in the lattice parameters. This
brings about an increase in unit cell volume by about 0.9% over the range
of concentration studied here.
In order to determine how the incorporation of AI into the
BP04 lattice and the existence of the aluminum phosphates affects the
#11229.117/DEM.WPD
CA 02292703 1999-12-20
53
stability of BP04 in aqueous surrounding, the solubility of the specimens
was measured. The test was carried out according to the procedure
described elsewhere [in Mikhailenko et al, 1998 (supra)j, which comprised
stirring a 1wt.% slurry of the solid in distilled water for 15 h, followed by
separation of the transparent liquid aliquot from the precipitated solid, its
evaporation and weighing of the soluble solid residue. It is seen from
Figure 17 that the thermal treatment dramatically decreases the apparent
solubility of AI-BP04, reducing it to less than 4% at any composition after
calcination at 1000oC in exactly the same manner as for BP04.
Comparing the solubility of the AI-BP04 samples, calcined at lower
temperature, with that of BP04 indicates, that the aluminum boron
phosphates are essentially less soluble. Thus, introducing the aluminum
enhanced the stability of boron phosphate in water.
Electrical conductivity
The conductivity of the samples was calculated from
their resistance which was derived from the extrapolated low intersect of
the high frequency semi-circle on a complex impedance plane with Re(Z)
-axis, (data not shown). These values usually were in good agreement
with the ordinates of the plateaus in the Bode plot of conductivity ( Re(Y)
vs frequency). Fig. 18 presents the electrical conductivity of AI-BP04,
preliminary calcined at 400oC and 1 OOOoC, as function of water content.
Measurements were performed at room temperature. The arrows in the
figure indicate the average values of water uptake by the solids
conditioned at 100%RH and room temperature. These values depend
mainly on calcination temperature but vary within rather narrow limits of
about 5% for the solid compositions studied here. It is clear from the
#11229.1171DEM.WPD
CA 02292703 1999-12-20
54
figure that the conductivity of AI-BP04 is sensitive to adsorbed water. It
increases with water/solid ratio from Q=10-7 to 8x10-3 S cm-1 for
specimens, calcined at 400oC. The conductivities of the solids calcined
at 1000oC change over a smaller range of values from 10-6 up to 10-3 S
cm-1. The specimens, calcined at higher temperature, showing the water
uptake of ca 6% at 100%RH, are much less hydrophilic than the ones
calcined at 400oC with uptake of ca 25% at saturation. The latter samples
almost all reach their top conductivities at saturation or at water content
only slightly above 100%RH. As to the solids, calcined at 1000oC, their
conductivities at 100%RH are below 10-5 S cm-1. They however continue
to increase up to 10-3 S cm-1 when extra moisture is added.
After calcination at 400oC, the specimens with low AI
content (2BAP100, 5BAP90, 5BAP100) have essentially the same
conductivity as the reference BP04 whereas their stability in water was
significantly higher (Fig. 17). However the solids containing AI appear to
be less stable with respect to calcination than pure BP04, as their
conductivity decreases faster with Tcalc than that of BP04. Treatment at
1000oC brings about a decrease in the conductivity of all AI-BP04
specimens to values about half an order of magnitude lower than the
starting points at 400oC. The less conductive specimens are the ones
which have higher AI/B ratios (data not shown). This can be seen more
clearly from the dependence of Q on the AI content presented in Fig. 19
in the form of curves, corresponding to different calcination temperatures.
It can be seen from the figure that for all series of specimens the
conductivity rapidly drops with an increase in AI content.
#11229.1171DEM.WPD
CA 02292703 1999-12-20
MAS NMR spectroscopy
MAS NMR spectroscopy was employed in this work in
order to examine the structural changes of the solids during calcination
and as a result of their interaction with water. The spectra obtained from
5 the 27AI, 31 P, 11 B and 1 H nuclei were recorded for dehydrated and
rehydrated 20BAP100 and 20BAP90 calcined at 400 and 1000oC and
were related to the spectra of the same nuclei (excepting 27AI) recorded
for the reference BP04 sample.
"B MAS NMR spectra (not illustrated) of all samples
10 studied were invariant under the treatment (calcination or hydration) and
exhibited a single resonance with an isotropic chemical shift of b = -3.7
-3.8 ppm from BF3Et20. This signal is assigned to tetrahedrally
coordinated B04 with the shift consistent with a = -3.3 registered for
anhydrous commercial BP04 from AIfaChem [28]. These results evidence
15 that in AI-BP04 boron coordination is little affected by hydration and does
not depend on the thermal treatment and aluminum content in AI-BP04.
A similar result was previously observed for BP04 [Mikhailenko et al.,
1998 (supra)] where 11 B MAS NMR spectra did not change after thermal
treatment and hydration. It should be mentioned that the 11 B chemical
20 shift range is narrow [29], so that the partial hydration of BP04 species
may escape notice, provided B remains tetracoordinated.
The 27AI NMR spectra of both hydrated and dehydrated
samples of 20BAP100 and 20BAP90 calcined at 400oC were very much
alike (data not shown).
25 After hydration 27AI NMR MAS spectrum of 20BAP90
(data not shown) remained identical to that of the dehydrated parent solid,
#11229.117/DEM.WPD
CA 02292703 1999-12-20
56
while in the spectrum of 20BAP100 the intensity of the peak from
6-coordinate aluminum became bigger and a downshift shoulder
appeared at b = 8 ppm. This clearly indicates that adsorbed water alters
the coordination environment of at least part of the aluminum atoms.
The tendency of AI-BP04 to hydrolysis as it can be seen
from Fig. 17 is critically dependent on calcination temperature and
aluminum content. It decreases with calcination obviously due to the
increased degree of crystal perfection, namely decreased lattice
distortion, longer chains of B04-P04 tetrahedra and larger crystallite size
[Mikhailenko et al., 1998 (supra)]. Aluminum phosphates introduced into
BP04 acts in a similar way, substantially decreasing the solubility of these
solids (Fig. 17). The difference in stability between BP04 and AI-BP04
in water is most significant at low calcination temperature (400oC), and
then decreases as temperature is raised, becoming negligible when the
solids are calcined at 1000oC.
It is believed that this occurs because at comparatively
low calcination temperature the structure of the solids is poorly ordered,
and more stable A104 tetrahedra, taking terminal positions, prevent boron
phosphate from hydrolysis and subsequent dissolution. Calcination
performed at higher temperature brings about the formation of better
crystallized bulk aluminum phosphate, which is less dispersed in BP04
crystals and thus has less impact on its solubility. However the BP04
phase itself, when calcined at higher temperature becomes sufficiently
water resistant, and as a consequence there is no difference between the
solubility of BP04 and AI-BP04 calcined at 1 OOOoC.
#11229.117/DEM.WPD
CA 02292703 1999-12-20
57
In any event, without being limited by a particular theory,
the results presented herein confirm that the partial replacement of boron
with aluminum in boron phosphate brings about an increase in inertness
of the solid towards water. In some cases it is achieved at the expense
of a decrease in its conductivity. However samples with AI/B(5/95
calcined at T<600oC where found possessing the same conductivity as
pure BP04, in spite of their lesser solubility compared to boron
phosphate. This observation reinforces the view that a compromise
between high proton conductivity and high stability of boron phosphate is
feasible, and can be achieved, among other approaches, through
chemical modification of the solid.
EXAMPLE 20
Composite membrane comprising the
basic filler material PEI.
Polyetherimide (PEI) has been chosen as a further
component for preparation of polymer blend membranes. PEI is a high
temperature engineering thermoplastic, which is miscible with PEEK,
producing blends with increased toughness over both components3o.
Blending of sulfonated PEEK (SPEEK) with aminated polymers can lead
to formation of hydrogen bonds through proton, shared by sulfonic acid
and amino groups, and can thus affect the properties of the polymer. In
reference3' it was shown, that blending of SPEEK with PEI involves a
reduction of swelling in aqueous media and therefore brings about an
improvement in the mechanical properties of SPEEK membranes.
#11229.117/DEM.WPD
CA 02292703 1999-12-20
58
One way to build up the proton conductivity of an
organic polymer is to mix it with an inorganic acid. This approach was
commonly employed for modification of a wide variety of polymers.
However, as it follows from the review by Lassegues , it has met with
only limited success, as usually the obtained conductivities did not reach
high enough values. More recently doping with a liquid acid was shown
to improve essentially the electrical properties of PEM~ in some
instances. For SPEEK, which possesses a rather high intrinsic
conductivity, such a blending could be beneficial, allowing a further
increase in the conductivity of the membranes. The instant example is an
attempt to study the impact of blending with a basic filler such as PEI
and/or doping with HCI and H3P04 on the electrical properties of the
SPEEK based polymer membranes.
Experimental
Polyetherimide (PEI) Utem was supplied by General
Electric Co. in form of extrudates. Polyether ether ketone (PEEK) from
Polyscience Inc. was sulfonated at room temperature using concentrated
H2S04 (95-98%) according to a procedure previously described. The
duration of the reaction was varied from 24 to 112 h in order to increase
the sulfonation degree. H2S04, HCI, H3P04 and all solvents were of
reagent grade or better and were used as received.
Membrane preparation
The pure SPEEK membrane and blend SPEEK/PEI
membranes were prepared by the solution casting. The polymers, taken
in appropriate proportions, were dissolved at room temperature in
dimethylacetamide (DMAc) to form a 10% solution, which was then stirred
#11229.117IDEM.WPD
CA 02292703 1999-12-20
59
for 8 hours. After evaporation of most of the solvent the resulting blend
was cast onto a glass plate using a casting knife. The cast membranes
were dried at room temperature overnight and then for 4-6 h at 60oC and
for 12 h more at 80-120oC. The degree of sulfonation (D.S.) was
determined both by elemental analysis for sulfur and by titration using the
following procedure. 2-5 g of the SPEEK was kept in 1 M aqueous NaOH
for 1 day and then back titrated with 1 M HCI using phenolphthalein as
indicator. Doping with HCI and H3P04 was performed by soaking the
membranes in aqueous 3M acids for 3 weeks. The membranes were then
washed and stored in distilled water.
Electron Microscopy
The morphologies of the composite polymer membranes
were investigated using a scanning electron microscope (JSM-849,
JEOL). Specimens for the SEM were prepared by freezing the dry
membrane samples in liquid nitrogen and breaking them to produce a
cross-section. Fresh cryogenic fractures of the samples were vacuum
spray-coated with a thin layer of Au/Pd prior to viewing in SEM.
Water absorption of membranes
The water absorption of SPEEK membranes was
determined from the difference in weight (V1n between the dry and the
swollen membranes. The membrane, cast from bMAc solution after
drying, was weighed and then soaked in water until the weight remained
constant. It was then taken out, wiped with blotting paper and weighed
again. The percentage of water absorbed was calculated with reference
to the weight of the dry specimen: (Wwet/Vlldry-1 )x100%.
#11229.117/DEM.WPD
CA 02292703 1999-12-20
Conductivity measurements
The proton conductivity of the membrane samples was
measured by the AC impedance spectroscopy over the frequency range
of 1-107 Hz with oscillating voltage 50-500 mV, using a Solarton 1260
5 gain phase analyzer. A sample of the membrane with diameter 13 mm
was placed in an open, temperature controlled test cell, where it was
clamped between two blocking stainless steel electrodes with a
permanent pressure of about 3 kg/cm2. The disadvantage of an open cell
is that the specimen may sustain dehydration during the measurement.
10 However it allows to provide good electrode-specimen contact (by
applying sufficient thrust using an external load) and gives an access to
a larger temperature range (typically up to 150oC). Besides, the thin
specimen discs (100-500 ?m) are tightly compressed between blocking
electrodes, and can lose water only through their edges, which is
15 negligible at low temperature over the experimental time scale.
The conductivity ? of the samples in transverse direction
was calculated from the impedance data, using the relation ( = d/RS
where d and S are the thickness and face area of the sample
respectively, and R was derived from the low intersect of the high
20 frequency semi-circle on a complex impedance plane with the Re(Z) axis.
The impedance data were corrected for the contribution from the empty
and short-circuited cell.
Results
Sulfonation of PEEK up to D.S. below 60% yields
25 membranes with poor conductivity, typically of the order of 10-4 S/cm at
room temperature. Therefore, only membranes based on SPEEK with
#11229.117/DEM.WPD
CA 02292703 1999-12-20
61
higher D.S. are further studied and discussed. In Table 5, the
composition of the specimens are listed along with their thickness and
swelling in water at ambient temperature. The conductivity of membranes
untreated and doped with HCI and H3P04 is also indicated in the table.
u> > zzs. ~ ~ ~roeM.wPo
CA 02292703 1999-12-20
Table.J'r. Characterization of the membranes
62
PEI Thickness Water uptake,Conductivity,
of S/cm
PEEK D.S.,content, hydrated wt.%
wt.% membrane, at 25C at 100C
~m
65 7.9x 10'4 a) 1 x 10-3
a)
8
- 300 33 2.8x10-3 b) .
1.7x10'2
b)
3.2x10-3 ~) 2.4x10'2
~)
70 1.4x 10'3 a) 6x 10'3 a)
8
- 350 40 4.1x10'3 b) .
2.5x10-2
b)
5.7x10'3 ~~ 4.1x10-2
)
1.0x10-3 a) 1.3x10-2
a)
0 300 47 3.0x10-3 b) 5.7x10'2
b)
5.1 X 10 3 7.Ox 10'2
) c)
2 450 61 2.2x10-3 a) 1.9x10-2
a)
1.8x10-3 a) 2.3x10'2
a)
5 600 S1 4.2x10-3 b) 6.4x10-2
6)
72 7.1x10-3 ~) 8.3x10-2
)
9.1x10 a) 1.4x10-2
a)
1 S 400 42 2.9x l 0-3 3 .1 x 10'2
b) b)
4.1x10-3 ~) .3.0x10'2
)
7.9x10 a) 4.8x10-3
a)
25 250 31 1.0x10-3 b) 7.3x10'3
b)
3.1x10-3 ) 2.1x10-2
~)
a) non-doped
b) doped with H3P04
~) doped with HCl
CA 02292703 1999-12-20
63
SPEEK membranes doped with HCI and H3P04
As can be seen from Table 5, the room temperature
conductivity of untreated SPEEK membranes increased when D.S. is
raised from 65 to 70%. The sample with D.S.=72% displayed slightly
lower conductivity at room temperature, however at 1 OOoC it exhibited the
maximum value as compared to the other two membranes. Treatment
with the acids further increased the conductivity up to 2-3 times for
65%SPEEK and 3-5 times in the case of 70% and 72%SPEEK. In all
three cases doping with a stronger acid, HCI, resulted in a larger increase
of conductivity than when membranes were treated with less strong
H3P04. It is worth mentioning that the SPEEK specimens doped with the
acids were tested after storage in water at room temperature for several
month, which did not affect their conductivity. This testifies that HCI and
H3P04 form solutions with SPEEK and the polymer can retain them firmly
enough. Thus this method of enhancing the conductivity should be
applicable to actual electrochemical devices.
SPEEKIPEI membranes
The morphology of polymer blends of unsulfonated
PEEK with PEI was previously studied in ref.~°. It was shown that
phase
separation occurs during PEEK crystallization, which leads to PEI
segregation into amorphous phase between the PEEK texture units. The
sulfonated PEEK studied in the present work was fully amorphous, as
confirmed by X-ray diffraction. Nevertheless, phase segregation occurs
in a similar way in the SPEEK/PEI blend polymers, data not shown. The
microphotographs display the formation of small (less than 1?rn) spherical
particles of PEI at 5%PEI content and their growth up to ~1?m at 15%
#~ ~ 22s. ~ »roeM.wPo
CA 02292703 1999-12-20
64
PEI and to ~2-3 ?m at 25% PEI. The distance between PEI particles in
the SPEEK matrix apparently remains unchanged.
The analysis of the temperature dependence of
conductivity for blend membranes is depicted in Fig. 20. It can be seen
that the thermally stimulated enhancement of the conductivity slows down
above 80-90oC and for some specimens a little decay is observed above
110-120oC. This behavior suggests that some dehydration of the
samples occurs at the higher temperatures, reached in these
experiments. As all measurements were carried out with the same
heating rate of ~3o/min, these data can be used as an indication of the
capacity of the membranes to retain water.
Blending SPEEK with PEI affected the conductivity of
the blend membranes as can be seen from Fig.20. However the
dependence of conductivity on the PEI content in the blend membranes
is not monotonous. This follows from the fact that the curve of?? vs. T for
pure SPEEK in Fig. 20 occupies an intermediate position between the
curves, corresponding to the other specimens. The dependence of
conductivity on PEI content in blend membranes at 25 and 100oC
indicates that the conductivity passes .through the maximum at small PEI
content (ca 2-5%) and then decreases down to values below that of pure
SPEEK. This maximum was not detected in ref.3° because only blends
with PEI contents above 10% were studied there.
SPEEK/PEI membranes doped with HCI and H3P04
The influence of doping with the acids on the
conductivity of the SPEEK/PEI blend membranes can be followed from
Fig. 21. As in the case of pure SPEEK specimens, the conductivity of
#11229.117/DEM.WPD
CA 02292703 1999-12-20
blend membranes significantly increased with doping and again,
treatment with HCI generally produced larger enhancement than that with
H3P04. It is worth noting that the membrane, containing 5%PEI exhibited
the maximum conductivity, which was close to 10-1 S/cm despite
5 dehydration above 80oC.
In distinction to SPEEK, PEI does not contain any
protonated species and is commonly regarded as an effective insulator.
Blending with an insulator is expected to bring about a decrease in
conductivity, which turned out to be not true in the case of SPEEK/PEI
10 blend membranes, where the dependence of conductivity on PEI content
follows a curve with a maximum. This effect is however well known in the
art for inorganic solid electrolytes, where a drastic increase in conductivity
is often observed upon introduction of a dispersed insulating phase into
a pure ionic material.
15 Blending sulfonated PEEK with PEI results in an
increase in membrane water uptake at low PEI content, and a decreased
water absorption when the PEI concentration is above 5%. The proton
conductivity of blend membranes, both doped and non-doped with HCI
and H3P04 , expressed as a function of PEI content, follows the same
20 trend. Doping with the acids usually enhanced the membrane conductivity
by several times, and the effect of doping with HCI was generally more
significant than that with H3P04. The SEM study has shown that PEI
forms spherical particles, dispersed in the SPEEK matrix, while the DSC
analysis evidenced, that at the same time PEI partially dissolves in
25 SPEEK, affecting the swelling properties of the latter. This explains the
unusual behavior of the blend membrane with reference to water
#11229.117/DEM.WPD
CA 02292703 1999-12-20
66
absorption and conductivity. Indeed, the introduction of a small quantity
of PEI brings about an increase in the membrane capacity to absorb
water due to formation of new adsorption sites along the interface
between the two polymer phases. Further increase in PEI content results
in a reduced membrane swelling due to the increased fraction of PEI
dissolved in the SPEEK matrix
Thus, the introduction of PEI to an acidic polymer such
as PEEK in the composite membranes of the present invention could
enhance the efficiency and/or stability of these membranes.
EXAMPLE 21
Composite membrane comprising POD
Aromatic polyoxadiazole (POD) was tested as a material
for PEMFC. The PODs are known to be thermoplastics, possessing good
strength, fatigue resistance and hydrolytic stability [34-36]. It was
previously shown, that thin films based on aromatic PODs can be
electrochemically doped to obtain an electrically conducting polymer. It
was assumed that in the process of doping, the oxadiazole rings can act
as an electron acceptor forming stable anionic radicals. Despite the fact
that many procedures of POD synthesis have already been explored and
the properties of a number of POD derivatives have been studied [34-
36], still little is known about the electrical conductivity of this group of
polymers. The present study represents an attempt to study the electrical
conductivity of both doped and undoped PBO membranes as function of
the preparation procedure, mechanical treatment and temperature.
#11229.1171DEM.WPD
CA 02292703 1999-12-20
67
Synthesis of POD membranes
Membrane synthesis was performed according fo the
reaction route described in [35]. Pure POD was synthesized from
hydrazine sulfate and terephtalic acid in concentrated H2S04. TBI-POD
was prepared by the same procedure using tert-butyl-isophtalic acid
instead of terephtalic acid. Since the mechanical properties of TBI-POD
were somewhat better than pure POD, we prepared co-polymers,
containing both repeat units. Membranes were prepared from the polymer
solution in sulfuric acid by casting onto a glass plate and then immersion
into water for phase inversion. The membrane designation is presented
in Table 6, where the concentration of polymer solution in the acid before
phase inversion is presented along with the pressure values used for
membrane mechanical compression:
#11229.117/DEM.WPD
CA 02292703 1999-12-20
68
Table.. Preparation conditions, mechanical pressure applied during post-
treatment and thickness
of POD based membranes
Polymer Thickness,
pm
Sample POD TBI-POD concentrationPr
, , essure,
deli nationwt % vvt % in acid ton/cm2 undoped doped
wt % .
A8 25 75 8 0 S50 730
3.6 120 220
5.7 100 170
0 170 180
A 10 25 75 _ 3.4 I 00 120
10
5.7 40 60
B8 0 100 7.5 0 300 400
3.4 130
4.5 - 125
5.7 70 85
CA 02292703 1999-12-20
69
Conclusion
Polyoxadiazole based membranes were synthesized
and tested for electrical conductivity in both as-prepared and H3P04
doped forms. The undoped materials were found to exhibit a rather low
conductivity when conditioned in saturated water vapor, while the
membranes impregnated with phosphoric acid retained a high
conductivity of the order of 10-1 S/cm in air and after heating up to
150oC even after 1 month of storage. Taking into consideration that
Nafion shows a sharp decrease in proton conductivity above 100oC due
to dehydration, it is apparent that the POD based membranes are of a
distinctive interest for PEMFC application.
The mechanical properties of the membranes, prepared
using the phase inversion technique were found to be rather poor due to
the presence of large pores. However, it turned out to be possible to
control the porosity using mechanical compression, which eliminated
macropores and imparted better flexibility and strength to the polymer.
Thus, polyoxadiazole could be used in the composite
membranes described above and could contribute to its stability and
efficiency.
Although the present invention has been described
hereinabove by way of preferred embodiments thereof, it can be modfied,
without departing from the spirit and nature of the subject invention as
defined in the appended claims.
#11229.117/DEM.WPD
CA 02292703 1999-12-20
REFERENCES:
1. H.P.Dhar, On solid polymer fuel cells, J.EIectroanaLChem.,
357(1993)237.
5
2. A.J. Appleby, Recent developments and application of the polymer fuel
cell, PhiLTrans.Royal Soc. London, A 354(1996)1681.
3. J.P.Shoesmith,R.D.Collins, M.J.Oakley, D.K.Stevenson, Status of solid
10 polymer fuel cell system development, J.Power Source, 49(1994)129.
4. K.D. Kreuer, Proton conductivity: Materials and Applications, Chem.
Mater., 8(1996)610.
15 5. G.Pourcelly and C.Gavach, Pertluorinated membranes, in Ph.
Colomban (Ed.), Poton Conductors: Solids, membranes and
gels-materials and devices, Cambridge University Press, New York, 1992,
pp 294-310.
20 6. A.E.Steck and C.Stone, Membrane materials in fuel cells, 2nd Int.
Symp. on New Materials for Fuel Cell and Modern Battery Systems,
Montreal, Canada, July 1997, p.792.
7. O.Savadogo, Emerging membranes for electrochemical systems: (I)
25 solid polymer electrolyte membranes for fuel cell systems, J.New
Mat.Electrochem.Syst.,1 (1998) 66.
#11229.117/DEM.WPD
CA 02292703 1999-12-20
71
8. T. Kobayashi, M.Rikukawa, K.Sanui, N.Ogata, Proton conducting
polymers derived from poly(ether-ether ketone) and
poly(4-phenoxybenzoyl-1,4-phenylene), Solid State tonics, 106 (1998)
219.
9. J. Kerres, A. Ullrich and Th.Haring, New ionomer membranes and their
fuel cell application:1. Preparation and characterization, 3rd Int. Symp..
on New Materials for Electrochemical Systems, Montreal, Canada, July
1999, p.230.
10. S.D.Mikhailenko, S.Kaliaguine, J.B. Moffat, Electrical Impedance
studies of the ammonium salt of 12-tungstophosphoric acid in the
presence of liquid water, Solid State tonics, 99 (1997) 281.
11. O.Nakamura and LOgino, Electrical conductivities of some hydrates
of dodecamolybdophosphoric acid and dodecatungstophosphoric acid
and their mixed crystals, Mater. Res.Bull.,17 (1982) 231.
12. S.D.Mikhailenko, S.M.J.Zaidi and S.Kaliaguine, Development of
zeolite based proton conductive membranes for use in direct methanol
fuel cells, Report submitted to Natural Resources Canada, Ottawa, 1997.
13. X. Jin, M.T.Bishop, T.S. Ellis and F.E.Karasz, A sulfonated
poly(aryl-ether-ketone), British Polym. J., 17 (1985) 4.
#11229.117/DEM.WPD
CA 02292703 1999-12-20
72
14. C. Bailly, D.J. Williams, F.E. Karasz and W:J. McKnight, The sodium
salts of sulfonated poly(aryl ether ether ketone) (PEEK): preparation and
characterization, Polymer, 28(1987)1009.
15. M.T.Bishop, F.E.Karasz, P.S.Russo and K.H. Langley, Solubility and
properties of poly(aryl-ether-ketone) in strong acids, Macromolecules, 18
(1985) 86.
16. R.Nolte, K.Ledjeff, M.Bauer, and R. Mulhaupt, Partially sulfonated
poly(arylene ther sulfone)- A versatile proton conducting membrane
material for modern energy onversion technologies, J.Membrane Sci.,
83(1993) 211-220.
17. M. Drzewinkski and W.J. Macnight, Structure and properties of
sulfonated polysulfone lonomers, J.Appl. Polym.Sci., 30 (1985) 4753.
18. J.F.O'Gara, D.J.Williams, W.J.Macnight and F.E. Karasz, Random
homogeneous sodium sulfonate polysulfone lonomers: preparation,
characterization, and blend studies, J.Polym. Sci. B: Polym. Phys, 25
( 1987) 1519.
19. A.Noshay and L.M. Robeson, Sulfonated polysulfone, J.Appl. Polym.
Sci, 20 (1976) 1885.
20. N.Sivashinsky and G.B. Tanny, Ionic heterogeneities in sulfonated
polysulfone films, J.Appl. Polym. Sci, 28(1983)3235.
#11229.117/DEM.WPD
CA 02292703 1999-12-20
73
21. M. Day, J.D. Cooney and D.M. ales, Degradation of
poly(aryl-ether-ether-ketone), PEEK as monitored by pyrolysis-GC/MS,
TG/MS, J.Anal. & Appl. Pyrolysis, 18 (1990) 163.
22. S.Faure, N.Cornet, G.Gebel, R.Mercier, M.Pineri, and B.Sillon,
Sulfonated polyimide as novel proton exchange membranes for H2/02
fuel cell, 2nd Int. Symp. on New Materials for Fuel Cell and Modern
Battery Systems, Montreal, Canada, July 1997, p.818.
23. M.W.Park, J.C.Yang, H.S.Han, Y.G.Shul, T.H.Lee, and Y.Cho,
Heteropolyacid(H3PW12O40) incorporated solid polymer electrolyte for
PEMFC, Denki Kagaku, 64,6(1996)743-748.
24. B.Tazi and O. Savadogo, Preparation and characterization of a new
membranes based on Nafion, silicotungstic acid and thiophene, 3rd Int.
Symp. on New Materials for Electrochemical Systems, Montreal, Canada,
July 1999, p.259.
25. K.D. Kreuer, T. Dippel, and J. Maier, Membrane materials for
PEM-Fuel cells: A microstructural approach, Proc. Electrochem. Soc., 95
(1995) 241.
26. Y.Sone, P.Ekdunge and D. Simonsson, Proton conductivity of Nafion
117 as measured by a four-electrode AC Impedance method,
J.Electrochem. Soc., 143 (1996) 1254.
#11229.117/DEM.WPD
CA 02292703 1999-12-20
74
27. N.Knudsen; E.K.Andersen, LG.K.Andersen, P.Norby, E.Skou, Solid
State tonics, 61, 153(1993).
28. G.L.Tumer, K.A. Smith, R.J.Kirkpatrick, E.Oldfield, J.Magn.Res., 67,
544(1986)
29. G.Engelhardt, D.Michel, "High resolution solid-state NMR of silicates
and zeolites", J.~ley & sons, New York, 1987, p.332.
30. Hudson, S.D.; Davis, D.D.; Lovinger, A.J. Macromolecules 1992, 25,
1759.
31. Cui, W.; Kerres, J.; Eigenberger, G. Separation and Purification
technology 1998, 14, 145.
32. Lassegues, J.C. In Proton Conductors: Solids, membranes and
gels-materials and devices; Colomban, Ph., Ed.; Cambridge University
Press: New York, 1992, pp 311-328.
33. Zaidi, S.M.J.; Chen, S.F.; Mikhailenko, S.D.; Kaliaguine, S. Proc.3rd
Int. Symposium on New Materials for Electrochemical Systems, Montreal,
Canada, July 1999, pp 225-227.
34. Wainright, J.S.; Wang, J-T.; Weng, D.; Savinell, R.F., Litt, M. J
Electrochem Soc 1995,142,121.
#11229.117/DEM.WPD
CA 02292703 1999-12-20
35. E. R. Hensema, J. P. Boom, M. H. V. Moulder and C. A. Smoulders,
J. Polym. Sci: Polym. Chem., 32, 513(1994).
36. E. R. Hensema, M. E. R. Sena, M. H. V. Moulder and C. A.
5 Smoulders, J. Polym. Sci: Polym. Chem., 32, 527(1994).
37. B. Schulz, L. Brehmer, B. Dietzel and Th. Zetzshe, Reactive &
Functional Polymers, 30,353(1996)
#11229.117/DEM.WPD
