Handbook of Polymer Synthesis Second Edition Episode 7 pptx

7
Aromatic Polyethers
Hans R. Kricheldorf
Institute of Technical and Molecular Chemistry, University of Hamburg, Hamburg, Germany
I. INTRODUCTION
Aromatic polyethers are a group of high-performance engineering plastics which were
developed by several chemical companies over the past forty years. Poly(phenyl ene ether)s
mostly prepared by oxydative polycondensation of phenols, were the first class of aromatic
polyethers which was technically produced and commercialized [1–3]. Perfectly linear
poly(arylene ether)s free of side chains can only be obtained by oxidative coupling of
2,6-disubstituted phenols. Therefore, the poly(2,6-dimethylphenylene-oxide), usually
called PPO, is the most widely used poly(arylene ether). Its outstanding property is its
compatibility with polystyrene (and a few other polymers) so that it is mainly used as
component of blends [4]. The oxidative coupling of phenols has intensively been studied
between 1950 and 1980 (as discussed in the 1st edition of this handbook) but only few
research activities were observed during the past 10 years discussed in the subchapter
‘Various Aromatic Polyethers’. Therefor e the literature evaluated and discussed below
mainly concerns poly(ether sulfone)s and poly(ether ketone)s. Furthermore, semi aromatic
polyethers and aromatic polysulfides are included in this chapter which mainly covers the
literature of the years 1990 through spring 2000 (complementary to the first edition).
The characteristic properties and advantages of aromatic polyethers when compared
to aliphatic engineering plastics based on an aliphatic main chain are as follows. Aromatic
polyethers are less sensitive to oxydati on at all temperatures, and thus, also less
inflammable (poly(vinyl chloride) and poly(tetrafluoroethylene) are, of course, exceptions
in the case of aliphatic polymers). The thermostability of aromatic polyethers is higher.
For all these reasons the maximum service temperature of aromatic polyethers (200–
260

C) is twice as high as that of aliphatic polymers. Furthermore, aromatic polyethers
possess higher glass-transition temperatures (T
g

s) which may be as high as 230

C for
commercial poly(ether sulfone)s. The T
g
s of commercial poly(e ther-ketone)s are lower
(typically around 140–150

C) but all commercial poly(ether-ketone)s are semicrystalline
materials having melting temperatures in the range of 280–420

C. Either due to high T
g
or due to a high T
m
the heat distortion temperature of aro matic polyethers significantly
higher than that of aliphatic polyme rs. A few characteristic disadvantages should also
be mentioned. Most poly(ether-sulfone)s reported so far and all commercial examples
are amorphous with the advantage of a high transparency and the shortcoming of a high
sensitivity to the attack of organic solvents. The crystalline poly(ether ketone)s are rather
Copyright 2005 by Marcel Dekker. All Rights Reserved.
insensitive to organic solvents, but they are sensitive to a cleavage by UV-irradiation
quite analogous to low molar mass benzophenones. Finally, it should be mentioned that
a typical application of poly(ether sulf one)s and poly(ether-ketone)s is that as matric
material in composites. Glas s fiber or carbon fiber are used as reinforcing components.
II. POLY(ETHER SULFONE)S
Poly(ether-sulfone)s, PESs, may in principle be prepared via four different strategies:
1. Polycondensation of suitable monomers involving an electrophilic substitution
of an aromatic ring.
2. Polycondensation of suitable monomers involving a nucleophilic substitution of

a chloro, fluoro or nitroaromat activated by a sulfonyl group in para-position.
3. Chemical modification of suitable precursor polymers.
4. Ring-opening polymerization of cyclic oligo(ether-sulfone)s
The discussion of synthetic methods and structures presented below will follow this
order.
A. Syntheses via Electrophilic Substitution
The oldest approach known for the preparation of PESs is a polycondensation process
involving the electrophilic substitut ion of a phenyl ether group by an aromatic sulfonyl
chloride group [3,5,6]. Such polycondensations may be based, either on monomers
containing both functional groups in one molecule (Eq. (1)) or by a combination of a
nucleophilic and an electrophilic monomer (Eq. (2)). These polycondensations need to be
catalyzed by strong Lewis acids such as FeCl
3
, AlCl
3
,orBF
3
. Characteristic disadvantages
of this approach are the need of an expensive inert reaction medium and side reactions
such as substitution (including branching) in ortho position of the nucleophilic monomer.
Furthermore, this approach is not versatile and limited to a few monomers. Previous
research acti vities in this field were report ed in the 1st edition of the Handbook [3], but
more recently new activities were not observed.
ð1Þ
ð2Þ
Copyright 2005 by Marcel Dekker. All Rights Reserved.
B. Syntheses Via Nucleophilic Substitution
The most widely used approach to the preparation of PESs in both academic research and
technical production is a polycondensation process involv ing a nucleophilic substitution
of an aromatic chloro- or fluorosulfone by a phenoxide ion (Eq. (3)). Prior to the review

of new PESs prepared by nucleophilic substitution publications should be mentioned
which were concerned with the evaluation and comparison of the electrophilic reactivity of
various mono- and difunctional fluoro-aromats [7–10]. The nucleophilic substitution of
aromatic compounds may in general proceed via four different mechanism. Firstly, the S
N1
mechanism which is, for instance, characteristic for most diazonium salts. Secondly, the
elimination-addition mechanism involving arines as intermediates which is typical for the
treatment of haloaromats with strong bases at high temperature. Thirdly, the addition–
elimination mechanism which is typical for fluorosulfones as illustrated in equations (3)
and (4). Fourthly, the S
NAR
mechanism which may occur when poorly electrophilic
chloroaromats are used as reaction partners will be discussed below in connection with
polycondensations of chlorobenzophenones.
ð3Þ
ð4Þ
In the case of the addition–elimination mechanism the addition step with the
formation of a short lived Meisenheimer complex (Eq. (3)) is the rate determining step.
Hence, the electron density of the carbon directly bound to the fluorine (ipso position) is
decisive for the reactivity and thus, for the rate of the reaction. In two publications [7,8]
the
13
C NMR chemical shifts of various fluoroaromats were determined, compared and
shown to be useful indicators of the reactivity of the ipso-carbon. This conclusion was
confirmed by calculation of the electron density via the quantum semiempirical PM 3
method in the MOPAC software. In fact, a linear correlation between the calculated
electron density and the
13
CNMRd values was obtained. Furthermore, the
19

F NMR
chemical shifts were determined for numerous electrophilic fluoroaromats and again
a linear correlation with the calculated electron densities, on the one hand, and with the
13
C NMR chemical shifts, on the other, was found [7,8]. These studies proved that the SO
2
group is the strongest activating divalent group. Only the monovalent nitrogroup has
a stronger electron-withdrawing effect. The strong electron-withdrawing effect of the SO
2
group has also the consequence that the C-atom directly attached to it is sensitive to a
nucleophilic attack. With KF as reagent the cleavage of the PES backbone (back-reaction
of Eqs. (3) and (4) was observed at 280

C [9], but it is not clear if the cleavage will be more
favored by other cations such as Cs

. Finally, a publication should be mentioned [10]
reporting on a partial desulfonylation during the polycondensation of a special ketone-
sulfone type monomer.
The standard procedure used by most authors for syntheses of new poly(ethersul-
fone)s is based on the reaction of equimolar amounts of a difluoro (or dichloro)- sulfone
Copyright 2005 by Marcel Dekker. All Rights Reserved.
and a bisphenol with dry K
2
CO
3
(equimolar or slight excess) in polar aprotic solvents such
as N-methylpyrrolidone (NMP), dimethylacetamide (DMAc) dimethylsulfoxide (DMSO)
or sulfolane. In the paper [11] stoichiometric amounts of CsF were applied instead of
K

2
CO
3
. However, CsF has no advantage, but it is significantly more expensive. Following
the standard procedure with K
2
CO
3
two research groups used commercial 4,4
0
-
dichlorodiphenylsulfone (DCDPS) for the preparation of the PESs (5a) [12] and (5b)
[13]. The DCDPS was also taken as electrophilic reagent for the preparation of the
functionalized oligo(ether-sulfone)s which were modified at the chloro endgroups (6) [14].
Another class of functional PES (7) was synthesized from commercial 4,4
0
-Difluorodi-
phenylsulfone (DFDPS) and 1,1-bis(4-hydroxydiphenyl)ethene [15]. The pendant methy-
lene group allows for thermal crosslinking of these PESs. DFDPS in combination with
various diphenols and 4-fluoro-4
0
-hydroxydiphenylsulfone served as comonomers for
the preparation of copoly(ether-sulfone)s having the structure (8). Their properties were
evaluated and correlated with their composition and sequence [16].
ð5Þ
ð6Þ
ð7Þ
ð8Þ
Copyright 2005 by Marcel Dekker. All Rights Reserved.
Most new PESs reported during the past ten years were prepared from new sulfone

type monomers or at least from noncommercial monomers. For instance, the
difluoroketone-sulfone (9a) was polycondensed with the dihydrodiphenylketone-sulfone
(9b) [17]. The same monomers were later used by another resear ch group together with
a variety of new fluoro ketone type monomers [19]. Other authors synthesized the
naphthalene containing difluordiphenylsulfones (10a) and (10b) as reaction partners of
methyl substituted 4,4
0
-dihydroxybiphenyls [19]. Dichloro- or difluoro- diphenylsulfones
having a central biphenyl unit (11a) were polycondensed with various commercial
diphenols [20,21]. In one of these papers [20] PES derived from the bisphenol (11b) were
studied in detail.
ð9Þ
ð10Þ
ð11Þ
Thiophene based poly(arylene-ether-sulfone)s were synthesized from monomer (12)
and bisphenol-A [22]. Another class of unconventional monomers is outlined in the
formulas (13a) and (13b) [23,24]. These monomers have the advantage that its structure
can easily be varied at the imide ring, and reactions at the imide ring allow also a
modification of the PES itself. In any case these terphenyl monomers co nsiderably raise
the glass transition temperature (Tg). Another kinked structure is that of the indan
derivatives (14) [25]. They were polycondensed with bisphenol-A and other common
diphenols. Two research groups reported on syntheses of more or less fluorinated PESs.
Two different synthetic strategies were elaborated. In the first case a difluoro diphenyl
disulfone having a fluorinated aliphatic chain segment (15b) was synthesized by oxidation
of the corresponding disulfide (15a) [26]. The disulfone (15b) was then polycondensed with
Copyright 2005 by Marcel Dekker. All Rights Reserved.
a variety of diphenols by means of sodium carbonate in DMAc. The second strategy is
characterized by the preparation of PESs having pendant trifluoromethyl groups by
polycondensation of the monomers (16a), (16b) and (16c) [27,28].
ð12Þ

ð13Þ
ð14Þ
ð15Þ
ð16Þ
Copyright 2005 by Marcel Dekker. All Rights Reserved.
C. Chemical Modification
The normal route of nucleophilic substitution was also applied to syntheses of PESs
having a broad variety of pendant functional groups. In most publications dealing with
functional PES the reactive substi tuents were subjected to further modifications. Two
research groups were inter ested in sulfonated PES which may find potential application as
proton transporting membranes in fuel cells. Two different synthetic approaches were
explored. The first one is based on polycondensations of a sulfonated DCDPS (17a) with
preformed potassium salts of diphenols in DMAc at 170

C [29]. The second approach
consists of the sulfonation of preformed PESs [30,31]. When a PES derived from
hydroquinone was sulfonated exclusive monosulfonation of the hydroquinone unit was
observed (18a). Increasing reactivity of the sulfonating agent did not influence the degree
of substitution but the stability of the PES chain. With 91% sulfuric acid no degradation
was observed at room temperature, whereas chlorosulfonic acid and oleum caused severe
degradation. Fur thermore, PES derived from methyl hydroquinone, dimethylhydro-
quinone and trimethylhydroquinone were sulfonated with concentrated sulfuric
acid. Complete monosubstitution was found for mono- and dimethyl hydroquinone
(18b) and (19a), whereas the sulfonation of the trimethyl hydroquinone units (19b)
remained incomplete [31].
ð17Þ
ð18Þ
ð19Þ
Copyright 2005 by Marcel Dekker. All Rights Reserved.
Several studies dealt with syntheses of PESs having pendant amino groups. Quite

analogous to the syntheses of sulfonated PESs two strategies were explored, (a)
polycondensation of aminated monomers, and (b) modification of preformed PES. The
first strategy was realized with synthesis and polycondensation of the dichlorodiamino-
sulfone (20a) which was synthesized by hydrogenation of 2,2
0
-dinitro-4,4
0
-dichlorodiphe-
nylsulfone [32]. Another research group used the corresponding difluorodiaminosulfone
[33]. A further difluoromonomer having a pendant amino group is the phosphine oxide
(20b) which was used as comonomer together with DFDPS and various diphenols [34].
The second synthetic strategy was realized in such a way that performed PESs were
nitrated at the hyd roquinone unit and the nitrogroup was reduced by means of sodium
dithionite (21) [35]. Another approach is based on the synthesis of PES, having pendant
imide groups (22) [36–38]. Variation of the amine (for instance via transimidization) allows
a broad variation of the pendant functional groups including the introdu ction of an amino
group.
ð20Þ
ð21Þ
ð22Þ
Five more papers reported on various modifications of PES involving introduction
and substitution of chloro or bromoatoms. For instance, bromination of the bisphenol-A
unit in a commercial PES yield ed the dibromoproduct (23), which was treated with
butyllithium. The lithiated PES was then reacted with methyliodide [39] or with tosylazide
[40]. The resulting azide groups were finally reduced to amino groups (24). Another
modification of brominated PESs utilized palladium complexes as catalysts for the
Copyright 2005 by Marcel Dekker. All Rights Reserved.
introduction of alkin-type substituents (25). These substituents have the purpose to enable
a thermal cure via cyclization or polymerization of the alkin groups [41].
ð23Þ

ð24Þ
ð25Þ
Two papers reported on the chloromethylation of PESs and the further modification
of the chloromethyl groups. In the first paper [42] the combination of octylchloromethyl
ether and SnCl
4
was used to introduce the CH
2
Cl groups, the combination of
octylbromomethyl ether and SnBr
4
yielded CH
2
Br groups (26a) and combinations of
chloromethylether þ SnBr
4
or bromomethyl ether and SnCl
4
producing a statistical array
of chloro and bromomethyl substituents. Reactions with potassium tert butoxide yielded
pendant tert butyl ether group s (26b) with sodium acetate pendant acetate groups were
obtained and after alkaline saponification CH
2
OH groups (26b). Furthermore, pendant
tosylate groups (27a) and diethylphosphonates (27b) were prepared. With sodium cyanide
pendant nitrile groups were formed (28a) which were saponified to yield CH
2
CO
2
H

Copyright 2005 by Marcel Dekker. All Rights Reserved.
groups. Finally the oxidation of chloromethyl groups with dimethylsulfoxid/NaHCO
3
or with Cr
2
O
2
7
was studied (yielding aldehyde groups (28b) [42]. In the second paper
triflicacid was used as solvent and catalyst in combination with butyl or octyl chloro-
methyl ether. This system is of course too expensive for any large scale experiments
or technical production of functionalized PES. For homo- or copolyether containing
hydroquinone an exclusive monosubstitut ion of the hydroquinone unit was found (29a),
and finally the transformation of the chloromethyl groups into triethylammonium
groups (29b) was studied [43]. PESs having pendant aldehyde groups were prepared
from (co-)polycondensations of the diphenol (30a). The aldehyde groups were almos t
quantitatively transformed into azomethine groups (30b) [44]. In another publication [45]
unsaturated PESs were prepared from 4,4
0
-dihydroxy-trans-stilbene and DFDPS and
treated with H
2
O
2
in the presence of a tungsten catalyst whereby epoxide groups suitable
for chemical or thermal crosslinking were obtained (31). Finally, a publication dealing
with the influence of energy rich irradiation (x-ray, electrons, AR

and N
2


) on PES
should be mentioned [46].
ð26Þ
ð27Þ
ð28Þ
Copyright 2005 by Marcel Dekker. All Rights Reserved.
ð29Þ
ð30Þ
ð31Þ
D. Various Synthetic Methods
Numerous publications describe syntheses and characterization of telechelic oligo(ether
sulfone)s which served as building blocks of triblock copolymers, multiblock copolymers
or networks. OH-terminated oligomers (32) were prepared by polycondensations of
DCDPS with an excess of bisphenol-A [47–49]. These oligo(ether sulfone)s were reacted
with commercial bisepoxides to yield epoxy networks [47]. They also proved to be useful
for syntheses of multiblock poly(ether-esters) [48,49]. The polyester blocks either consisted
of poly(ethylene terephthalate) or of LC-poly(ester-imide)s (33). The LC-blocks played the
role of a reinforcing component in the PES matrix and showed interesting mechanical
properties. Telechel ic poly(ether-sulf one)s having C–F endgroups were also prepared
by co polycondensation of 4-fluoro-4
0
-trimethylsiloxy-diphenylsulfone with small amounts
of DFDPS [50]. The molecular weight was controlled by the feed ratio of DFDPS.
When small amounts of 4,4
0
-bis(2,4-difluorobenzoyl)diphenyl ether were used as como-
nomers, four armed stars with C–F endgroups were obtained. Small amounts of silylated
1,3,5-trihydroxybenzene as comonomer yielded three-armed stars having OSiMe
3

end-
groups [50].
Copyright 2005 by Marcel Dekker. All Rights Reserved.
ð32Þ
ð33Þ
Several research groups prepared oligo(ether-sulfone)s having primary amino
endgroups [51–55]. m-Aminophenol, p-aminophenol or a higher aminophenol (34)
served as endcapping agents. These oligomeric diamines were polyco ndensed with various
aromatic dicarboxylic acid dichlorides to yield polyamides [53] or they were poly-
condensed with bisanhydrides yielding polyamides such as (35) [54]. Poly(ether-sulfone-
amide)s were also prepared by the invers e approach [56]. In this case two ‘sulfone
dicarboxylic acids’ (36) were synthesized and polycondensed with various aromatic
diamines via the triphenylphosphite pyridine method.
ð34Þ
ð35Þ
ð36Þ
Triblockcopolymers derived from a central block of PES and two wings of
poly(phenylene oxide) were also reported [57]. For this purpose oligo(ether sulfone)s were
synthesized with an excess of DFDPS, so that two C–F endgroups were obtaine d (37).
These oligomers were then polycondensed with poly(phenylene oxide)s having one
trimethylsiloxy endgroup whereby CsF served as catalyst (38). Multiblock copolymers
Copyright 2005 by Marcel Dekker. All Rights Reserved.
consisting of PES and polydimethylsiloxane blocks were prepared from OH terminated
oligo(ether sulfone)s and diethylamine terminated siloxane blocks (39,40) [58]. The same
approach was reported to yield the poly(ether sulfone disilane)s (41) starting from a
bis(diethylamino)disilane [59]. These polysilanes deserve interest because of their
photosensitivity. Grafting of polysiloxane blocks onto PES (based on bisphenol-A) was
achieved in such a way that the PES was lithiated with nBuLi and reacted with
chlorodimethylvinylsilane [60]. The Si-H endgroup of a polysiloxane was then added onto
the pendant vinyl group (42,43). In this connection a work describing the radical grafting

of styrene onto maleimide or nadimide endcapped oligo(ether sulfone)s should be
mentioned [61]. In this way PES reinforced polystyrene foams with open pores were
obtained. In two papers oligo(ether sulfone)s and oligo(etherketone)s having two
acetylenic endgroups were described [62,63]. These oligomers (45) designed to yield
thermostable networks upon thermal cure were prepared by means of the new endcapping
agents (44).
ð37Þ
ð38Þ
ð39Þ
ð40Þ
Copyright 2005 by Marcel Dekker. All Rights Reserved.
ð41Þ
ð42Þ
ð43Þ
ð44Þ
ð45Þ
An entirely new synthetic approach is based on the oxidative coupling of
4,4
0
-diphenoxydiphenylsulfones (46a) or 4,4
0
-diphenylsulfidodiphenylsulfone (46a),
(R
1
–R
4
¼ H) [64]. These studies were extended to bisnaphthyloxysulfones of structure
(47) [65]. This approach involves a radical-cationic mechanism resulting from the single
electron transfer reaction (SET) of a p-electron from the aromatic monomers to FeCl
3

which plays the role of oxidation and coupling catalyst (so-called Sholl reaction).
Nitrobenzene served as the standard solvent and the temperature was varied from 20 to
100

C, but the molecular weights remained low (M
n
< 3000 Da) for all monomers of
structure (46a), whereas high molecular weights (M
n
 38,000 Da) were obtained from
the polycondensations of monomers (47). Finally, a new polycondensation method
should be mentioned yielding poly(ketone sulfone)s free of ether groups (48) and (49) [66].
Copyright 2005 by Marcel Dekker. All Rights Reserved.
ð46Þ
ð47Þ
ð48Þ
ð49Þ
E. Ring-Opening Polymerization (ROP)
All synthetic strategies discussed above have in common to be step growth polymeriza-
tions. Over the past ten years a new strategy was elaborated and explored based on the
ring-opening polymerization of cyclic oligo(ether sulfone)s, OESs. This chain growth
Copyright 2005 by Marcel Dekker. All Rights Reserved.
polymerization has the following advantages and disadvantages when compared to step-
growth polymerizations. The main problem is the synthesis of the cyclic monomers, above
all, when large quantities are needed. The advantages are, firstly, a polymerization process
which does neither need solvents, nor produce byproducts. Therefore, the ROP approach
is well suited for the reaction-injection molding (RIM) technology. Secondly, the ROP of
strained cyclic OESs offers the chance to prepare PESs with very high molecular weight
(M
n

> 10
5
Da). Thirdly, sequential copolymerizations with other cyclic monomers may
yield a variety of block copolymers.
Cyclic OESs were prepared in four different ways. Firstly, an electrophilic acylation
of 1,4-bisphenoxybenzene was performed under high dilution (50) [67]. Secondly,
1-chloro-4
0
-hydroxy diphenylsulfone was dimerized and cyclized in the presence of
K
2
CO
3
(51) [68]. Thirdly, several cyclic OESs were prepared by condensation of diphenols
and dihalosulfones via nucleophilic substitution under high dilution (52) [69–75]. Either
mixtures of cyclic OES were isolated and used for ring-opening polymerizations [69,70] or
monodisperse cycles were isolated and characterized [71–75]. Fourthly, preformed PES
was subjected to back-biting degradation catalyzed by CsF in DMF at 155

C. At low
concentrations, large fractions of cyclic OES were obtained and monodisperse cycles (from
the dimer to the hexamer) were isolated by column chromatography [76,77]. Two papers
[73,75] describe detailed studies of ring-opening polymerizations conducted in bulk at high
temperatures. Unfortunately, cyclic OESs possess high melting temperatures (up to
500

C), and only in cases of nonsymme trical cycles the reaction temperatures could be
lowered to 290

C. Such high temperatures have two disadvantages. Firstly, high fractions

of cycles remain unreacted for thermodynamic reasons, and gel particles are formed due to
partial crosslinking. Anionic polymerizations in concentrated solutions below 250

C have
not been studied yet.
ð50Þ
ð51Þ
Copyright 2005 by Marcel Dekker. All Rights Reserved.
ð52Þ
Finally, a synthetic approach should be mentioned yielding polyamides containing
cyclic OESs as part of the repeating unit [78]. The cyclic dicarboxylic acid (53c) was
prepared from (53a) and (53b) by a conventional procedure and polycondensed with
4,4
0
-diamino-diphenylmethane using the triphenylphosphite-pyridine method.
ð53Þ
Copyright 2005 by Marcel Dekker. All Rights Reserved.
III. POLYETHERKETONES
Most research activities in the field of aromatic polyethers published over the past ten
years concern poly(e therketone)s (in this review the abbreviation PEK is used for all
poly(etherketone)s, not only for those having one ether and one keto group in the
repeating unit). Quite analogous to PESs the synthetic methods reported for PEKs may be
subdivided into three groups:
1. polycondensations involving an electrophilic substitution (i.e., acylation of a
phenoxy group)
2. Polycondensations involving a nucleophilic substitution of a chloro-, fluoro- or
nitro-aromat activated by a keto group in para position)
3. modification of suitable precursor polymers.
A. Syntheses Via Electrophilic Substitution
Various PEKs were prepared via electrophilic substitution processes such as that

exemplarily outlined in equation (54) [79]. The problems of this approach are in principle
the same as in the case of PESs. An inert expensive solvent is needed, it is difficult to reach
high conversions without side reactions and the number of useful monomers is lower than
in the case of syntheses based on nucleophilic substitution reactions. The electrophilic
polycondensations may be subdivided into two different methods. Firstly, acid chlorides
are used as electrophilic monomers in combination with a Lewis acid. Secondly, free
carboxylic acid served as monomers in combination with an acidic dehydrating agent.
None of the polycondensation methods described in this section is new, and origin and
early exploration of these methods has been reviewed in the 1st edition of this handbook
(Chapter 9).
ð54Þ
In a publication of 1988 [79] (not reviewed before) polycondensations of phenoxy-
benzoyl chloride (Eq. (54)) or polycondensations of diphenylether with isophthaloylchlor-
ide, terephthaloyl chloride and a phenolph thalein dicarboxylic acid dichloride (55) were
studied. AlCl
3
served as catalyst and the solvent was varied. It was found that the
homogeneous polycondensation in nitrobenzene gave lower molecular weights than the
heterogeneous reaction in CH
2
Cl
2
or (ClCH
2
)
2
, (DCE). The chain growth continues in
the precipitated AlCl
3
-oligomer or -polymer complexes. Similar results were found by

another group [80] which did not know the first paper [79]. Other authors [81], compared
two reaction media: AlCl
3
/CH
2
Cl
2
and H
2
F
2
/BF
3
. The faster reaction (but similar mol-
weights) were found in the H
2
F
2
system. The influence of HCl or Lewis bases on AlCl
3
catalyzed polycondensatins of diphenylether and terephthaloylchloride was also studied
[81]. New structures were obtained by AlCl
3
catalyzed polycondensations of the oligo-
ethers (56a,b) with isophthaloylchloride or adipoylchloride [82]. Two other research groups
studied AlCl
3
catalyzed polycondensations of isopht haloyl chloride, terephthaloyl chloride
or naphthalene-1,6-dicarbonyl chloride with diphenyl ether or with the oligoether (57)
in much detail [83–88]. Potential defects in the chemical structure (as revealed by

1
H and
Copyright 2005 by Marcel Dekker. All Rights Reserved.
13
C NMR spectroscopy) and the morphology of the particles which crystallized directly
from the reaction mixture were intensively investigated [84–88].
ð55Þ
ð56Þ
ð57Þ
Polycondensations of dicarboxylic acids with dehydrating agents may be subdivided
into two methods, either a solution of P
4
O
10
in methanes sulfonic acid (Eatons reagent)
[89–93] or neat triflic acid were used [94–97]. By means of Eatons reagent 3-phenoxyenzoic
acid was polymerized (58) [89], and the dicarboxylic acids (59a–d) were polycondensed
with the oligoethers (60a,b) [90]. Furthermore a small series of polyethers were reported
having alternating sequences of keto, ether and sulfone groups between para func-
tionalized benzene rings [10,91]. PEK’s containing CF
3
groups were prepared from the
dicarboxylic acid (61a) and diphenyl [92], and PEKs containing phosphazene rings were
synthesized analogously from (61b) [93]. All these polycondensations were conducted at
80–120

C and gave only low (M
n
< 5000 Da) to moderate (M
n

< 10,000 Da) molecular
weights. Using triflic acid as catalyst and reaction medium the polycondensation of
terephthaloyl chloride with diphenyl ether was studied [94]. The most interesting result of
this study is the finding that addition of triflic anhydride or P
4
O
10
enhances the molecular
weight by a factor of 4 or 5. Another group conducted systematic studies of structure
reactivity relationships of various monomers or model compounds in triflic acid catalyzed
(poly)condensations [95]. Selected monomers used in this study are presented in the
formulas (62a,b) and (63a,b). This work was extended to cocondensations of the
carborane monomers (64a,b) with monomers of the structure (62a,b) [96]. The physical
and thermal properties of the carborane containing PEKs were also studied. Finally
ferrocene containing PEKs (65) should be mentioned which were prepared from ferrocene
dicarboxylic acid by means of Eatons reagent or triflic acid [97].
ð58Þ
Copyright 2005 by Marcel Dekker. All Rights Reserved.
ð59Þ
ð60Þ
ð61Þ
ð62Þ
ð63Þ
ð64Þ
Copyright 2005 by Marcel Dekker. All Rights Reserved.
ð65Þ
B. Syntheses Via Nucleophilic Substitution
1. Mechanistic Studies
Numerous PEKs were synthesized by the nucleophilic substitution of aromatic fluoro-
or chloroketones, and in this connection several research groups conducted detailed

mechanistic studies [7,8,98–110]. Using
13
C and
19
F NMR spectroscopy combined with
model reactions and computer calculations of electron densities the activating power of
CO-groups for F in para position was compared to that of other electron -withdrawing
groups [7,8] and a significantly weaker activation effect was found. When the reactivities
of the dihalobenzonaphthones (66a–d) were compared in polycondensation process with
diphenols the reactivity decreased in the given order (a ! d). The failure of the dichloro
compound (66d) to yield PEKs was attributed to the low reactivity (i.e. electrophilicity) of
the Cl–C bond in aromatic nucleophilic substitution (S
N
AR). However, several authors
found in detailed mechanistic studies [99–103] that the chlorobenzophenones easily
undergo a radical side reaction (67–69) yielding saturated chain ends. The extent of this
side reaction depends very much on the solvent and to a lesser extent on the redox
potential of the phenoxide ions. The following order of decreasing usefulness (i.e.
decreasing molecular weights of the isol ated PEKs) of polar solvents was found:
DPSU > DMAc > NMP > TMU > DMPU
(diphenylsulfone, dimethylacetamide, N-methylpyrrolidone,tetramethylurea,1,3-dimethyl-
perhydropyrimidinone 2).
ð66Þ
ð67Þ
ð68Þ
ð69Þ
Copyright 2005 by Marcel Dekker. All Rights Reserved.
In DPSU the radical side reactions are almost completely avoidable [10].
Furthermore, addition of a radical scavenger may be helpful to raise the molecular
weights [100]. Another approach consists of the use of special phase transfer catalysts

(70a,b) which promote the polycondensation of chlorobenzophenones and diphenols in
the presence of K
2
CO
3
[104]. These pyridinium salts were selec ted because they are stable
up to 300

C even under alkaline conditions. Another version of this approach is the
combination of these pyridinium salts with an amount of KF. Activation of the phenolic
OH-groups and under certain reaction conditions a halogen exchange takes place so that
the far more reactive fluoroketones are formed as reaction intermediates [105]. However
the activation of KF by means of the phase-transfer catalysts (68a–c) may have the
additional effect, that the fluoride ions begins to cleave the PEK backbone at temperatures
as low as 160

C. From other studies [10,106,107] it was known that KF alone attacks the
PEK chains only at temperatures 300

C. Transetherification, catalyzed by phenoxide
ions was also studied by several authors [105–108]. Another important aspect investigated
in two papers [109,110] is the influence of the reaction medium on the molecular weight in
polycondensations exclusively involving fluoroketones and the S
N
AR mechanism. In the
first paper [109] difluorobenzil (71a) was polycondensed with free diphenols and K
2
CO
3
in four different solvents DMSO and sulfolane gave the best results, whereas cleavage

of the PEK backbone was found in NMP and DMPU. However, excellent molecular
weights were obtained in NMP when silylated diphenols (71b) and a catalytic amount of
CsF were used as reaction partners of (71a). In the second paper it was reported that
DMPU is advantageous over NMP when less reactive electrophiles than fluoroketones or
fluorosulfones are used (see Section III.F).
ð70Þ
ð71Þ
C. Various Structures
Most papers reporting on syntheses of PEKs de al with a systematic variation of their
structure with the purpose to elucidate structure property relationships. In the present
review the discussion of these papers has been subd ivided into the following groups:
1. PEKs prepared from 4,4
0
-difluorobenzophenone (DFBP) and various diphenols
[111–122]
Copyright 2005 by Marcel Dekker. All Rights Reserved.
2. PEKs prepared from new fluoroketone monomers and commercial diphenols
[123–148]
3. Fluorinated PEKs [149–154]
4. Liquid-crystalline PEKs [155–157]
5. Telechelic oligomers, block-copolymers and networks
6. Hyperbranched PEKs.
All these PEKs were synthesized via the standard procedure, K
2
CO
3
(rarely in
combination with Na
2
CO

3
) was used as catalyst and HX acceptor in polar solvents such
as, DMSO, DMAc or NMP combined with toluene for the azeotropic removal of water.
However in Section III.c.2 addition al chain extension methods will be discussed.
In two publications [111,112] polycondensation of DFBP with hydroquinone,
resorcinol or 4,4
0
-dihydroxybenzophenone were described. In addition to the homopoly-
mers a series of copolymers with systematic variation of the hydroquinone/resorcinol ratio
was studied. Anothe r publication [113] reported on an analogous series of copolymers
prepared from DFBP and mixtures of hydroquinone and 1,5-dihydroxynaphthalene. The
role of various side reactions was also discussed. In several publications substituted
diphenols were used, mainly with the purpose to improve the solubility and to reduce
the melting temperature. Typical examples are PEKs derived from resorcinol having a
pendant adamantly group (72a) [114] or PEKs prepared from the substituted hydro-
quinones (72b–e) [115]. Polyelectrolytes of structure (73) were prepared by copolyconden-
sation of a sulfonated hydroquinone and unsubstituted hydroquinone [116]. The free
sulfonic acid served as binding site for the fixatio n of basic NLO chromophors, such as
(74a,b) and (75a,b) in the form of their pyridinium salts. Several PEKs showing improved
solubilities compared to their unsubstituted analogs resulted from polycondensations of
DFBP and methylated dihydroxybiphenyls (76a,b) [117]. Phenyl substituted biphenyl diols
(77a–c) were also used as comonomers of DFBP, and these monomers imparted high T
g
s
combined with good solubilities and high thermostabilities into the PEKs [118]. In another
paper several PEKs and PES were prepared from DFBP (or DFDPS) and 4,4
0
-dihydroxy
m-terphenyl [119] to obtain amorphous thermostable polyethers. Amorphous, but also
fluorescent PEKs were the result, when phenolphthaleine and the substituted phenolphtha-

leins (78a,b) were used as comonomers of DFBP [120]. Particularly bulky cardomono mers,
such as the fluorene derivatives (79a,b) have, of course, again the consequence that the
pertinent PEKs are amorphous, soluble in numerous solvents and highly thermostable
[121]. Finally, PEKs derived from the hydroxyphenylphthalazin (79c) need to be mentioned
[122]. In this case the PEK backbone includes C–N bonds in addition to ether groups.
ð72Þ
ð73Þ
Copyright 2005 by Marcel Dekker. All Rights Reserved.
ð74Þ
ð75Þ
ð76Þ
ð77Þ
ð78Þ
ð79Þ
Most syntheses of PEKs showing new structural elements were based on new or
noncommercial ‘fluoromonomers’. A difluorodiketone (80a) with a kinked structure
designed to reduce the melting temperatures of the PEKs derived from it was prepared
from isophthaloyl chloride and fluorobenzone [123]. The extremely kinked group of
monomers having structure (80b–d) required a more cumbersome synthesis. The resulting
PEKs were amorphous and possessed high glass transitio n temperatures (T
g
s) when
derived from (80d) [124,125]. Two research groups [126–128] reported on alkyl substituted
PEKs prepared from the lengthy ‘fluoromonomers’ (81a–c). This substitution pattern
considerably improves the solubility and eliminates the crystallinity, but the T
g
s remain
rather low (around 150

C). Further studies concerned ‘fluoromonomers’ derived from the

biphenyl moieties (82a,b) [129], (83a) [130], (83b) [131]. Several publications reported on
Copyright 2005 by Marcel Dekker. All Rights Reserved.
syntheses and polycondensations of new fluoromonomers derived from naphthalene. For
instance, the 2,6-substituted naphthalene monomers (84a), were described in Refs. [132]
and [133]. Syntheses and polycondensations of the 1,5-substituted naphthalenes (84b,c)
were reported in Refs. [134–136]. The chemistry of the ‘1,8-naphthalene monomer’ (85a)
was described in Ref. [137]. From the tetrasubstituted monomer (85b) a kind of comb-like
PEK was prepared [138]. PEKs derived from a monomer having pendant naphthyl groups
were prepared from (86) and had high T
g
s [139]. New ‘fluoromonomers’ derived from
indane (87a,b) were synthesized from 4-methyl-a-methylstyrene [140,141]. The PEKs
derived from them were as expected amorphous. Two research groups were interested in
polyethers having alternating sequences of ketone, ether and sulfone groups [142,143]. For
their syntheses mainly the monomers (88a or b) were used. Another research group [144]
reported on syntheses of phosphorous containing PEKs from monomer (89a). In this work
and in publications discussed below the fluorinated bisphenol-A (89b) was used as one of
the comonomers. Several authors had interest in PEKs containing heterocycles in the
backbone. Thiophene containing PEKs were obtained from monomers (90a or b) [145],
[146], and benzofurane based PEKs or PESs were prepared from monomers (91a or b)
[147]. PEKs having pyridine or isoquinoline rings in their backbones were obtained by
polycondensations of the monomers (92a) [133] or (92b) [132] and (92c) [148].
ð80Þ
ð81Þ
ð82Þ
Copyright 2005 by Marcel Dekker. All Rights Reserved.