Macromol. Rapid Commun.

2001, 22, 1463–1473 1463

Review: Polyethers, both aliphatic, such as poly(ethylene

oxide), poly(propylene oxide), etc., and wholly aromatic
ones, such as poly(phenylene oxide)s, are commercially
important materials. Polymers belonging to the former
class are primarily synthesized via a ring-opening poly-
merization route, while those belonging to the latter are
prepared via either oxidative coupling or nucleophilic aro-
matic substitution approaches. Polyethers that contain
both aromatic and aliphatic units in their backbone are far
less common. This review will discuss some of the recent
advances in the preparation of polyethers, primarily
focussing on those where the ether linkage is generated
during polymerization. Although the standard ring-open-
ing polymerization (ROP) route toward aliphatic poly-
ethers has witnessed several interesting developments in
recent years, it will not be covered in this review. The last
section deals with a new melt-transetherification approach
for the preparation of poly(xylylene alkylene ether)s
developed in our laboratory.

Recent Developments in Polyether Synthesis

Manickam Jayakannan,a S. Ramakrishnan*
Department of Inorganic and Physical Chemistry, Indian Institute of Science, Bangalore 560012, India
Fax: 080-360-0683; E-mail: [email protected]

Keywords: melt transetherification; oxidative coupling; polyethers; reductive polycondensation; ring-opening polymeriza-
tion;

Introduction Poly(arylene ether)s

Polyethers can be broadly classified into three categories Poly(arylene ether)s, which have diaryl ether linkages,
based on the type of ether linkage present in the back- can be further subdivided into three types, depending on
bone, i. e. (i) diaryl ether linkage, (ii) aryl-alkyl ether link- their repeat unit structure, as indicated below.
age, and (iii) dialkyl ether linkage. The most extensively
investigated polymers among these belong to categories
(i) and (iii). We discuss here some of the recent advances
made under each subheading. Although there are several
reviews that discuss specific classes of polyethers in great
detail,[1 – 5] we attempt to present an overview highlighting
the recent developments in the various categories, using
specific examples for illustration. Hence, this review
aims at tracing the progress in synthetic methodology and
mechanistic understanding, rather than being an exhaus-
tive survey describing the various types of polyethers that
have been reported. Polyethers belonging to Type A, by structural imposi-
a
Current address: Laboratory of Macromolecular and Organic tion, are most often prepared via self-condensation of an
Chemistry, Eindhoven University of Technology, Eindhoven, AB-type monomer, while those belonging to the other
The Netherlands. two categories are generally prepared by an (AA+BB)-

Macromol. Rapid Commun. 2001, 22, No. 18 i WILEY-VCH Verlag GmbH, D-69451 Weinheim 2001 1022-1336/2001/1812–1463$17.50+.50/0
1464 M. Jayakannan, S. Ramakrishnan

type condensation route. A wide range of polymers has

been prepared by varying the monomer structures, utiliz-
ing both approaches. Here again, rather than presenting
an exhaustive survey of the various types of structures
prepared, our emphasis is to follow the development in
the synthetic strategies by highlighting selected exam-
ples.

Type A Poly(arylene ether)s

Poly(phenylene oxide) (PPO) or, more correctly,
poly(2,6-dimethylphenylene oxide) (2), was the first Scheme 1.
commercially successful wholly aromatic polyether. The
process for its preparation was developed by General cally stable copper complex based on N,N 9-di-tert-butyl-
Electrics Company, USA, over half a century ago ethylenediamine (DEBDA) was the key to the commer-
(Scheme 1).[6 – 8] It involves the oxidative coupling of 2,6- cial success of the process. PPO generated in this manner,
dimethylphenol (1), in the presence of stoichiometric via the oxidative coupling route, would ideally be
amounts of oxygen (oxidant), copper(I) chloride (cata- expected to possess a single phenolic end-group. How-
lyst) and an amine ligand, which is either pyridine or ever, due to the unavoidable occasional occurrence of
N,N,N 9,N 9-tetramethylethylenediamine (TMEDA). The C1C coupling, it has been estimated that typically about
amine ligand plays an important role in governing the 3% of the polymer chains possess phenolic end groups at
fate of the polymerization: coupling can occur via the for- both chain ends.[1] An interesting review by Hay,[9] one of
mation of either a C1O bond leading to the polymer or a the key players in the discovery of this polymerization,
C1C bond that leads mainly to a small dimeric molecule provides excellent insight into the chronological develop-
(3). It is generally seen that higher amine concentrations ment of this methodology along with a representative
lead to the predominance of C1O bond formation yield- sampling of the range of structural variation to this basic
ing the expected polyether. The generation of a hydrolyti- backbone that has been achieved.

Manickam Jayakannan was born in Madurai, India, in 1972 and received his B.S. and M.S.
degrees from SVN College, Madurai Kamaraj University, in 1993 and 1995, respectively. He
obtained his Ph.D. degree from the Indian Institute of Science, Bangalore, under the super-
vision of Prof. S. Ramakrishnan for studies on branched and linear polyesters and polyethers.
Since August 2000, he has been working as a post-doctoral fellow in the group of Prof. Rene
A. J. Janssen, Eindhoven University of Technology, The Netherlands. His current interests
are centered around the development of novel methodologies for polymer synthesis and new
p-conjugated low-band gap materials for solar cell applications.

S. Ramakrishnan received his B.S. degree from SIES college, Bombay University, in 1980,
and his M.S. degree from the Indian Institute of Technology, Bombay, in 1982. Subsequently,
he went to the University of Massachusetts, Amherst, where he carried out his graduate stu-
dies under the supervision of Prof. J. C. W. Chien in the area of conjugated polymers. After
obtaining his Ph.D. in 1988, he worked as a post-doctoral fellow for two years in the Corpo-
rate Research Laboratory, Exxon Research and Engineering Company, Annandale, New Jer-
sey. During this time he worked on utilizing borane-containing monomers to prepare functio-
nalized polyolefins via Ziegler-Natta and ROMP techniques. In 1990, he returned to India to
take up a lecturer position at the Indian Institute of Science, Bangalore, where he currently
holds a position as an Associate Professor. His research interests are focused on the develop-
ment of new polymerization methodologies, hyperbranched polymers, conjugated polymers,
and molecular imprinting, as well as self-associating and self-assembling macromolecules.
Recent Developments in Polyether Synthesis 1465

recent phase-transfer catalysis (PTC) modification leads

to well-defined PPOs with controllable molecular weights
and functionality.[15] Chain-growth in this case takes place
via the addition of the phenoxy radical to a phenolate
Scheme 2.
anion (Scheme 3). This process, therefore, has been
referred to as an anion-radical reaction as opposed to the
radical-radical coupling that is believed to occur in stan-
dard oxidative coupling polymerization.[2] The success of
the PTC approach relies on the low concentration of the
phenoxy radicals in the organic (polymerization) phase
(due to the low concentration of the oxidant – air), and
hence no C1C coupling is seen to occur.
Further, by the use of various end-blocking phenols,
such as 2,4,6-trimethylphenol or 4-tert-butyl-2,6-
dimethylphenol (8) as comonomers, the molecular
weights of the resulting polymers are conveniently con-
trolled leading to PPOs (9), with single phenolic end-
groups that can be further functionalized (Scheme 4).[16]
On the other hand, the use of a Bisphenol A-type mole-
cule (10) as a chain-initiating phenol results in the forma-
tion of telechelic PPOs (11) with two phenolic chain
ends.[17]
Scheme 3.
The other more commonly used route to prepare poly-
(arylene ether)s utilizes the conventional nucleophilic
An alternate approach to this class of poly(arylene aromatic substitution (SNAr) process. This approach has
ether)s is via the Ullmann condensation of bromo-substi- been most successfully applied to the preparation of sev-
tuted phenols, for instance 4-bromophenol (4) (Scheme eral commercial engineering thermoplastics, such as
2). This approach to PPO was explored very early by Sta- poly(ether ketone)s and poly(ether sulfone)s, etc.[3] The
matoff but was not commercially exploited.[10] Decades requirement of an electron deficient (activated) aromatic
later, McGrath and coworker have re-explored this halide to affect the ipso substitution makes it difficult to
approach to prepare simple PPO and other poly(arylene prepare simple wholly aromatic (Type A) polyethers
ether)s from un-activated aromatic dihalides.[11] Copoly- using this approach. However, after the demonstration
merization using different bisphenols permitted them to that perfluoroalkyl substituents can effectively activate
prepare a range of polymers with varying levels of crys- aromatic halides towards SNAr-type ipso substitution,[18]
tallinity. it became apparent that wholly aromatic polyethers can
Another mechanistic pathway for the preparation of indeed be readily prepared via this route. A very recent
poly(arylene ethers)s from bromophenols proceeds via a example of such a polymer prepared via the self-conden-
radical anion intermediate (SRN19 mechanism).[12 – 14] These sation of an AB-type monomer, 49-hydroxy-4-nitro-3-tri-
polymerizations occur in the presence of catalytic fluoromethylbiphenyl (12), illustrates the potential of
amounts of an oxidative initiator, such as lead oxide, using pendant perfluoroalkyl groups (Scheme 5).[19] Reac-
potassium ferricyanide, or even in the presence of air. tion in this case proceeds readily by the nucleophilic dis-
While some of the earlier attempts using this route lead to placement of the nitro-group by the phenolate. Interest-
uncontrolled functionality and branching,[13] the more ingly, this displacement occurs smoothly in the absence

Scheme 4.
1466 M. Jayakannan, S. Ramakrishnan

Scheme 5.

of any side reactions, which are often associated with

reactive nitrite ions at the high polymerization tempera-
tures.

Type B Poly(arylene ether)s

Polyethers having two types of aryl groups are typically
prepared via an AA + BB-type polycondensation route.
The reaction pathways that could be used to prepare such
structures fall into either an Ullmann-type coupling, or the Scheme 6.
SNAr route wherein pendant activating groups are present.
McGrath and coworkers demonstrated the utility of the
Ullmann route to prepare polyethers 14, from Bisphenol A
and 4,49-dibromobiphenyl.[11] The Ullmann coupling con-
ditions further permitted the use of a range of other un-
activated dihaloaromatics as well as halophenols, such as
4-bromophenol, as comonomers to access copolyethers
with wide-ranging crystallization propensities.

More recently, SNAr-type polymerization has also been

effectively used to prepare such wholly aromatic poly- Scheme 7.
ethers. In these cases the dihalo(dinitro)aromatics are
activated toward SNAr-type polycondensation by the pre-
sence of pendant electron-withdrawing substituents. In nyl group, as evidenced by their comparable Hammet r
one of the early examples of such an approach pendant values (rCF3 = 0.54, rC2O = 0.5).[4] The successful genera-
benzoyl groups were utilized.[20] Thus, using 2,5-difluoro- tion of high-molecular weight polymers via the standard
4-benzoylbenzophenone (15) and hydroquinone, wholly SNAr route amply testifies to the efficacy of this group in
aromatic polyethers of high molecular weights (ginh = 0.8 activating the aromatic ring toward nucleophilic substitu-
dL/g) were prepared. tion. The very high oxidative stability of these perfluoro
Polyether 16 derived from hydroquinone was highly groups is an added advantage from the point of view of
crystalline and soluble only in solvents such as penta- enhanced thermal stability.
fluorophenol and methanesulfonic acid. On the other A wide range of wholly aromatic polyethers (20) were
hand, polyether 17 from methylhydroquinone was less prepared under standard SNAr polymerization conditions,
crystalline and readily soluble in a range of common using polar aprotic solvents, such as N-methylpyrrolidone
organic solvents like chlorinated hydrocarbons, tetrahy- (NMP), and a base, such as K2CO3 (Scheme 7). High-
—
drofuran (THF), N,N-dimethylacetamide (DMAc), etc., molecular weight polymers (M n = 30 000–40 000, PD =
clearly illustrating the role of the random (regio-irregular) 2.5, as determined by means of gel permeation chromato-
placement of the methyl substituent along the polymer graphy (GPC) using PS standards) with very high Tgs
backbone. (200–250 8C) and decomposition temperatures A 500 8C
Pendant perfluoromethyl groups have also been used to were readily prepared. The use of pendant activating
prepare Type B polyethers. Banerjee et al. prepared poly- groups thus precludes the incorporation of these groups
ethers from difluoroaromatic monomers that were acti- (like carbonyl and sulfone) into the backbone, which is
vated by pendant trifluoromethyl groups present in the essential to the generation of a range of engineering ther-
ortho positions.[21] The electron-withdrawing effect of the moplastics, such as poly(ether ether ketone)s (PEEK) and
trifluoromethyl group is similar to that of a ketone carbo- poly(ether sulfone)s.
Recent Developments in Polyether Synthesis 1467

Type C Poly(arylene ether)s phenone chain-ends, produced via this reductive dehalo-
Polyethers belonging to this class are “structurally” not genation, and the molecular weight of the resulting poly-
simple poly(arylene ethers)s but contain an additional mer, thus confirming the occurrence of the SRN1 path-
(activating) functionality in their backbone, such as a way.[22] The occurrence and extent of the SRN1 pathway
ketone, sulfone, etc. depends on both the type of dihalide and phenolates used.
The extent of reaction via this pathway is high when
dichloroaromatics and simple hydroquinone are used,
while it is far less when bisphenols and fluoroaromatics
are used. Percec et al. have provided a detailed discussion
rationalizing this behavior and its dependence on the
monomer structures.[2]
Apart from the conventional activating groups, several
new electron-withdrawing substituents have been utilized
to enable polyether formation via the SNAr route. As
The approach toward the preparation of such poly- mentioned in the previous section, by the use of perfluor-
ethers primarily relies on the SNAr-type polycondensation oalkyl groups several new poly(aryl ether)s (21) have
of activated aromatic dihalides by bisphenolates.[4] Poly- been prepared – a few recent examples are shown
merization most often follows a simple SNAr mechanism below.[23] Several of these polymers exhibit very high
via an intermediate Meisenheimer complex, the formation thermal stability and mechanical properties that are com-
of which is the rate-determining step in the polymeriza- parable to typical high-performance engineering thermo-
tion process (Scheme 8). Detailed mechanistic studies by plastics.
Percec and coworkers, however, clearly demonstrate that
the reaction in some instances proceeds via an SRN1 path-
way, wherein the aromatic dihalide accepts an electron
from the phenolate to form a radical anion.[2] As shown in
Scheme 8, the radical anion can either react with the phe-
noxy radical to generate the Meisenheimer complex or
loose the halide ion to generate a phenyl radical. The gen-
eration of a phenyl radical, which subsequently termi-
nates via hydrogen abstraction from a hydrogen donor
(usually the solvent), leads to the generation of an unreac-
tive benzophenone chain end. This process leads to a con-
siderable decrease in the molecular weight of the result-
ing polymer. Through a set of detailed experiments using
NMR spectroscopy, Percec and coworkers clearly
showed a correlation between the concentration of benzo-

As shown in Scheme 9, heterocyclic rings, such as qui-

noxalines, oxadiazoles, thiadiazoles and triazoles, have
also been used effectively as activators of the bis(arylha-
lide) monomers (22).[24]

Scheme 9.

Other activating groups that have been used include the

cyano group[25] and a fairly novel group containing 3-
Scheme 8. cyclobutene-1,2-dione (24):[26]
1468 M. Jayakannan, S. Ramakrishnan

using CsF and potassium 4-benzoylphenolate appeared

over a decade ago,[33] but the routes to cyclomers were
low-yielding and required rather high polymerization
temperatures. An interesting alternative for poly(arylene
ether ketone)s was developed by Hay and coworkers[34]
based on the use of 1,2-dibenzoylbenzene to promote
cyclomer formation. The resulting cyclomers were read-
Use of Masked Bisphenols ily polymerizable via ROP to yield high-molecular
—
weight polymers (M w = 300 000, PD A 10; GPC using
Silylated bisphenols in the presence of cesium fluoride, in polystyrene standards). Two problems associated with
lieu of the bisphenolate salt, serve as an excellent alterna- this approach is the rather broad molecular weight distri-
tive for the preparation of poly(arylene ether)s under vir- bution (reflective of branching) and incomplete conver-
tually neutral conditions (Scheme 10).[27] Kricheldorf and sion during ROP even at high temperatures.
coworkers developed this route and have utilized it to
carry out polymerization in the melt with a variety of
bisphenols to prepare a wide range of poly(aryl ether)s,
including poly(aryl sulfone)s,[28] poly(aryl ether
ketone)s,[29] poly(aryl ether pyridine)s[30] and poly(aryl
ether benzonitrile)s.[31] Hay and coworkers similarly uti-
lized carbamate (also carbonate) derivatives as masked
bisphenols.[32] The carbamate groups are readily cleaved
in the presence of KHCO3/K2CO3 at 155–165 8C to gener- A more recent approach toward isoPEK utilized dilu-
ate the potassium phenolate in situ. In the presence of tion methods to generate mixtures of cyclic oligomers via
activated bis(arylfluoride)s high-molecular weight poly- the self-condensation of 4-fluoro-3-hydroxybenzophe-
mers ([g] = 0.6–0.9 dL/g) are obtained in relatively short none.[35] The mixture of cyclomers was readily converted
—
periods. to high-molecular weight polymers (M w = 60 000–
180 000 PD = 2–5; GPC using polystyrene standards) at
high conversions using potassium (or cesium) 4-benzoyl-
phenolate.

Hyperbranched Poly(aryl ether)s

Nucleophilic aromatic substitution has also been used
effectively to generate hyperbranched poly(arylene
ether)s. The first example was that by Miller et al.,[36]
who prepared the hyperbranched analog of isoPEK 29
Scheme 10. (Scheme 11).

Poly(arylene ether)s via Ring-Opening Polymerization

Traditionally, ROP has been applied primarily to the
synthesis of aliphatic polyethers, such as poly(ethylene
oxide)s, poly(oxytetramethylene), etc. In recent years,
ROP routes for the preparation of engineering thermo-
plastics have received a great deal of attention.[5] The low Scheme 11.
melt viscosity of the cyclic oligomers (cyclomers), which
can be readily transformed to high-molecular weight Alternate routes using 3,5-difluoro-4-hydroxybenzo-
polymers without the release of volatiles, presents several phenone (30) also resulted in a variant of the above struc-
processing advantages, e. g., for use in reactive injection ture (Scheme 12). Reversing an AB2-type monomer,
molding. Furthermore, this approach is better suited for Hawker et al. prepared an A2B-type monomer 32, which
the fabrication of fiber-reinforced thermoplastic compo- when polymerized yielded hyperbranched polyethers
sites because of the potentially improved fiber-wetting with phenolic instead of fluoro end groups (Scheme
properties of the low-molecular weight cyclomers. It can 12).[37] This reversal had a considerable influence on their
also lead to improved consolidation. solubility as well as their glass transition temperatures. It
Early reports of all-aromatic diarylene ether ketones is interesting to note that, despite the fact that in the for-
and sulfone cyclomers, and their SNAr-mediated ROP mer case involving 30, the fluoroaromatic nucleus is not
Recent Developments in Polyether Synthesis 1469

Scheme 14.
Scheme 12.

Scheme 13. Scheme 15.

activated toward an SNAr-type displacement (by virtue of Subsequently, AB-type self-condensation routes based
the carbonyl group being in the meta position), polymeri- on 3-bromomethylphenols (37) to generate linear poly-
zation proceeds under typical SNAr conditions, although (benzyl ether)s was demonstrated by FrØchet and cowor-
the molecular weights were observed to be lower (31 had kers.[41] This direct alkylation using phenolates, however,
— —
an M n = 20 000, PD = 1.8, while 33 had M n = 43 000, PD resulted in significant C-alkylation, and leads to the gen-
= 2.3). eration of branched structures 38. Using masked silyl ether
Using an improved silyl-masked AB2 monomer (34), modification 39 and KF/K2CO3 as the catalyst in the pre-
Thompson et al. prepared hyperbranched poly(ether imi- sence of 18-crown-6, this approach yielded linear poly-
—
de)s 35 (Scheme 13) of high molecular weights (M n = ethers 40, essentially devoid of C-alkylation (Scheme 14).
20 000, PD = 3.5; GPC using polystyrene standards).[38] The approach was further extended to the preparation
Polymerization was found to be very rapid when cata- of hyperbranched poly(benzyl ether)s starting from the
lyzed with CsF at 240 8C, using diphenylsulfone as the corresponding AB2 monomer 5-bromomethyl-1,3-dihy-
solvent. Further studies on this polymerization pointed to droxybenzene (41) (Scheme 15).[42]
the occurrence of a post-polymerization transetherifica-
tion process that leads to a decrease in the degree of
branching along with an increase in polydispersity.[39] Poly(alkyl ether)s
Generally, wholly aliphatic polyethers are prepared via
Poly(aryl alkyl ether)s the ROP of the appropriate cyclic ether monomers. Con-
densation routes that involve the generation of ether link-
Polyethers belonging to this class have received relatively
age, for instance via the Williamson route, work reason-
less attention. The methods used to prepare such poly-
ably well only with phenolates, but fair poorly as a poly-
ethers rely primarily on the Williamson-type etherifica-
merization methodology when aliphatic alcohols are
tion process.
used.
An interesting condensation route was recently devel-
oped by Yokozawa and coworkers for the preparation of
poly(xylylene ether)s.[43] This approach was based on a
The generation of high-molecular weight polymers via reductive polycondensation of a aromatic dialdehyde
this approach has been possible only more recently under using trialkylsilane as the reducing agent and trityl per-
phase-transfer catalysis (PTC) conditions. The PTC chlorate as the catalyst (Scheme 16).
method for polycondensation was developed by Percec The polymerization, as is believed, proceeds via the
and coworkers for the preparation of liquid crystalline formation of the alkoxysilane, which further condenses
polyethers 36, containing biphenyl mesogens.[40] with the aldehyde to generate the ether linkage via a
hemiacetal intermediate (Scheme 17). The relative rates
of formation of the alkoxysilane via reduction and the
subsequent reaction of these to form the hemiacetal was
postulated to control the formation of high-molecular
weight polymers.
1470 M. Jayakannan, S. Ramakrishnan

Transetherification Approach
We have recently been working on an alternate condensa-
tion route toward poly(alkyl ether)s, which involves a
Scheme 16. transetherification reaction. A transetherification process,
in general, may be represented as shown in Scheme 20.

Scheme 20.

In the context of polycondensation, this would imply

the reaction of a di-ether with a diol to generate the poly-
mer and a low-molecular weight alcohol (Scheme 21). As
Scheme 17. in the case of transesterification, this process would be
driven to completion by a continuous removal of the low-
boiling alcohol.

Scheme 21.

An important aspect for ensuring the success of such a

Scheme 18. strategy is the design of a di-ether monomer in which
only one of the C1O linkages of the ether group is rela-
tively labile so that rapid transetherification can occur
leading to the formation of a high-molecular weight poly-
mer. Di-ether monomers having methyl benzyl ether moi-
eties are ideally suited for such a process because of the
Scheme 19. facile generation of the benzylic cation in such molecules
under acid-catalyzed conditions. A problem that is
Based on the intermediacy of the silyl ethers, this encountered in such a process is Friedel-Crafts alkylation
approach was further extended for the preparation of that can take place under the polymerization conditions
other polyethers containing aliphatic segments, along applied, leading to the formation of crosslinked products.
with the xylylene units, by the use of appropriate a,x-bis- This is, however, easily circumvented by the use of fully
(silyloxy)alkanes 46 (Scheme 18).[44] The self-condensa- substituted derivatives. Thus, completely substituted bis-
tion of isophthalaldehyde (leading to relatively larger (benzyl methyl ether)s (such as 50; Scheme 22) readily
values of x in 47) was precluded only when bulky silanes,
such as Ph3SiH, were used as the reducing agent.
In a more recent report, Yokozawa et al. showed that
the intermediacy of the hemiacetal is not necessarily the
only pathway to generate the polymer: bis(trimethylsilyl)
p-xylylene ethers yield the corresponding polyethers
directly by a loss of hexamethyldisiloxane as the conden- Scheme 22.
sate in the presence of trifluoromethanesulfonic acid.[45]
A problem associated with this approach is the inability undergo this transetherification polycondensation with a
to completely circumvent Friedel-Crafts alkylation that variety of diols in the presence of p-toluenesulfonic acid
can lead to the formation of crosslinked products, which (PTSA) to give polyethers of moderate molecular weights
—
is typically observed during prolonged polymerization (M w = 20 000–40 000, PD = 3–5; GPC using polystyrene
times. standards).[47] An important feature of this approach is the
In another interesting extension of this approach, solvent-free melt condensation that is readily affected. A
Yokozawa et al. demonstrated the possibility of simulta- wide range of aliphatic diols as well as oligo(ethylene
neously introducing pendant units by the use of other silyl glycol)s were shown to be effective in generating linear
nucleophiles instead of simple trialkylsilanes, such as polyethers.
cyanotrimethylsilane, allyltrimethylsilane or even silyl Typically, melt polycondensation is readily performed
enol ethers during polymerization (Scheme 19).[46] at 180 8C, driving the equilibrium to the right by the con-
Recent Developments in Polyether Synthesis 1471

Scheme 23.

tinuous removal of methanol either under nitrogen purge temperatures as high as 180 8C in the presence of PTSA,
or under reduced pressure. At lower temperatures around clearly show that no self-condensation takes place.[49]
150 8C, the molecular weights of the resulting polymers Thus, the transetherification approach permits the genera-
are lower. Molecular-weight distributions, in some cases, tion of structurally well-defined polyethers that incorpo-
are often broad and the polydispersity indices are signifi- rate various aliphatic diol segments. More importantly,
cantly higher than the expected value of 2 for an ideal the transetherification approach is a solvent-free melt-
step-growth polymerization process. This could be due to condensation route. The problem of Friedel-Crafts alkyla-
branching and/or other side-reactions. Possible side reac- tion that Yokozawa et al. face is also a problem in the
tions that could occur under such polymerization condi- transetherification methodology but is readily remedied
tions are the dehydrative self-condensation of the alipha- by the use of peralkylated monomers.
tic diols and/or the cleavage of the aryl ether linkages The use of monomers based on relatively inexpensive
present in some of the benzyl ether monomers (50-type starting materials, such as durene and isodurene, makes
I). The presence of similar broad distributions in the case this approach fairly attractive in terms of being able to
of durene and isodurene-based monomers (50-type II) generate a wide range of polyethers via melt condensa-
suggested that the latter is not the only reason, if taking tion. For instance, by the use of poly(ethylene glycol)s
place at all. The occurrence of some these side reactions (PEG) of varying molecular weights a series of segmen-
was probed by an analysis of the depolymerization prod-
ucts. Because of the intrinsic equilibrium nature of this
polymerization process, the polymers were readily trans-
formed to the starting monomers by reversal of the equili-
brium in the presence of excess methanol (Scheme 23).[48]
The analysis of the products of such a degradative
methanolysis showed the presence of the two expected
monomers only. No evidence was seen from NMR spec-
troscopy for the occurrence of diol self-condensation. ted poly(ethylene oxide)s (SPEO) were readily pre-
Mass-spectral analysis also confirmed that this type of pared.[50] Glass transition temperatures and crystallization
self-condensation might happen to a minor extent only. propensities of these types of SPEOs are determined by
The broad distribution may therefore be due either to the the PEG segment length, and therefore make them poten-
presence of cyclic compounds or some alternate process. tially useful for solid-polymer electrolyte applications.
Further detailed studies using techniques, such as Furthermore, the modification of the pendant groups
MALDI-TOF mass spectrometry, are likely to reveal the using phenolic-type monomers can lead to other interest-
exact origin. ing structures, such as branched PEO equivalents 54.
It is important to recognize certain similarities between More recently, this approach was also shown to be use-
our transetherification approach and the approach devel- ful for the preparation of hyperbranched polyethers.[51]
oped by Yokozawa et al.,[45] using xylylene bis(trimethyl- Utilizing a readily available starting material, such as
silyl ether)s. In the latter, a self-condensation of the bis(- mesitol, an AB2-type monomer was easily synthesized
trimethylsilyl ether) monomer occurs to give the corre- and its polymerization yielded the appropriate hyper-
sponding poly(xylylene ether) by exclusion of hexam- branched polyether (Scheme 24). The degree of branch-
ethyldisiloxane, but in the former, control experiments ing in these systems was shown to be fairly high around
with the bis(methoxy methyl)benzene monomers, even at 0.75. The incorporation of other spacers, such as oli-

Scheme 24.
1472 M. Jayakannan, S. Ramakrishnan

go(ethylene oxide) units, can lead to highly branched [3] For some recent reviews see: [3a] S. Maiti, B. Mandal,
PEO equivalents, which may be expected to have very Prog. Polym. Sci. 1986, 12, 111; [3b] M. Bruma, J. W.
Fitch, P. E. Cassidy, J. Macromol. Sci., Rev. Chem. Phys.
low crystallization propensity. These types of amorphous
1996, C36, 119.
PEO equivalents could find potential use for solid-poly- [4] For an excellent review on poly(arylether) syntheses, see:
mer electrolyte applications. J. W. Labadie, J. L Hendrick, M. Ueda, ACS Symp. Ser.
1996, 624, 210.
[5] D. J. Brunelle, in: Ring Opening Polymerisation: Mechan-
Conclusions ism, Catalysis, Structure, Utility, D. J. Brunelle, Ed., Han-
ser, New York 1993, chapter 11.
The area of polyethers, although very mature, continues [6] A. S. Hay, H. S. Blanchard, G. F. Endres, J. W. Eustance,
to be a very active and exciting one. In this review, we J. Am. Chem. Soc. 1959, 81, 6335.
have traced the development of the various approaches [7] U.S. 3306874 (1967), inv.: A. S. Hay.
for the synthesis of polyethers. We have primarily con- [8] U.S. 3306875 (1967), inv.: A. S. Hay.
[9] A. S. Hay, J. Polym. Sci., Part A: Polym. Chem. 1998, 36,
centrated on step-growth polymerizations, although we 505.
did make a brief mention about the exciting develop- [10] U.S. 3228910 (1966), inv.: G. S. Stamatoff.
ments in ROP routes to poly(arylene ethers)s. We chose [11] M. J. Jurek, J. E. McGrath, Polym. Prepr. (Am. Chem.
to leave out the very many interesting developments in Soc., Div. Polym. Chem.) 1987, 28, 180.
the ROP approaches for the preparation of simple alipha- [12] G. D. Staffin, C. C. Price, J. Am. Chem. Soc. 1960, 82,
3632.
tic polyethers, such as the living and immortal polymeri- [13] C. C. Price, N. S. Chu, J. Polym. Sci. 1962, 61, 135.
zation routes, that were developed and perfected during [14] H. S. Blanchard, H. L. Finkbeiner, G. A. Russell, J. Polym.
the last decade. This topic, we believe, amply deserves an Sci. 1962, 58, 469.
independent review. [15] V. Percec, T. D. Shaffer, J. Polym. Sci., Polym. Lett. Ed.
Two new approaches for accessing polyethers, which 1986, 24, 439.
[16] V. Percec, J. W. Wang, J. Polym. Sci., Part A: Polym.
contain dialkyl ether linkages, specifically benzyl alkyl Chem. 1991, 29, 63.
ethers, were highlighted. One relies on reductive polycon- [17] V. Percec, J. W. Wang, Polym. Bull. 1990, 24, 493.
densation, while the other one is based on a transetherifi- [18] J. W. Labadie, J. L. Hendrick, Polym. Prepr. (Am. Chem.
cation process. The latter, although restrictive in the Soc., Div. Polym. Chem.) 1990, 31, 344.
choice of the bis(benzyl ether) monomer, is a melt-con- [19] I. S. Chung, S. Y. Kim, Macromolecules 2000, 33, 9474.
[20] K. Konno, N. Deguchi, K. Yonetake, M. Ueda, P. E. Cas-
densation methodology, which permits the use of a wide sidy, J. W. Fitch, J. Polym. Sci., Part A: Polym. Chem.
range of aliphatic diols and oligo(ethylene glycol)s as 1997, 35, 605.
comonomers. Furthermore, the ready availability of per- [21] S. Banerjee, G. Maier, Chem. Mater. 1999, 11, 2179.
alkylated raw materials, such as durene and isodurene, [22] V. Percec, R. S. Clough, P. L. Rinaldi, U. E. Litman,
for the preparation of the requisite monomers makes this Macromolecules 1991, 24, 5889.
[23] S. Banerjee, G. Maier, M. Burger, Macromolecules 1999,
even more attractive, should some interesting commercial 32, 4279.
applications for these polyethers emerge. [24] J. L. Herdrick, J. W. Labadie, Macromolecules 1990, 23,
1561.
[25] F. M. Mercer, A. Easteal, M. Bruma, Polymer 1997, 38,
707.
Acknowledgement: S.R. would like to thank Prof. E. W. [26] Z. Y. Wang, L. Suzzarini, Macromolecules 1996, 29, 1073.
“Bert” Meijer for hosting him for three months as a “Philips [27] H. R. Kricheldorf, Makromol. Chem., Makromol. Symp.
Visiting Professor” at Eindhoven University of Technology in 1992, 54/55, 365.
early 2001, during which time most of this review was written. [28] H. R. Kricheldorf, G. Bier, J. Polym. Sci., Part A: Polym.
We would also like to thank the Department of Science and Chem. 1983, 21, 2283.
Technology, New Delhi, for funding (SP/S1/G-07/99). [29] H. R. Kricheldorf, G. Bier, Polymer 1984, 25, 1151.
[30] H. R. Kricheldorf, G. Schwarz, J. Erxleben, Makromol.
Chem. 1988, 189, 2255.
[31] H. R. Kricheldorf, J. Meier, G. Schwarz, Makromol.
Received: August 14, 2001 Chem., Rapid Commun. 1987, 8, 529.
Revised: October 9, 2001 [32] Z. Y. Wang, H. N. Cavalho, A. S. Hay. J. Chem. Soc.,
Chem. Commun. 1991, 1221.
[33] H. M. Colquhoun, C. C. Dudman, M. Thomas, C. A.
O’Mahoney, D. J. Williams, J. Chem. Soc., Chem. Com-
mun. 1990, 336.
[1] D. Aycock, V. Abolins, D. M. White, in: Encyclopedia of [34] [34a] K. P. Chan, Y. F. Wang, A. S. Hay, Macromolecules
Polymer Science and Engineering, 2nd ed., H. F. Mark, N. 1995, 28, 653; [34b] Y. F. Wang, K. P. Chan, A. S. Hay, J.
M. Bikales, C.G. Overberger, G. Menges, Eds., John Wiley Polym. Sci., Part A: Polym. Chem. 1996, 34, 375.
& Sons, New York 1988, Vol. 13, p. 1. [35] M. F. Teasley, D. Q. Wu, R. L. Harlow, Macromolecules
[2] For an excellent discussion on the mechanistic aspects, the 1998, 31, 2064.
reader is referred to: V. Percec, J. H. Wang, R. S. Clough, [36] T. M. Miller, T. X. Neenan, E. W. Kwock, S. M. Stein, J.
Makromol. Chem., Makromol. Symp. 1992, 54/55, 258. Am. Chem. Soc. 1993, 115, 356.
Recent Developments in Polyether Synthesis 1473

[37] C. J. Hawker, F. Chu, Macromolecules 1996, 29, 4370. [46] [46a] T. Yokozawa, K. Takenoya, Macromolecules 1996,
[38] D. S. Thompson, L. J. Markoski, J. S. Moore, Macromole- 29, 497; [46b] K. Takenoya, T. Yokozawa, Macromole-
cules 1999, 32, 4764. cules 1998, 31, 2906; [46c] T. Yokozawa, L. Niimi, K.
[39] D. S. Thompson, L. J. Markoski, J. S. Moore, I. Sendijare- Takenoya, Macromol. Chem. Phys. 1998, 199, 2453.
vic, A. Lee, A. J. McHugh, Macromolecules 2000, 33, [47] M. Jayakannan, S. Ramakrishnan, Macromol. Chem. Phys.
6412. 2000, 201, 759.
[40] V. Percec, H. Nava, H. Jonsson, J. Polym. Sci., Part A: [48] M. Jayakannan, S. Ramakrishnan, J. Polym. Sci., Part A:
Polym. Chem. 1987, 25, 1943. Polym. Chem. 2001, 39, 1615.
[41] K. E. Uhrich, J. M. J. FrØchet, Polymer 1994, 35, 1739. [49] M. Jayakannan, Ph.D. Thesis, Indian Institute of Science,
[42] K. E. Uhrich, C. J. Hawker, J. M. J. FrØchet, S. R. Turner, Bangalore 2000.
Macromolecules 1992, 25, 4583. [50] M. Jayakannan, S. Ramakrishnan, J. Polym. Sci., Part A:
[43] T. Yokozawa, F. Nakamura, Makromol. Chem., Rapid Polym. Chem. 2000, 38, 2635.
Commun. 1993, 14, 167. [51] M. Jayakannan, S. Ramakrishnan, Chem. Commun. 2000,
[44] T. Yokozawa, F. Nakamura, Macromolecules 1995, 28, 1967.
4668.
[45] T. Yokozawa, M. Nishimori, T. Endo, Macromol. Chem.
Phys. 1996, 197, 1361.