Handbook of Polymer Synthesis Second Edition Episode 14 pptx

14
Polymers for Organic Light Emitting
Devices/Diodes (OLEDs)
O. Nuyken, E. Bacher, M. Rojahn, V. Wiederhirn and R. Weberskirch
Technische Universita
¨
tMu
¨
nchen, Garching, Germany
K. Meerholz
Universita
¨
tZuKo
¨
ln, Ko
¨
ln, Germany
I. INTRODUCTION
Facing the 21st century, the development of new techniques that are able to display data
faster, more detailed and in mobile applications, is one of the pro spering scientific fie lds.
One approach for lightweight, flexible, power-efficient full-color displays are organic light
emitting diodes (OLEDs). Such devices with their low driving voltage, bright color
and high repetition rate (e.g. for video-application) are ideal for usage in miniature
displays as well as in large area screen [1–3]. The basic principle of these devices are
electroluminescent ‘semiconducting’ organic materials packed between two electrodes.
After charge injection from the electrodes into the organic layer and charge migration
within this layer, electrons and deficient electrons (so called ‘holes’) can recombine to form
an excited singlet state. Light emission of the latter is then a result of relaxation processes
[4–6]. To achieve high electroluminescence efficiencies, the materials have to fulfill several
specific requirements including low injection barriers at the interface between electrodes
and organic material, balanced electron- and hole-density and mobility and high lumines-

cence efficiency. Furthermore, the recombination zone should be located away from the
metal cathode to prevent annihilation of the exited state. Since no material known to date
is able to meet all these criteria, modern OLEDs consist — besides the transparent
substrate (e.g., glass, PET), anode (most commonly indium tin oxide, ITO) and metal
cathode (e.g., Mg–Ag-alloy) — of several organic layers for charge injection, transport
and/or emission [7,8] (the principal set-up is shown in Scheme 1).
ð1Þ
Copyright 2005 by Marcel Dekker. All Rights Reserved.
In such multilayer diodes, each layer can be separately optimized concerning
injection barriers, charge mobility and density and quantum efficiency. Much of the
motivation for studying organic materials stems from the potential to tailor desirable
optoelectronic properties and process characteristics by manipulation of the primary
chemical structure. Objecting optimal charge transport, recombination probability
and light emission and consequently a maximum external efficiency of the device, various
substances have been developed, modified and tested in the last few years. For hole
transport/electron blocking layers, triarylamine- and pyrazoline-structures (see Scheme 2)
were found to be most promising [9–11].
ð2Þ
For electron transport/hole blocking purposes, a wide variety of electron-deficient
moieties are well known, e.g., 1,3,4-oxadiazoles [12], 1,2,4-triazoles [13], 1,3-oxazoles,
pyridines and quinoxalines [14] (see Scheme 3). Materials with conjugated p-electron
system (e.g., styrylarylenes, arylenes, stilbenes, oligo- and poly(thiophene)s — see
Scheme 3) are widely used as co mbined charge transport and luminescence layers as
well [12,15].
Basic structures of electron transport=hole blocking materials and oligomeric
and polymeric mater ials for charge transport and luminescence
ð3Þ
Two basic principles are commonly used for the preparation of OLEDs: the
sublimation method, in which the organic layers are prepared by vapor deposition results
in well-defined layers of excellent purity but tolerates only low molecular mass molecules

with high temperature stability [16]. The less expensive preparation out of solution,
requires soluble substances or precursors [17] and is therefore widely used in combination
Copyright 2005 by Marcel Dekker. All Rights Reserved.
with polymers because of their homogeneity, good layer-building-properties and long-
term form stability resulting in a long device lifetime.
The goal of this article is to describe the scope and limitations of synthetic routes
that have been used to produce suitable oligomers and polymers for LED application.
The polymers in this article will be discussed on the basis of their backbone structure and
the synthetic strategy of their formation and are divided into completely p-conjugated
polymers, non-conjugated polymers and polymers with defined segmentation (see
Structure 4).
ð4Þ
II. n-CONJUGATED POLYMERS
Since the discovery of electrically conductive polymers by Heeger, MacDiarmid and
Shirakawa et al. in 1977 [18] — resulting in the Nobel Prize in Chemistry 2000 [19] —
p-conjugated systems have a major role in the field of so called ‘plastic electronics’. Key
property of these polymers is the conjugated double bond along the polymeric backbone,
allowing charge migration afte r injection via electrodes.
A. Poly(p-phenylene-vinylene)s (PPV)
The first polymers used for light emitting diodes — discovered by Friend and Holmes et al.
in 1990 [20] — and still the most common ones used in recent devices, are completely
p-conjugated poly( p-phenylene-vinylene)s. These polymers — which can be used in single
layer devices as both charge-transport and green emitting materials — will be discussed on
the synthetic strategy of their formation.
1. Precursor Routes
Unsubstituted poly(phenylene-vinylene)s (PPVs) are insoluble in any known solvent.
To improve solubility and with that processability unsubstituted PPVs were first synthe-
sized using precursor routes like the so called Wessling- (or sulfonium-) route [21–24].
Accordingly, the condensation is performed with solubilized monomers, and a soluble
polymeric intermediate is formed. The latter is converted to PPV in a final reaction step,

that is preferentially carried out in the solid state, allowing the formation of homogeneous
PPV films or layers. Following this route, a soluble precursor polymer with excellent
film forming pro perties is obtained by base induced polyreaction of p-xylylene-a,a
0
-
bisdialkylsulfonium salts. After spin coating, the precursor polymer is converted by
polymer analogous heat induced elimination to the corresponding PPVs (Scheme 5).
ð5Þ
Copyright 2005 by Marcel Dekker. All Rights Reserved.
In general, any functionalized poly( p-xylylene) with leaving group in the a-position
to the aromatic moieties can be used as precursor, as long as they fulfill the basic
requirements of OLED-techniques (i.e., solubility, transparency, excellent film forming
properties, good thermal stability after processing, etc.). Commonly used as leaving
groups beside the sulfonium group are halogens [25,26] (so called ‘Gilch-procedure’),
hydrohalogenides [27], alkoxides [28] and alkylsulfinyles (known as ‘Vanderzane-
procedure’) [29].
To avoid unwanted side reactions and damages of other device-layers during thermal
conversion (e.g., by ox idation or reaction with volatile corrosive elimination products),
organic-solvent soluble PPV derivatives such as poly(2-methoxy-5-(2
0
-ethylhexyloxy)-p-
phenylenvinylene (MEH-PPV) or poly 2,5-dihexyloxy-p-phenylenevinylene (DH-PPV)
(Scheme 6) have been developed. These materials can be spin-coated from solution after
the conversion step. Another advantage of these PPV-derivatives is the possibility to
modify the electronic properties of the film with different substitution patterns. Therefore
all kind of organic substituents have been introduced into the aromatic system to alter
the structure of the aromatic buildin g block, including alkoxy-, alkyl-, cholestanoxy and
silicium containing groups [30–35] (Scheme 6).
ð6Þ
A precursor route not involving heteroatoms in the precursor polymers has also

been developed. It is based on the oxidation of soluble poly( p-xylylene)s to corresponding
PPVs by using stoichiometrical amounts of 2,3-dichloro-5,6-dicyano-1,4-benzochinone
(DDQ) (Scheme 7) but is restricted so far to a-phenyl-substituted poly( p-xylylene)s [36].
ð7Þ
Beside spin-coating-based preparation techniques, the so-called chemi cal-vapor-
deposition-route (CVD) has gained considerable attention as a solvent free preparation
process. Following this route, the starting materials are pyrolized after vaporization,
followed by CVD and polymerization of the monomers on the substrate. Finally, the
Copyright 2005 by Marcel Dekker. All Rights Reserved.
halogeno-functionalized poly( p-xylylene) is converted to PPV by polymer-analogous
thermoconversion (Scheme 8) [25,37,38].
ð8Þ
2. Polycondensation and C–C-Coupling Routes
Some drawbacks of the precursor routes mentioned above have been overcome by the
use of polycondensation- and C–C-bond-coupling reactions. To produce soluble PPV-,
poly(thiophene)-, or poly(pyrrol) derivatives for spin coating preparation, various types
of transition metal catalyzed react ions, such as the Heck-, Suzuki-, and Sonogashira-
reaction, Wittig- and Wittig–Horner-type coupling reactions, or the McMurry- and
Knoevenagel-condensation have been utilized.
A typical example of the Pd catalyzed Heck reaction of 1,4-dibromo-2-phenylbenzol
with ethylene to obtain the poly(phenylphenylene vinylene) [39] is depicted in Scheme 9.
A common drawback of this reaction-type is the insufficient regioselectivity, resulting in
1,1 diarylation of the product (>1%, depending on the substituents) [40].
ð9Þ
In order to avoid this problem, the Suzuki coupling is used as well to obtain various
substituted PPVs. Therefore an aromatic diboronic acid or ester and dibromoalkylene are
reacted in the presence of a Pd catalyst as depicted in Scheme 10 [41].
ð10Þ
Cyano derivatives of PPV with high oxidation potential are commonly
synthesized by Knoevenagel condensation of substituted terephthaldehyde with

Copyright 2005 by Marcel Dekker. All Rights Reserved.
benzene-1,4-diacetonitriles yielding an alternating copolymer type pro duct (see
Scheme 11) [42]
ð11Þ
Schlu
¨
ter et al. described the synthesis of soluble PPV derivatives from substituted
aromatic dialdehydes via McMurry-type polycondensation reaction. With this low valent
titanium catalyzed reaction (see Scheme 12), the obtained pro ducts are characterized by
a double bond cis/trans ratio of about 0.4 and an average degree of polymerization of
about 30 [43].
ð12Þ
Phenylic substituents at the vinyl ene positions — increasing both solubi lity of the
polymer and stability of the double bond — can be achieved by reductive dehalogenation
polycondensation of 1,4-bis(phenyldichlormethyl)benzene derivatives with chromium(II)-
acetate as reducing agent [44] (see Scheme 13).
ð13Þ
A further route leading to unsubstituted PPV was published by Grubbs et al. [45],
utilizing ring-opening olefin metathesis reaction as shown in Scheme 14. Starting from
Copyright 2005 by Marcel Dekker. All Rights Reserved.
bicyclic monomers with bicyclo(2.2.2)octadiene skeleton, the ring-opening metathesis
polymerization (ROMP) is performed with Schrock-type molybdenum carbene catalysts.
The obtained, well defined, nonconjugated soluble precursors, containing carboxylic ester
functions, are then thermally converted to the conjugated PPV.
ð14Þ
The Wittig reaction (see Scheme 15) is also a commonly used method for yielding
PPV derivatives from arylene bisphosphonium salts and bisbenzaldehydes. Since only
products of moderate molecular weight are obtained, more interest in this reaction is
given in the field of spacer segregated poly( p-phenylene vinylene)s with defined conjuga-
tion length (see III.A) [46]. An improvement concerning the degree of polymerization is

obtained by the Horner modification of the Wittig procedure (‘Wittig–Horner reaction’).
Following this route, the bisphosphonium salt is replaced by bisphosphonates or aromatic
bisphosphine oxide monomers [47].
ð15Þ
Due to the side chain induced twist within the main chain the effective conjugation
length is notably effected in soluble PPVs. A strategy to overcome this problem and to
develop more rigid conjugated systems has been presented by Davey and co-workers in
1995 who prepared poly(phenylene-ethynylene)-type polymers according to the following
scheme (Scheme 16) [48].
ð16Þ
Copyright 2005 by Marcel Dekker. All Rights Reserved.
3. Other Poly(phenylene-vinylene)s
Oligo- and poly(m-phenylene-vinylene) derivatives are not accessible via the polymeriza-
tion approach analogue to the Wessling- or Gilch-route. Accordingly, other methods
as the reductive dehalogenation polycondensation or the Wittig-type reaction as shown in
II.A.1 and II.A.2. are used for their formation. Despite their increased solubility, the
1,3-phenylene-units within the poly(m-phenylene-vinylene)s act as conjugation barriers so
that their usage in OLED techniques is very limited.
Oligomers of (o-phenylene-vinylene)s can be obtained using various C–C-coupling
and polycondensatio n methods. For higher oligomers and polymers, the Stille-type
coupling of 1,2-diiodobenzene or 1,2-bis(2-iodostyryl)benzene with bis(tri-n-butylstannyl)-
ethylene was introduced by Mu
¨
llen et al. [49] (see Scheme 17).
ð17Þ
The o-phenylene-vinylene-struct ure represent an intermediate case between the
p- and m-derivatives, allowing an extended p-conjugation and simultaneously disturbing
it by the non-planar geometry between the vinylene units. Utilizing three different alkyle
chains leads to the PPV-copolymer ‘‘Super Yel low’’ — commercially available from
Corion Organic Semiconductors GmbH — which shows the best efficiency and lifetime of

PPV-derivatives upto now (see Scheme 18):
ð18Þ
B. Heteroaromatic Systems
Heteroaromatic systems, such as the widely used poly(thiophene)s can be obtained by
simple oxidat ive polymerization of the soluble monomers or oligomers either by
electrochemical means or oxidizing agent such as FeCl
3
[50,51]. This common route is
also used to synthesize a variety of mono- and dialkyl-, -alkoxy-, and -alkylsulfonic acid
substituted and therefore soluble poly(thiophene)s [52–57 ] (Scheme 19) and can also be
utilized to obtain poly(pyrrole)s. The disadvantage of this polymerization methods
however is the regiorandom structure of the polyme ric product with non-reproducible
properties.
Copyright 2005 by Marcel Dekker. All Rights Reserved.
ð19Þ
For better defined poly(thiophene) structures a variety of organometallic mediated
synthesis have been introduced. Most widely employed are Grignard-type organo-
magnesium compounds in addition to a nickel catalyst. Highly regioregular head-to-tail
3-alkylpoly(thiophene)s are obtained following the synthetic route of McCul lough et al.
(see Scheme 20).
ð20Þ
Polymers — prepared via the polymerization of 2-bromomagnesio-5-bromo-3-
alkylthiophenes — exhibit enhanced conductivity and optical properties when compared
with regiorandom materials [58,59]. Another approach to regioregular alkylpoly(thio-
phene)s is the usage of zinc instead of magnesium in nickel- or palladium catalyzed
polymerizations [60,61]. Due to the improvements, these synthetic methods are by far the
most valuable synthetic routes to these materials. In contrast, the regioselective synthesis
of substituted poly(pyrrole)s was not reported to date.
Heterocyclic, electron deficient conjugated syst ems like poly(1,3,4-oxadiazole)s,
poly(1,3-oxazole)s and poly(1,2,4-triazole)s are applied in organic light emitting diodes as

electron transport and hole blocking layers. The synthetic strategies for their formation
are as manifold as the structures themselves, reaching from polymerization of functional
monomers to polymer analogue formation of the conjugated system (e.g., by ring closure
dehydration, dehalogenation, etc.). For further details is referred to the reviews of
Schmidt et al. [14] and Feast et al. [62].
C. Light Emitting Polymers (LEPs) Based on Polyfluorenes
A second important class of p-conjugated polymers are polyfluorenes, which were
obtained the first time by oxidative polymerization of 9-alkyl- and 9,9-dialkylfluorenes
with ferric chloride [63]. These polymers showed low molecular weight and some degree of
branching and non-conjugated linkages through positions other than 2 and 7.
A very success ful way to improve regiospecificity and to minimize branching
was the synthesis through transition-metal-catalyzed reactions of monomeric
2,7-dihalogenated fluorenes. The palladium-catalyzed synthesis of mixed biphenyles
from phenylboronic acid and aryl bromide discovered by Suzuki et al. [64] tolerates a large
variety of functional groups and the presence of water. This method can also be used to
prepare perfectly alternating copolymers.
Copyright 2005 by Marcel Dekker. All Rights Reserved.
1. Polyfluorene-Homopolymers
Polyfluorenes with alkyl substituents at C9 are soluble in conventional organic solvents
such as aromatic hydrocarbons, chlorinated hydrocarbons and tetrahydrofuran, which
made them useful to prepare thin films for OLEDs. As a consequence many efforts
have been undertaken to synthesize a large number of high-m olecular-weight, 9-mono-, or
disubstituted very pure fluorene-based polymers.
ð21Þ
9,9-Disubstituted 2,7-bis-1,3,2-dioxaborolanylfluorene is allowed to react with a
variety of dibromoarenes in the presence of a catalytic amount of (triphenylphosphine)
palladium (Scheme 21). The improved process yields high-molecular-weight polymers with
a low polydispersity (<2) in less than 24 h reaction time, whereas the conventional Suzuki
coupling process can take up to 72 h and more to deliver polymers of modest molecular
weights. Optimized LEDs based on these polymers, made by improved Suzuki poly-

fluorene chemistry, exhibited light emission exceeding 10,000 cd/m
2
with a peak efficiency
of 22 lm/W [3].
2. Polyfluorene Copolymers
The described synthesis of polyfluoren homopolymers allows also the design of alternating
copolymers. Instead of the 2,7-dibromofluorene a variety of dibromoarenes can be used
in the Pd-catalyzed C–C coupling polymerization reaction. An important group of
comonomers are tertiary aromatic amines, which have been known as excellent hole-
transport materials and have found many applications as photoconductors and in LEDs.
The resulting alternating copolymers are all blue emitters, exc ellent film formers and
show high hole mobilities [65]. These materials can be used as emitters as well as hole
transporters in LED devices.
This alternating copolymer concept has been extended to other conjugated mono-
mers as shown in Scheme (22). All synthesized copolymers [66] are of high molecular
weight, are highly photoluminescent and their emissive colours can be qualitatively
correlated to the extent of delocalization in the comonomers. For example the thiophene
copolymer emits bluish green light, but the bithiophene copolymer emits yellow light.
Copyright 2005 by Marcel Dekker. All Rights Reserved.
ð22Þ
No other polymer class offers the full range of color with high efficiency, low
operating voltage and high lifeti me when applied in a device. The polyfluorene-based
materials seem to be very viable for commercial applications. A special group within the
polyfluorenes are poly-spiro-deriva tives. They offer a wide range of accessible colours.
Compared to standard polyfluorenes they are morphologically more stable and do not
form aggregates as easily (see Scheme 23):
ð23Þ
D. Poly(p-phenylene)s
Poly( p-phenylene) (PPP) represents a wide class of interesting conjugated polymers for
PLED applications. To be exact, the formerly described polyfluorenes also belong to this

class of polymers. Wide bandgaps are typical for PPPs and allow emission of blue light.
Since the design of efficient long-lived blue emitters remains a significant challenge to the
field, polymers, such as poly-p-phenylenes, are attractive candidates for consideration.
As with the PPVs, most PPPs are characterized by their insolubility and infusibility,
properties that were a considerable hindrance towards structural charact erization and
processing. Thus research activities were directed to form PPP films via soluble thermally
converted precursor polymers on the one hand and the development of soluble, substituted
PPPs on the other hand.
First attempts to generate poly( p-phenylene) were undertaken by Kovacic et al. in
the 1960s [67]. He reported the oxidative treatment of benzene with copper(II)-chloride in
the presence of strong Lewis acids (e.g. aluminum trichloride) which led to a condensation
of the aromatic rings by forming radical cations as reactive intermediates. The benzene
units are preferentially connected in the 1,4-position, but crosslinking and oxidative
condensation to highly condensed aromates and a maximum degree of condensation of
about 10 make this reaction interest ing only for historical aspects.
1. PPPs by Transition-Metal-Catalyzed Condensation Reactions
The availability of newer, more effective methods for aryl–aryl coupling has been an
important driving force for the development of new synthetic strategies for PPPs and
other polyarylenes. Transition metal catalysis, such as the Pd(0)-catalyzed aryl–aryl
coupling developed by Suzuki [63] and nickel(0)-catalyzed or -mediated coupling
Copyright 2005 by Marcel Dekker. All Rights Reserved.
according to Yamamoto [68] have been employed most successfully. An example for the
Ni(0)-catalyzed coupling is the coupling of 1,4-dibromo-2-methoxycarbonylbenzene to
poly(2-methoxycarbonyl -1,4-phenylene) as a processable PPP precursor [69]. The aro-
matic polyester PPP precursor is then converted to carboxylated PPP and thermally
decarboxylated to PPP with copper(II)-oxide catalysts (Scheme 24).
n
ð24Þ
A second, very fruitful synthetic principle for structural ly homogenous, processable
PPP derivatives involves the preparation of soluble PPPs by the introduction of solubi-

lizing side groups. The pioneering work here was carried out in the late 1980s, when
soluble poly(2,5-dialkyl-1,4-phenylene)s were prepared for the first time [70]. The Suzuki
aryl–aryl cross-coupling method (Scheme 25), adapted to polymers by Schlu
¨
ter, Wegner
et al., made it possible to synthesize solubilized PPPs with a dramatically increased
molecular weight of up to 100 phenylene units.
ð25Þ
Soluble PPPs not only contain alkyl substituents, they were also synthesized with
alkoxy groups and with ionic side groups like carboxy and sulfonic acid functions, which
are able to form PPP polyelectrolytes [71].
It is also possible to synthesize chiral PPPs as Scherf et al. reported [72]. They are
composed of chiral cyclophane subunits, made by a Suz uki-type aryl–aryl cross-coupling
reaction of the corresponding diboronic acid and dibromo derivatives. The monomers
containing cyclic –O–C
10
H
20
–O– loops were separated into the pure enantiomers and used
to generate the corresponding stereoregular iso- and syndiotactic PPP-derivatives
(Scheme 26). The isotactic derivative possesses a chirality of its main chain.
ð26Þ
Copyright 2005 by Marcel Dekker. All Rights Reserved.
An important aspect concerning the electronic properties of the PPP is the influence
of substituents at the phenylene units. In unsubstituted PPP, there is a twist angle of 23

between adjacent phenylene units [73] . This seems to be significant, but the p-overlap
is a function of the cosine of the twist angle, so a fair amount of conjugative interaction
remains even at 23


. If substituents are placed along the PPP backbone (e.g., at the
2- and 5-positions), the solubility is enhanced, but the p-overlap is reduced dramatically.
The resulting twist angles reach from 60

to 80

depending on the length of the alkyl
substituents [74].
The described facts show the synthetic demands for being able to prepare process-
able, and structurally defined PPPs, in which the p-conjug ation remains nearly intact or is
even increased compared to the parent PPP system. To realize this principle it is necessary
to prepare structures in which the aromatic subunits could be obtained in a planar or only
slightly twisted conformation in spite of the introduction of substituents. One of the first
examples was the synthesis of polyfluorenes via oxidative coupling of fluorene derivatives
as described above [63]. Another possibility to reach this aim is the preparation of
‘stepladder’ PPPs.
Monomers like the 2,7-dibromo-4,9-dia lkyl-4,5,9,10-tetrahydropyrenes (Scheme 27)
represent suitable starting monomers for the realization of such ‘stepladder’ structures.
These difunctionalized tetrahydropyrene monomers were first prep ared by Mu
¨
llen et al.
[75] and reacted in a Yamamoto-type coupling [76]. Reaction of the dibromide with a
stoichiometric amount of a low-valent nickel(0) complex gave a poly(4,9-dialkyl-4,5,9,10-
tetrahydropyrene-2,7-diyl) (PTHP) as a new, completely soluble type of PPP derivative,
in which each pair of neighboring aromatic rings is doubly bridged with ethano linkages.
The solubilizing alkyl substituents are attached at such positions on the periphery of
the molecule that they cannot cause twisting of the main chain. The number-average
molecular weight was M
n
¼ 20,000, corresponding to 46 THP units.

n
ð27Þ
The luminescence characteristic of PTHP suggests that it is a potential candidate
for the active component in OLEDs. Investigations showed the appearance of a quite
intense blue-green electroluminescence with a quantum yield of up to 0.15% (single layer
construction ITO/PTHP/Ca).
The ‘stepladder’ concept can be logical continued towards a completely planar
ladder polymer to minimize the mutual distorsion of adjacent main chain phenylene units.
The complete flattening of the conjugated p-system by bridging all the phenylene subunits
should then lead to maximum conjugative interaction. As with the PTHP systems, alkyl
or alkoxy side chain should lead to soluble polymers. This idea was realized first in 1991
with the first synthesis of a soluble, conjugated ladder polymer [77]. The preparation is
according to a so-called classical route, in which an open-chain, single stranded precursor
polymer was closed to give a double stranded ladder polymer (Scheme 28). In the synthesis
Copyright 2005 by Marcel Dekker. All Rights Reserved.
of this LPPP, the precursor polymer is initially prepared by Suzuki aryl–aryl coupling of
an aromatic diboronic acid and an aromatic dibromoketone.
ð28Þ
The cyclization to structurally defined, soluble LPPP takes place in a two-step
sequence, consisting of a reduction of the keto group followed by ring closure of the
secondary alcohol groups in a Friedel–Crafts-type alkylation. The resulting ladder
polymer has an average molecular weight of 25,000, corresponding to 65 phenylene units.
LPPP is characterized by unusual electronic and optical properties as a consequence of
planarization of the chromophore. The absorption maximum undergoes a bathochromic
shift to a l
max
value of 440–450 nm for the p!p* transition compared to PPP with
l
max
¼ 336 nm [78]. The photoluminescence of LPPP in solution is a very intensive blue,

but the bulk properties are surprising different: Although efficient LEDs can be assembled,
the emission of the solid state film is yellow in the case of photoluminescence and
electroluminescence. In comparison to the former descripted PTHPs the quantum yield
is with ca. 1% much higher [79].
2. Other Routes to Poly( p-phenylene)s
Recently the most popular synthetic routes to PPPs are the transition-metal-catalyzed
condensation reactions discussed above, but several other syntheses were developed to
generate PPP and its derivatives.
About 40 years ago, Marvel et al. described [80] the polymerization of
5,6-dibromocyclohexa-1,3-diene to poly(5,6-dibromo-1,4-cyclohex- 2-ene), followed by a
thermally induced, solid state elimination of HBr with formation of PPP (Scheme 29).
The products, however, indicate some structural defects like incomplete cyclization
and crosslinking.
ð29Þ
More than two decades later, Ballard et al. developed an improved precursor route,
starting from 5,6-diacetoxycyclohexa-1,3-diene (Scheme 30), the so called ICI route
[81,82]. The soluble precursor polymer is then aromatized thermally to PPP via
elimination of two molecules of acetic acid per structural unit. The polymerization of
the monomer, however, does not proceed as a uniform 1,4-polymerization: beside the
Copyright 2005 by Marcel Dekker. All Rights Reserved.
regular 1,4-linkages about 10% of 1,2-linkages are formed as a result of a 1,2-
polymerization of the monomer.
ð30Þ
An improved precursor route to high molecular weight, structurally regular PPP
by transition metal-catalyzed polymerization of a cyclohexa-1,3-diene derivative to a
stereoregular precursor polymer was described by Grubbs et al. [83] and MacDiarmid [78]
et al (Scheme 31). The final step of the reaction sequence is the thermal, acid-catalyzed
elimination of acetic acid, to convert the precursor into PPP. They obtaine d PPP films of a
definite structure, which were unfortunately contaminated with large amounts of
polyphosphoric acid, the acidic reagent employed.

ð31Þ
Another possibility to receive PPP derivatives is the Bergman cyclization (Scheme
32), starting from substituted enediynes, e.g., 1-phenyl-hex-3-en-1,5-diyne, leading to
poly(2-phenyl-1,4-phen ylene). It is also possible to synthesize the structurally relat ed
poly(2-phenyl-1,4-naph thalene) from 1-phenylethynyl-2-ethynylbenzene [84,85].
ð32Þ
Although PPPs and its derivatives reveal extraordinarily high thermal and oxidative
stabilities, corresponding single-layer OLEDs exhibit only low electroluminescence
efficiencies. Higher efficiencies have been achieved by preparing polymer blends or by
virtue of two-layer OLED-constructions. External efficiencies up to 3% were determined
for an ITO/PVK/poly(2-decyloxy-1,4-phenylene)/Ca — OLED [86,87].
These recent achievements are the result of new and more efficient synthetic methods,
which permit chemo- and regioselective syntheses and allow molecular weights high
enough to cast films with good integrity resulting in reasonable efficiencies [88].
III. CONDUCTING POLYMERS WITH ISOLATED CHROMOPHORES
A major challenge in the last few years remained the development of materials that
combine the processability of polymers with the defined optical and electrical properties
of low molecular weight chromophores. Strategies to optimize electrooptical and
mechanical properties in OLEDs have generally implemented the use of chromophores
with defined conjugation length either inserted in the polymer backbone or alternatively
Copyright 2005 by Marcel Dekker. All Rights Reserved.
attached as pendant side groups. By manipulation of the primary chemical struc ture,
processing characteristics as well as optoelectronic properties, i.e. regulating the HOMO
and LUMO energy levels allowing fine tuning of charge injection properties can be
controlled.
A. Nonconjugated Polymers with Side Chain Chromophores
One approach to obtain electroluminescent material with defined conjugation length is to
link the chromophore as side chains to a nonconjugated polymer backbone. This concep t
was found to be of great advantage for the transformation of crystalline materials to
amorphorous ones and combines high loading efficiency without segregation of the

fluorophores. Moreover, as the synthesis occurs without any transition metal catalyst, that
can be difficult to remove and might act as quencher, the obtained polymers show high
purity grade.
Based on this approach, suitable monomers with defined electron or hole trans-
porting units were prepared and polymerized by different methodologies. Following this
strategy polymethacrylates, polystyrenes, polynorbornenes, polysiloxanes, polyethers, and
polyesters with different side chain chromophores have been synthesized.
Free radical polymerization of functionalized methacrylate monomers containing
charge transporting units has been reported by several groups [89–92]. Electron trans-
porting as wel l as hole transporting material have been synthesized by this method
(Scheme 33). A tri-functional copoly(methacrylate) bearing a blue emitting distyryl-
benzene chromophore, an aromatic oxadiazole electron transporting unit and a cross-
linkable cinnamate showing strong electroluminescence and good solution process ability
has been synthesized. Although the luminescent properties would be satisfactory for
application in electroluminescent devices, one of the major problems related to this
application, namely stability could not be achieved with polymethacrylate backbone
polymers. It was shown by Cacialli et al. [91] that both the methacrylate backbone and
the chromophore groups were susceptible to alteration processes.
ð33Þ
Copyright 2005 by Marcel Dekker. All Rights Reserved.
Styrene derivatives represent a second important class of monomers that can
be polymerized by free radical polymerization. By doing so, copolymers with TPD and
oxetane side chains were obtained where the oxetane group was used for covalent cross-
linking [93] (Scheme 34). The resulting film showed excellent mechanical integrity proving
this method to be very useful to improve mechanical properties. Moreover, oligo(PPV)s
[94] or electron transporting units such as oxadiazoles [95] were also introduced as pendant
groups with a polystyrene backbone.
ð34Þ
A second possibility to attach the chromophores to the polymer backbone is the
polymer analogue modification with the appropriate electron or hole transporting unit.

Complete substitution of poly( p-acetoxystyrene) was achieved through Williamson
condensation with chloromethylstilbene resulting in a polymer with M
n
¼ 22,000 Da,
M
w
/M
n
¼ 2.1 and T
g
¼ 71

C (Scheme 35 [96]).
ð35Þ
Copyright 2005 by Marcel Dekker. All Rights Reserved.
A similar synthetic concept has been used to attach pyrene emitting groups on
poly(methylhydrosi loxane) by hydrosilation in the presence of chloroplatinic acid
[89,97]. (Scheme 36). The reaction was complete after one week giving a polymer with
M
n
¼ 6800 Da with an emission wavelength of 500 nm in a single-layer device, appearing
light blue.
ð36Þ
Moreover, dimethylsilane modified oxadiazole units were used to functionalize
polystyrene-block-polyisoprene copolymers via Heck reaction or hydrosilation. The
degree of side chain functionalization varied from 4–44% [98].
Anionic polymerization of styrene bearing TPD-like side chains has also been
reported [99] (Scheme 37). A wide range of hole transporting polymers were prepared
with high T
g

s ranging from 132–151

C to increase thermal and long-term stability of the
device.
R
1
¼ FR
2
¼ CH
3
ð37Þ
The ring-opening metathesis polymerization of side chain functionalized norbornene
monomers was applied to obtain electroluminescent polymers (Scheme 38) [100,101].
Homo- and copolymers with 25 or 50 repeat units were prepared with M
n
ranging from
19,400–53,000 Da and M
w
/M
n
¼ 1.02–1.04. The color of emission could be fine tuned by
Copyright 2005 by Marcel Dekker. All Rights Reserved.
varying the monomer ratio.
ð38Þ
Poly(norbornenes) containing pendant triarylamines have also been synthesized
by ruthenium catalyzed ring-opening metathesis polymerization [102]. By varying the
polarity and the length of the linker between polymer backbone and the triarylamine
functionality the device characteristics could be tuned (Scheme 39).
ð39Þ
Copyright 2005 by Marcel Dekker. All Rights Reserved.

The substitution of ester groups by less polar ether functionalities enhances thereby
external quantum efficiencies, lowers the operating voltage and impro ves the stability of
the device. Further improvement is obtained by reducing the length of the alkyl linker.
A further interesting concept is the synthesis of a polymer having a fully conjugated
backbone and pendant side-chain chromophores, combining electron-transport, hole-
transport and light emitting properties in a single polymeric material. Hybrid polymers of
this type with oxadiazole side chains were obtained by Bao et al. [103] via Heck
(Scheme 40) or Stille reaction (Scheme 41). The polymers prepared via the Heck reaction
showed a molecular weight M
n
¼ 28,500 Da with a polydispersity of 3.65 whereas the Stille
coupling resulted in lower molecular weight polymers of 8100 Da with polydispersities
of 1.67. The polymer with a PPV-like backbone indicated yellow-orange light emission
whereas polymers containing a thiophene group emit red-orange light. Both polymers
showed better external quantum efficiencies and better charge injection properties
compared to those without any side cha in.
ð40Þ
Copyright 2005 by Marcel Dekker. All Rights Reserved.
ð41Þ
B. Main Chain Polymers with Defined Segmentation
The effective conjugation length in main chain polymers can be controlled in two ways,
either by steric hindrance of p-conjugated segments that are associated in a non-coplanar
way or by introducing conjugation interrupters. Both concepts have been used frequently
in the past decade to gain better control over the specific device properties, including
electroluminescence, photoconductivity and photovoltaic effects.
Segmented conjugated polymers have the advantages over fully conjugated ones that
their electronic properties are independent of the degree of polymerization and can be
easily tuned by varying the substituent s or the conjugation length.
1. Conjugated Main-chain Polymers with Twisted Conformation
Since the p-overlap is a function of the cosine of the twist angle of adjacent aromatic units

two approaches have been used in the past to control the effectice conjugation length,
either by changes in the polymer geometry or topology. A twisted conformation of the
polymer backbone was achieved by introduction of alkyl or alkoxy substituents in 2,5
positions along the PPP backbone causing twist angles of 60–80

[104].
A second powerful approach are meta linkages in PPV (Scheme 42) that led also
to the interruption of conjugation due to the non-coplanar arrangement of adjacent
conjugated repeat units [105,106]. This concept was also applied to prepare chiral
polymers using 1,1-binaphthyl units [107,108] (Scheme 41) or conjugated polymers with
anthracene units [109–111].
ð42Þ
Copyright 2005 by Marcel Dekker. All Rights Reserved.
ð43Þ
2. Main Chain Polymers with Non-conjugated Interrupters
More important from a synthetic point of view, however, is the concept of non-
conjugative interrupters. In such a polymeric material the optoelectronic properties can
be tailored by the proper selection of the chromophore unit whereas the physical
properties can be adjusted by the non-chromophoric part. Burn et al. [112] introduced
the notion of isolated chromophores in 1992 by selectively eliminating one of two
leaving groups of the precursor polyme r to give a conjugated–non-conjugated polymer.
The design of such polymers with controlled chromophore length is quite a
large area, as chromophores and spacers can be combined in an almost infinite way.
Chromophores can be hole-transporting triarylamine or electron-transporting ones
such as oxadiazol es, with ethers, esters [113], amines, amides, imides [114–116], silanes,
fluorenylidene as possible con jugation interrupters [117,118].
Flexible alkyl spacer through ether linkage were introduced by Wittig reaction
between the corresponding aldehyde and triphenylphosphonium salt. This has been
used for example to synthesize various polymers where the oligomeric PPV segments
are separated by polymethylene spacers, which show blue or blue-green emission

(Scheme 44) [119].
ð44Þ
Copyright 2005 by Marcel Dekker. All Rights Reserved.
Polymers with other substituents [117,120] or similar structures [121,122] were
also synthesized by this method.
Hadziioannou et al. introduced a dialkylsilyl spacer between PPV-like segments [123]
via Heck reaction (Scheme 45), where R is an alkoxy or an alkyl group. For a ¼ 1,
the polymer can be obtained via Wittig reaction between the silicon containing
dialdehyde and the appropriate triphenylphosphonium halide [124].
ð45Þ
The same spacer has also been used to interrupt the conjugation between
oligothiophenes [125,126]. In this case, the insertion of the dialkylsilyl group occurs
via a Wurtz coupling or a palladium-catalyzed polycondensation. The general structure
of the obtained polymers is depicted in Scheme (46) where R represents an alkyl chain.
ð46Þ
This class of polymers shows not only excellent environmental stability but can
also be easier processed with the addition of long, flexible hydrocarbon side chains in
comparison to most of the conjugated polymers. Moreover, the electroluminescence
characteristics in single-layer devices indicated a strong dependence on the number
of thiophene monomers and were found to be gradually shifted from green to red with an
increasing number of thiophene units from three to seven.
Fluorenylidene linkages have been widely used as conjugation interrupters due to
their stiff conformation and to the ability to increase solubility of the resulting polymer
and were therefore incorporated in several PPP based polymers by the Suzuki poly-
condensation reaction [127]. Following this concept a blue emitting polymer based on
Copyright 2005 by Marcel Dekker. All Rights Reserved.
fluorenylidene segmented oligo( p-phenylene)s was synthetized by Nuyken et al.
[128,129] as described in Scheme (47).
ð47Þ
The fluorenylidene linker concept was also applied to the synthesis of triarylamine

segmented polymers via a Hartwig–Buchwald reaction [132] (Scheme 48).
Dba=dibenzylidene aceton
ð48Þ
Moreover, Miller et al. introduced fluorenylidene linker as well as other conjugation
breakers such as hexafluoropropylidene [127] via the nickel-mediated Yamamoto reaction
[133] as described in Scheme (49).
Copyright 2005 by Marcel Dekker. All Rights Reserved.
COD=cyclooctadiene; bpy = bipyridine
ð49Þ
Two copolymers were synthes ized with different ratios of the fluorene-based
monomer and the fluorenylidene linker. Both polymers wer e obtained in high yields
and high molecular weight with M
w
¼ 55,000–89,500 Da and exhibited excellent thermal
stability. Glass transition temperatures ranged from 153

C to 197

C with decomposition
temperature (5% weight loss measur end by TGA analysis) of 440 to 450

C.
Wegner et al. developed a class of polymers containing sequences linked by
ethylene, vinylene or ethynylene groups [130]. PPV-like polymers containing an
adamantane spacer group (Scheme 50) [131] are also accessible via the palladium
catalyzed Suzuki reaction. The polymers possessed excellent thermal stability and shows
an onset of thermal decomposition temperature at 362

C and a T
g

at 151

C due to the
chain rigidity.
n
ð50Þ
IV. CONCLUSIONS
In the past decade a great deal of attention has been focused on the preparation and
characterization of p-conjugated oligomers and polymers due to their potential
application as novel materials for optoelectronics. Initial attempts have been employed
oxidative polymerization methods which led very often to low molecular weight
compounds with poor defined polymer structure. The availability of newer, more efficient
Copyright 2005 by Marcel Dekker. All Rights Reserved.