Vol. 3

LIGHT-EMITTING DIODES

87

LIGHT-EMITTING DIODES

Introduction

The role of polymers in light-emitting diodes (LEDs) traditionally has been asso-
ciated with the use as a structural material. Epoxy resins are routinely applied
as the transparent “packaging material” in conventional LEDs in which light is
generated at the junction of two inorganic semiconductors. The polymer provides
the structural integrity of the device and serves as the optical medium that en-
ables the extraction of light (1). During the last decade, however, LEDs have been
developed in which the semiconducting, electroluminescent (EL) material itself
is a polymer. These devices are usually referred to as polymer LEDs or PLEDs
and have attracted significant attention in both academic and industrial environ-
ment. The principal interest in the use of EL polymers is based on their promise
to combine the ease of processability and mechanical flexibility of macromolecular
materials with the exceptional, readily-tailored properties of organic semiconduc-
tors. PLEDs offer the prospective of low production costs, light weight, large area,
fast switching times, low power consumption, high brightness, large viewing an-
gle, and crisp colors. Thus, the new technology may feature distinct advantages
over existing flat-panel display devices such as conventional emissive displays, in-
organic LEDs, and liquid crystal displays (LCDs). As a result, the list of potential
applications of PLEDs ranges from small, low information-content displays for
simple electronic devices such as pagers and portable phones to high resolution
video-rate displays for use in desktop monitors and ultrathin television sets. In ad-
dition, PLEDs may be employed as flat lighting devices, for example in automotive
rear lights or as backlights in LCDs.

Since the first demonstration of a PLED in 1990 (2), the field has progressed

with an almost unbelievable pace, and the technology has matured to the edge

Encyclopedia of Polymer Science and Technology. Copyright John Wiley & Sons, Inc. All rights reserved.

88

LIGHT-EMITTING DIODES

Vol. 3

of commercial exploitation (3). In view of the over 5000 publications on PLEDs
published to date, it is clearly outside the scope of this article to review this excit-
ing progess in detail. Rather, a comprehensive overview focusing on the general
concepts, as well as the current state-of-the-art, is given. For a more detailed de-
scription regarding the chemical, physical, and engineering aspects of PLEDs, the
reader is referred to a selection of excellent reviews which have appeared in the
recent past (4–11). It should be noted that the discussion of LEDs based on low
molecular weight organic emitters has been omitted here for apparent reasons.
However, the latter share many of the features of PLEDs (12,13), and despite
the more demanding device manufacturing processes of low molecular weight or-
ganic LEDs, their technological application is slightly more advanced than that
of PLEDs (3–5,9).

Device Physics and Device Design Considerations

An LED is an electrooptical device that exploits the EL effect of a fluorescent,
semiconducting material and emits electromagnetic radiation in the ultraviolet,
visible, or infrared regime of the electromagnetic spectrum when an electric cur-
rent is applied (1) (see LEDs (Light Emitting Diodes)). Conventional LEDs, com-
mercially introduced in the 1960s, are based on inorganic semiconductors such as
GaAs, GaP, AlGaAs, GaAsP, and InGaP. Electroluminescence in organic materials
was originally discovered in the early 1960s (14), but useful devices were demon-
strated only some 20 years later (12,13). Electric-field-induced light emission from
a fluorescent, semiconducting polymer was first observed in 1990, when an elec-
trical field was applied to a thin film of poly(p-phenylene vinylene) (PPV) (2).

The LED reported then materialized the most simple device configuration of

a PLED which, in accordance with the number of organic layers used, is usually
referred to as a single-layer device (Fig. 1). The semiconducting polymer film
(typical thickness 50–500 nm) is sandwiched between two electrodes, of which
at least one has to be (semi)transparent in order to allow the generated light
to escape from the device. Adopting a standard Schottky-configuration, a high
work function anode such as semitransparent indium tin oxide (ITO) and a low
work function cathode such as Al, Mg or Ca are employed, leading to nonlinear
rectification and, thus, rendering the device into a diode. If an external voltage,
or bias, is applied, oppositely charged carriers—holes and electrons—are injected
above a threshold voltage from the electrodes into the conduction band (CB) and
the valence band (VB), respectively, of the semiconducting polymer (Fig. 2). These
electronic levels correspond to the electron affinity (EA) and ionization potential
(IP) of the polymer. Because of the applied electric field (typically of the order

Vol. 3

LIGHT-EMITTING DIODES

89

h

ν

Transparent substrate (glass, polymer)

Anode (indium tin oxide)

EL polymer

Cathode (low-working function metal)

h

ν

Transparent substrate (glass, polymer)

Anode (indium tin oxide)

Hole-transporting polymer

Electron-transporting EL polymer

Cathode (low working function metal)

Fig. 1.

Schematic drawings of the configurations of a single-layer PLED (top) and a two-

layer PLED (bottom).

EA

Anode

IP

Cathode

⌽

C

⌽

A

VB

CB

EG

EV

Fig. 2.

Simplified schematic representation of the electronic energy levels in a single-

layer PLED. CB and VB are the conduction band and valence band, respectively, of the
semiconducting polymer, which equal to the ionization potential (IP) and electron affinity
(EA) relative to vacuum level (EV). The work functions (or Fermi levels) for anode (



A

) and

cathode (



C

) and the band gap (EG) are also indicated.

of 10

6

V/cm, corresponding to an operating voltage of a few volts), these carriers

travel through the polymer toward the oppositely charged electrode. On their
way, the charge carriers will either recombine within the emitting polymer layer
with an opposite charge and form singlet and triplet excited states (excitons), or
they reach the opposite electrode and discharge without having contributed to
the EL effect. The singlet excited states created through charge-recombination
are identical with those generated through photoexcitation, and, as in the case
of fluorescence, they relax to the electronic ground state under the emission of
light. The color of the latter is governed by the energy difference between the
singlet excited state and the electronic ground state, also referred to as the band

90

LIGHT-EMITTING DIODES

Vol. 3

gap (EG) of the EL polymer. The triplet excited states, by contrast, decay through
nonradiative processes.

The theoretical internal quantum efficiency of a PLED,

η

int

, is defined as the

number of photons generated per injected electron:

η

int

= η

PL

η

r

χ

(1)

η

int

is primarily governed by the (solid-state) fluorescence quantum efficiency

of the polymer,

η

PL

, which can exceed 65% (15), the fraction of injected charge

carriers that form excited states,

η

r

, and the fraction of singlet excited states

created (as opposed to triplets),

χ. Unfortunately, spin statistics dictate a χ of

only 1/4, although recent work suggests that this limitation can, in principle,
be overcome (16). To make things worse, internal reflection effects, which are
governed by the device’s geometry and the refractive index of the light-emitting
polymer (n), allow only a fraction of the generated light to be extracted from the
device. Thus, in a typical device configuration, the external quantum efficiency of
a PLED,

η

ext

, is lower than

η

int

by a factor of 2n

2

(17):

η

ext

= η

int

/2n

2

(2)

Approximating the refractive index of the EL polymer with a value of 1.4

allows to put up an upper limit of about 5% for

η

ext

, which still compares favor-

ably with the typical performance of conventional inorganic LEDs (0.2–8%) (1). It
should be noted that special geometries may be used to improve the light extrac-
tion from the device (18). For engineering purposes, it is often desirable to express
the performance of an LED through its power efficiency

η

pow

:

η

pow

= η

ext

E

/U

(3)

η

pow

is the optical power output divided by the electrical power input, and is

calculated from

η

ext

using the applied voltage U and the average energy E of the

emitted photons. Since the human eye is usually the detector for visible LEDs,
the key performance parameter of an LED is the luminous efficiency

η

lum

(units:

lm/W) which is a product of

η

ext

and the relative eye sensitivity curve S which has

been defined by the Commission Internationale de L’Eclairage, CIE):

η

lum

= η

ext

S

(4)

Reference values that are frequently used to evaluate the performance of PLEDs

are the luminous efficiencies of conventional inorganic LEDs (0.15–20 lm/W) or
unfiltered incandescent lamps (17 lm/W) (1). Finally, the absolute brightness of an
LED, or luminance (usually quoted in cd/m

2

), and luminance per current density

(cd/A) are also frequently used parameters to describe the performance of LEDs.
Here, the brightness of a laptop computer (200 cd/m

2

) or of a TV cathode ray tube

(500 cd/m

2

) are often employed as benchmarks (19).

As can be deduced from equation (1), the optimizations of charge-injection

and charge-transport (both of which directly influence the fraction of injected
charge carriers that form excited states,

η

r

) are important aspects for the design

Vol. 3

LIGHT-EMITTING DIODES

91

of efficient PLEDs. In order to allow for efficient charge-injection and, concomi-
tantly, low operating voltages, the work functions of cathode and anode should
match the EA and IP (or in semiconductor terminology, conduction band and va-
lence band) of the semiconducting emitter, respectively, as closely as possible.
Unequal barrier heights at the two electrode–polymer junctions are undesirable
because one carrier will be injected preferentially, leading to an increased driv-
ing voltage and a reduced luminous efficiency. At the other hand, most organic
semiconductors are p-type conductors, allowing holes to be transported more effi-
ciently than electrons. Consequently, substantial research efforts have been made
regarding the identification of optimal electrode–polymer combinations. A gen-
eral possibility, of course, is the chemical modification of the emitting polymer,
with the intent to tailor—besides the color-determining band gap—its EA, IP,
and/or charge-transport characteristics. Another strategy is the design of two-
layer (or even three-layer) devices in which, for example, a hole-transporting poly-
mer and an electron-transporting, light-emitting polymer are combined (7). The
resulting heterojunction between the two semiconducting polymers allows to con-
trol the charge-injection rates and helps to confine the charge recombination to
the emitting layer (Fig. 1). In addition, this architecture effectively reduces prob-
lems associated with the migration of oxygen and metal ions from the electrodes
into the organic layer and the related creation of quenching sites in the emission
zone. Consequently, the two-layer architecture has become the preferred PLED
embodiment.

Materials

The design of PLEDs relies on a variety of different materials, including a trans-
parent substrate, the (semi)transparent anode, a hole-transporting layer in case
of the usual two-layer structure, the EL polymer itself, an additional electron-
conducting/hole-blocking layer in case of a three-layer device, the cathode, and
finally some additional packaging material(s).

Polymeric materials such as polyester-based films have initially been her-

alded as desirable substrates because of the unique potential to create mechan-
ically flexible displays (20). However, moisture and oxygen have to be rigorously
excluded from the device so as to prevent rapid degradation of the EL polymer
and the cathode. Thus, in view of the notoriously poor barrier characteristics of
polymeric substrates, glass is the current substrate and packaging material of
choice because of its unmatched barrier characteristics. Nevertheless, recent stud-
ies have shown that the barrier characteristics of polymeric substrates can greatly
be improved by applying a coating of silicon nitride or silicon oxynitride through
a vapor deposition process (21).

Essentially all PLED devices described in the literature rely on ITO as the

high work function anode material, although a number of alternative transparent
conducting oxides have become available (22). ITO offers good electrical conduc-
tivity and a transmittance of

>85% at a film thickness of <150 nm, and allows the

generated light to escape from the device. ITO-coated glass is routinely used as
the electrode material in LCDs and a variety of other applications and is readily
available at comparably low cost. Most importantly, ITO offers a work function

92

LIGHT-EMITTING DIODES

Vol. 3

Fig. 3.

Chemical structures of poly(vinylcarbazole) (PVK), poly(aniline) (PANI), and

poly(3,4-ethylene dioxy thiophene) (PEDOT) which are frequently used as the hole-
transport layer in two-layer devices.

of between 4.5 and 5.0 eV (relative to vacuum level), which matches the IP of
frequently used conjugated organic polymers and allows for reasonably efficient
hole-injection. Unfortunately, the electronic characteristics of ITO depend quite
strongly on the manufacturing conditions and cleaning treatments. In addition,
its surface structure can vary substantially, making the use of ITO often rather
cumbersome.

As mentioned before, the two-layer architecture has become the preferred

PLED embodiment (Fig. 1), thus requiring a first polymer layer with efficient
hole-transport characteristics (6). Poly(vinylcarbazole) (PVK) is an extensively
investigated, commercially available hole-conductor, which has been frequently
used in this context (Fig. 3). Alternatively, polymeric triphenyldiamines have been
employed (23). More recently, poly(aniline) (PANI) and poly(3,4-ethylene dioxy
thiophene) (PEDOT) which are commercially available from Ormecon and Panipol
Ltd. and Bayer AG, respectively, have found widespread application instead (4,10).
Used in their electrically conducting form, these polymers not only allow efficient
hole-injection and hole-transport, but also offer an effective electron transport. As
a result, the electric field across this layer is dramatically reduced, limiting the
undesired migration of charged species from the anode through the device (see
LEDs (Light Emitting Diodes)).

Light generation is achieved in the actual light-emitting layer which, in a

typical two-layer configuration, usually also acts as an electron-transport ma-
terial; the design considerations for EL polymers are discussed in detail below,
together with an overview of typical material systems.

Finally, a low work function cathode is required for electron-injection. A va-

riety of different cathode materials have been employed, including Ca, Li, Mg, Al,
Cr, or alloys of these metals, sometimes also combined with Ag. Since electron-
injection is often a limiting factor for the device efficiency, the metal with the lowest
work function of the above series (Ca, 2.9 eV) is a preferred option. However, while
Ca-based devices may work well in the laboratory, the practical application of this
metal is stifled by its sensitivity toward oxidation. Changing the cathode mate-
rial to Mg (work function 3.7 eV) or Al (4.3 eV) leads to improved device stability,
however, usually at the expense of an increase in driving voltage and a reduced
device efficiency.

Vol. 3

LIGHT-EMITTING DIODES

93

Development of EL Polymer Systems.

The characteristics of the emis-

sive layer are, of course, of predominant importance for optimum device perfor-
mance. Thus, the design of light-emitting polymers has received particular atten-
tion during the last decade. In view of the abundance of EL polymers that have
been synthesized and investigated to date (4–11), it is far beyond the scope of
this review to provide a comprehensive overview. Consequently, a number of se-
lected examples are presented which illustrate the various design concepts, are
based on the most commonly employed building blocks, and are relevant from an
application point of view. Important design considerations address the need for

(1) high intrinsic photoluminescence efficiency
(2) high purity (elimination of traps which provide nonemissive decay paths

for excitons)

(3) adequate charge-transport characteristics (low barrier for charge-injection,

reversible formation of charge carriers, balanced charge transport)

(4) appropriate color of emitted light
(5) high optical transparency (no re-absorption or scattering of the emitted

light)

(6) good processability
(7) high film quality (homogeneous thickness of typically 50–500 nm, no pin-

holes or cracks, smooth surface, compatibility to adjacent layers)

(8) high thermomechanical stability (no softening under operating conditions)
(9) long lifetime (high thermal, light, oxidative, and environmental stability)

(10) low price

This list of fundamental requirements has indeed been a veritable challenge

for materials design in terms of synthesis, understanding of structure–property
relationships, and processing. The majority of the vast number of polymeric EL
materials investigated to date origins from the family of conjugated polymers, ie,
macromolecules featuring an extended

π-conjugated backbone (Figs. 4a and b).

Important examples of classes of conjugated polymers used as emitting layer in
PLEDs (Fig. 5) include poly(arylene)s, poly(arylene vinylene)s, and, to some ex-
tent, also poly(arylene ethynylene)s (24). Unfortunately, the very same conjuga-
tion which renders these macromolecules semiconducting is also the origin of the
difficulties that were initially encountered in attempts to process the early, more
simple forms of these polymers into useful shapes, ie, high quality films. These
difficulties are directly related to the fact that conjugated polymers are relatively
rigid macromolecules which exhibit high melting enthalpies and low melting en-
tropies, resulting in melting temperatures that are typically far beyond thermal
decomposition. In addition, the nonpolar chemical nature of many conjugated poly-
mers permits only modest enthalpic interactions with potential solvents, which
prevents simple solution-processing. These problems have been overcome by a
number of different design strategies which, naturally, follow those already suc-
cessfully applied for nonconjugated rigid-rod polymers (25). The latter include (1)
the use of so-called precursor routes; (2) the derivatization of

π-conjugated poly-

mers with flexible (or alternatively bulky) side chains which increase the entropy

94

LIGHT-EMITTING DIODES

Vol. 3

Fig. 4.

Schematic representation of typical EL polymer architectures. (a) “Plain,” con-

jugated main chain, (b) hairy-rod structure, (c) semiflexible, “segmented” structure, and
(d) nonconjugated polymer with conjugated side chains.

,

π-conjugated moiety;

,

flexible segment; and

, solubilizing side chain.

Fig. 5.

Chemical structures (generalized) of different classes of conjugated polymers used

as EL materials.

gain upon dissolution (and also reduce interchain interactions) and, thus, render
the polymer soluble (Fig. 4b); (3) the introduction of flexible “spacer” units between
the conjugated moieties which reduce the overall rigidity of the polymer backbone,
again leading to improved solubility (Fig. 4c); and (4) the use of nonconjugated,
processable polymers which feature the conjugated moieties in the form of side
chains (Fig. 4d) or (5) as guest molecules which are not covalently connected to
the matrix polymer but are rather dissolved or dispersed in the latter.

The so-called precursor route represents a particularly attractive way to

synthesize and process underivatized

π-conjugated polymers which were initially

employed in many of the early studies on PLEDs, including PPVs and poly(p-
phenylene)s (PPPs) (26,27). The concept relies on the three steps of synthesiz-
ing tractable, nonconjugated “precursor” macromolecules comprising appropri-
ately selected leaving groups, processing of these precursor polymers according
to standard (mostly solution-based) techniques, and converting the precursor

Vol. 3

LIGHT-EMITTING DIODES

95

polymer into the final (intractable) conjugated species by (usually heat-induced)
abstraction of the leaving group. A classical example is the synthesis of yellow-
green-light-emitting PPV (see structure above) (28), which has been used and op-
timized by a variety of research groups (26,29). A special version of the precursor
route that has received considerable attention in connection with the prepara-
tion of PPV-based LEDs makes use of a chemical vapor deposition (CVD) process
(8,26). Here, the starting material is vaporized under high vacuum conditions
and pyrolized to result in suitable monomers, which subsequently undergo CVD
and polymerization on a substrate. This solvent-free process usually results in
high quality films of a precursor-PPV, which can be subsequently converted to
the conjugated form. However, all precursor methods suffer from major difficul-
ties in that (1) the conversion from precursor to

π-conjugated polymer is often

not quantitative (leading to defects which may interrupt the conjugation in the
polymer backbone), (2) the removal of low molecular weight conversion products
(which may adversely affect some of the polymer’s properties) is not trivial, (3) the
precursor polymers often exhibit a limited oxidative stability, and (4) generally
undesirable side reactions may occur. On the other hand, the process also allows
to tune the material properties to a certain extent. For example, partial conver-
sion of precursor/

π-conjugated polymer results in segmented polymers (Fig. 4c)

with improved performance and modified color of the emitted light, as described
below.

A different approach is the design of inherently soluble, “hairy-rod” EL

polymers according to the design strategy outlined in Fig. 4b. The latter allows
film processing by standard solvent-based methods, eg, spin- and dip-coating,
screen- and ink-jet printing, as well as Langmuir Blodgett (LB) techniques.
The electronic properties of these macromolecules are primarily governed by
the chemical structure of the conjugated polymer backbone itself, but a num-
ber of additional tools can be employed to further manipulate band gap (emission
color), charge-injection/charge-transport characteristics, thermomechanical sta-
bility, thermal/oxidative stability, and other relevant properties (4,6,8,11). The
incorporation of electron-deficient aromatic rings and the integration of electron-
withdrawing or electron-donating substituents into the conjugated system are
illustrative examples of how the electronic characteristics can be varied over a
wide range. Another important general tool for property tailoring is the control
of the effective conjugation length through the introduction of side chains which
exhibit steric interactions and force the backbone to twist, or by using copolymers
that comprise well-defined conjugated segments that are separated by aliphatic
“spacer” moieties (Fig. 4c). Such segmented conjugated polymers (Fig. 4c), non-
conjugated polymers with conjugated side chains (Fig. 4d), and also blends of non-
conjugated matrix polymers and low molecular weight EL compounds (in which
the crystallization-induced segregation of the EL guest molecules and the poly-
meric matrix poses a serious problem) were reviewed extensively in the literature
(4,6,8,11,26). The nonconjugated spacer moieties disrupt the electronic conjuga-
tion, concomitant with a shift of the band gap toward higher energies, ie, into the
blue regime of the visible spectrum, which with extended conjugated structures
has proven to be difficult to access. Moreover, confinement of excitons is achieved
by stifling the migration of the latter to quenching sites, which usually leads to

96

LIGHT-EMITTING DIODES

Vol. 3

Fig. 6.

Chemical structures of poly[2-methoxy-5-(2



-ethylhexyloxy)-p-phenylene viny-

lene] (MEH PPV) and poly(9,9-dioctyloxy fluorene).

an appreciable enhancement of luminescence efficiency. However, as mentioned
heretofore, the nonconjugated moieties disrupt the electronic conjugation, sug-
gesting intrinsic limitations with respect to charge transport (it should be noted
in this context that also polymers with a fully conjugated backbone structure fea-
ture only a finite effective conjugation length which in case of PPV, for example,
extends over less than 10 repeat units, cf Ref. 8).

From the multitude of organo-soluble conjugated EL materials investigated

to date, two classes of materials seem to have shown the most promising overall
matrix of properties, at least from an industrial point of view, namely derivatized
PPVs and poly(fluorene)s. The most widely studied soluble PPV derivative is the
orange-light-emitting poly[2-methoxy-5-(2



-ethylhexyloxy)-p-phenylene vinylene]

(MEH PPV) which was introduced in 1991 (30). It was the first soluble light-
emitting material to be used in a PLED (Fig. 6) (31). The alkyloxyy substituents
not only render the polymer soluble but also shift the EL emission to higher wave-
lengths when compared to the parent PPV. A wide variety of poly(2,5-dialkoxy-p-
phenylene vinylene)s and other 2,5-derivatized PPVs have since been investigated
(4,6), and a few PPV derivatives, including a phenyl-substituted-PPV, have been
developed for commercial exploitation for example by Covion and are used in the
pilot production of Philips Electronics and Uniax Corp. (32). Besides the PPVs,
poly(9,9-dialkyl fluorene)s (Fig. 6) have recently emerged as an extremely promis-
ing class of polymeric light-emitting materials (10), and a series of poly(fluorene)
homo- and copolymers have been developed by the Dow Chemical Co. (32).

Towards Technological Exploitation

Only a decade after the first demonstration of electroluminescence in polymeric
materials, the performance of PLEDs now appears to meet many of the require-
ments for applications at the low performance end of the display market, and
the field has matured to the point of commercial exploitation; the most current
overview of the state-of-the-art can be found in an outstanding review (4). A num-
ber of electronics and polymer companies, including Cambridge Display Technol-
ogy Ltd., Covion, Dow Chemical Co., Philips Electronics, and Uniax Corp., have

Vol. 3

LIGHT-EMITTING DIODES

97

disclosed prototypes or pilot products that offer a luminous efficiency

η

lum

between

10 and 22 lm/W at an (average) brightness between 100 and 1000 cd/m

2

(4,10,32).

Thus, these devices clearly meet the brightness requirements for many display
applications and also match the efficiency of conventional inorganic LEDs incan-
descent lamps. However, it should be noted that bright and efficient devices are
primarily available in the green range of the emission spectrum, while presently
available blue- and red-light-emitting materials result in devices which are sub-
stantially less bright and efficient (4,7). Thus, substantial progress regarding the
availability of different colors is required before full-color devices based on red,
green, and blue pixels can be made available.

Probably, one of the most critical aspects from an application point of view

is that of device lifetime. Although the requirements for the latter strongly de-
pend of the particular application, operating lifetimes of

>10,000 h are usually

targeted (4,5), often at elevated temperatures. Detailed studies conducted at a
variety of laboratories (4,10) have indeed demonstrated that this requirement
can, in principle, be met when adequately packaging the device. However, most
of the “successful” lifetime studies were conducted under constant current con-
ditions, and a gradual rise of the driving voltage was reported to maintain the
initial current and brightness. This situation not only suggests an irreversible
degradation of the device but also presents a challenge for the electronic device
drivers.

The most basic device architecture of PLEDs has been described above; most

technical applications rely on much more sophisticated designs. For example, al-
most every display application will demand arrays of many individually address-
able pixels, similar to conventional LCDs (Fig. 7). Pixilated PLEDs rely on the
creation of a “grid” of intersecting anode and cathode rows, which is usually gen-
erated by photolithographic patterning of the ITO anode into columns and vapor
deposition of the cathode rows through a corresponding mask. As it is well known
for LCDs (33), the devices can be operated in either passive- or active-matrix
mode. Passive-matrix addressing relies on the consecutive addressing of the rows
(typically within a period of 1/30 s or less) and on applying a simultaneous bias to
the columns holding active pixels. Where an active row and column intersect, the
electric field directly causes EL emission. This format, which is usually used for
low resolution applications, requires high scanning rates in order to avoid flicker-
ing of the image. Moreover, because each pixel is active only for a short fraction of
the cycle, it will require to generate very high intensity (ie,

>10,000 cd/m

2

) under

a high current density during its “on” state in order to produce an acceptable aver-
age luminance. The active matrix architecture moderates the need to cycle at high
rates by giving every pixel a “memory” in the form of a capacitor and a transis-
tor. While this architecture seems mandatory for high resolution and large-area
displays, it adds, of course, another layer of complexity to the device.

Another step toward high end application is, of course, the creation of multi-

color devices, featuring red, green, and blue pixels. A variety of approaches have
been suggested to achieve the latter, including (1) the use of uv or blue LEDs
and external color conversion through green and red-light-emitting fluorescent
dyes applied to the appropriate pixels, (2) the use of white LEDs in combination
with conventional absorbing color filters, (3) the stacking of individual, indepen-
dent device segments in vertical fashion, and (4) the use of advanced deposition

98

LIGHT-EMITTING DIODES

Vol. 3

Fig. 7.

Polymer LED passive matrix display; 160

× 160 pixels with pitch of 300 µm.

Picture provided by Nobel Laureate Dr. Alan J. Heeger, courtesy of Uniax Corp., Santa
Barbara, Calif.

processes that allow the deposition of different EL polymers in patterned fashion,
for example, by ink-jet printing (4).

Another exciting development has been the design of PLEDs which emit lin-

early polarized light and, among other applications, might find use as backlights
of conventional LCDs (8,34). The central feature in these devices is a high degree
of uniaxial supramolecular orientation of the light-emitting polymer layer, which
is most easily achieved by using thermotropic liquid-crystalline PLED materials
such as poly(9,9-dialkyl fluorene)s or segmented poly(arylene vinylene)s. Mon-
odomain alignment in thin films of these materials is usually accomplished by
using an alignment layer such as a rubbed poly(imide) film, leading to highly po-
larized photo- and electroluminescence, with dichroic ratios (ratio of the intensity
of emitted light polarized parallel and perpendicular to the orientation direction,
respectively) that can exceed 20.

After only a decade of development, PLEDs indeed seem to have matured

to the edge of commercial exploitation. The technology is expected to initially
address the market for small- and medium-sized alphanumeric displays, as used
in calculators, handheld telephones, pagers, car radios, personal digital assistants
information readouts on household appliances, and in automobiles, which has a
current annual volume of about US$2.5 billion, with high growth rates expected.

Vol. 3

LIGHT-EMITTING DIODES

99

The display technologies traditionally employed in these applications are LCDs,
vacuum fluorescence displays, and inorganic LEDs. Industrial sources estimate
that by 2004–2005, organic LEDs (including PLEDs as well as LEDs based on
small organic molecules) may capture between US$350 to US$700 million of that
market (3,35). In the long run, PLEDs also have the potential to replace current
high resolution video-rate displays in, eg, desktop monitors, ultrathin television
sets, and the like.

BIBLIOGRAPHY

1. M. G. Craford and F. M. Steranka, in G. L. Trigg, ed., Encyclopedia of Applied Physics,

Vol. 8, VCH, Weinheim, 1994, pp. 485.

2. J. H. Burroughes and co-workers, Nature 347, 539 (1990).
3. A. H. Tullo, C&EN 20 (June 26, 2000).
4. U. Mitschke and P. B ¨auerle, J. Mater. Chem. 10, 1471 (2000).
5. D. Kalinowski, J. Phys. D: Appl. Phys. 32, R179 (1999).
6. A. Kraft, A. C. Grimsdale, and A. B. Holmes, Angew. Chem. Int. Ed. 37, 403

(1998).

7. D. Bradley, Curr. Opin. Solid State Mater. Sci. 1, 789 (1996).
8. A.

Greiner,

ed.,

Polym.

Adv.

Technol.

(special

topic

issue)

9,

369–490

(1998).

9. S. Miyata and H. S. Nalwa, eds., Organic Electroluminescent Materials and Devices,

Gordon and Breach Science Publishers, Amsterdam, 1997.

10. M. T. Bernius, M. Inbasekaran, J. O’Brien, and W. Wu, Adv. Mat. 12, 1737

(2000).

11. A. Greiner and co-workers, Macromol. Chem. Phys. 197, 113 (1996).
12. C. W. Tang and S. A. VanSlyke, Appl. Phys. Lett. 51, 913 (1987).
13. C. Adachi, T. Tsutsui, and S. Saito, Appl. Phys. Lett. 56, 799 (1990).
14. M. Pope, H. Kallmann, and P. Magnante, J. Chem. Phys. 38, 2042 (1963).
15. B. R. Hsieh and co-workers, J. Am. Chem. Soc. 129, 231 (1998).
16. M. A. Baldo and co-workers, Nature 395, 151 (1998).
17. N. C. Greenham, R. H. Friend, and D. D. C. Bradley, Adv. Mater. 6, 491 (1994).
18. G. Gu and co-workers, Opt. Lett. 22, 396 (1997).
19. C. Bastiaansen, T. Tervoort, and C. Weder, eds., Macromolecular Symposia, Vol. 154,

Wiley-VCH, Weinheim, 2000.

20. G. Gustafsson and co-workers, Nature 357, 6378 (1992).
21. H. Kubota and co-workers, J. Lumin. 56, 87 (2000).
22. MRS Bulletin 25(8) (2000).
23. M. Thelakkat, R. Fink, F. Haubner, and H.-W. Schmidt, Macromolecular Symposia,

Vol. 125, Wiley-VCH, Weinheim, 2000, pp. 157.

24. C. Schmitz and co-workers, Adv. Funct. Mater. 11, 41 (2001).
25. J. Moulton and P. Smith, in M. Lewin and J. Preston, eds., Handbook of Fiber Science

and Technology III, High-Technology Fibers, Part C, Marcel Dekker, Inc., New York,
1993, pp. 275.

26. A. Greiner and co-workers, Acta Polym. 48, 1 (1997).
27. D. Gin and V. Conticello, Trends Polym. Sci. 4, 217 (1996).
28. U.S. Pat. 3,401,152 (1986), R. A. Wessling and R. G. Zimmerman. (to The Dow Chemical

Company).

29. D. D. C. Bradley, J. Phys. D: Appl. Phys. 20, 1389 (1987).

100

LIGHT–EMITTING DIODES

Vol. 3

30. F. Wudl, P. M. Allemand, G. Srdanov, Z. Ni, and D. McBranch, Materials for Nonlinear

Optics: Chemical Perspectives, American Chemical Society, Washington, D.C., 1991,
pp. 683.

31. D. Braun and A. J. Heeger, Appl. Phys. Lett. 58, 1982 (1991).
32. URL:

www.cdtltd.co.uk;

www.covion.com;

www.dow.com;

www.dupont.com;

www.philips.com; www.uniax.com.

33. L. K. M. Chan, in R. J. Brook, M. C. Flemings, and S. Mahajan, eds., The Encyclopedia

of Advanced Materials, Vol. 2, Elsevier Science Ltd., Oxford, 1994, pp. 1294.

34. M. Grell and co-workers, Adv. Mater. 11, 895 (1999).
35. D. E. Mentley, Laser Focus World 35(5), 111 (1999).

A

NDREAS

G

REINER

Universit ¨at Marburg Germany
C

HRISTOPH

W

EDER

Case Western Reserve University