Recent progress of molecular organic electroluminescent
materials and devices
L.S. Hung
a,*, C.H. Chen
baCenter of Super-Diamond and Advance Films (COSDAF) and Department of Physics & Materials Science,
City University of Hong Kong, Hong Kong, China
b
Department of Applied Chemistry and Microelectronics & Information Systems Research Center, National Chiao Tung University, Hsinchu 300, Taiwan, ROC
Abstract
Electroluminescent devices based on organic materials are of considerable interest owing to their attractive characteristics and potential applications to flat panel displays. After a brief overview of the device construction and operating principles, a review is presented on recent progress in organic electroluminescent materials and devices. Small molecular materials are described with emphasis on their material issues pertaining to charge transport, color, and luminance efficiencies. The chemical nature of electrode/organic interfaces and its impact on device performance are then discussed. Particular attention is paid to recent advances in interface engineering that is of paramount importance to modify the chemical and electronic structure of the interface. The topics in this report also include recent development on the enhancement of electron transport capability in organic materials by doping and the increase in luminance efficiency by utilizing electrophosphorescent materials. Of particular interest for the subject of this review are device reliability and its relationship with material characteristics and interface structures. Important issues relating to display fabrication and the status of display development are briefly addressed as well. #2002 Elsevier Science B.V. All rights reserved.
Keywords: Electroluminescence; Organic light-emitting devices; Carrier injection; Efficiency; Stability; OLED-based displays
1. Introduction
Organic electroluminescence (EL) is the electrically driven emission of light from non-crystalline organic materials, which was first observed and extensively studied in the 1960s [1,2]. In 1987, a team in Kodak introduced a double layer organic light-emitting device (OLED), which combined modern thin film deposition techniques with suitable materials and structure to give moderately low bias voltages and attractive luminance efficiency [3,4]. Shortly afterwards, in 1990 the Cambridge group of Friend announced a conducting polymer-based LED [5,6]. Since then, there have been increasing interests and research activities in this new field, and enormous progress has been made in the improvements of color gamut, luminance efficiency and device reliability. The growing interest is largely motivated by the promise of the use of this technology in flat panel displays. As a consequence, various OLED displays have been demonstrated. In this article, the reader will find a description of the latest trends in organic EL research covering all the new
*
Corresponding author. Present address: Department of Physics and Materials Science, City University of Hong Kong, Tat Chee Avenue, Kowloon, Hong Kong, China. Tel.:þ852-2788-7433; fax: þ852-2788-7830.
E-mail address: [email protected]. (L.S. Hung).
0927-796X/02/$ – see front matter # 2002 Elsevier Science B.V. All rights reserved. PII: S 0 9 2 7 - 7 9 6 X ( 0 2 ) 0 0 0 9 3 - 1
achievements and most important data, but the description is confined to molecular organic EL materials and devices.
The design of EL materials for used in OLEDs is critical to device performance. Great strides have been made towards the development and improvement of molecular materials for display applications. Intense research in both academia and industry over the last 3–4 years has yielded OLEDs with remarkable color fidelity, device efficiencies and operational stability.
Charge injection and transport are the limiting factors in determining operating voltage and luminance efficiency. In OLEDs, the hole current is limited by injection, and the electron current is strongly influenced by the presence of traps owing to metal–organic interactions. In order to enhance carrier injection the selection of efficiently electron-injecting cathode materials and the use of appropriate surface treatments of anodes are of great importance. The current statues of charge injection and transport will be discussed in this review.
The efficiency of an OLED is determined by charge balance, radiative decay of excitons, and light extraction. Significant progress has been made recently in developing phosphorescent emitters via triplet–triplet energy transfer, and high-efficiency OLEDs in various colors have been demonstrated. Light extraction is determined by the device structure and the refractive indices of the composed layers. Current research activity is directed toward various surface modifications that can increase extraction efficiency, while the problem of light-trapping remains unsolved in terms of its application to displays.
The most critical performance characteristic for OLEDs is the device operational lifetime. Continuous operation of OLEDs generally leads to a steady loss of efficiency and a gradual rise in bias voltages. Although OLEDs have achieved long operational stability, the material issues underlying the EL degradation are not fully understood. There are many studies of degradation process, and an overview will be presented in this article.
The remarkable advance in OLEDs has led to its application to flat panel displays. The promise of low-power consumption and excellent emissive quality with a wide viewing angle is unique among display technologies. At the moment, passive monochrome and multicolor displays are commercially available, and active-matrix full-color displays have been demonstrated. Some important issues relating to display fabrication and the status of display development will also be described.
2. Background
2.1. Device configuration and operation
An OLED has an organic EL medium consisting of extremely thin layers (<0.2 mm in combined thickness) sandwiched by two electrodes. In a basic two-layer OLED structure, one organic layer is specifically chosen to transport holes and the other organic layer is specifically chosen to transport electrons. The interface between the two layers provides an efficient site for the recombination of the injected hole–electron pair and resultant electroluminescence.
When an electrical potential difference is applied between the anode and the cathode such that the anode is at a more positive electrical potential with respect to the cathode, injection of holes occurs from the anode into the hole-transport layer (HTL), while electrons are injected from the cathode into the electron-transport layer (ETL). The injected holes and electrons each migrate toward the oppositely charged electrode, and the recombination of electrons and holes occurs near the junction in the luminescent ETL. Upon recombination, energy is released as light, which is emitted from the light-transmissive anode and substrate.
The heterojunction should be designed to facilitate hole-injection from the HTL into the ETL and to block electron injection in the opposite direction in order to enhance the probability of exciton formation and recombination near the interface region. As shown in Fig. 1, the highest occupied molecular orbital (HOMO) of the HTL is slightly above that of the ETL, so that holes can readily enter into the ETL, while the lowest unoccupied molecular orbital (LUMO) of the ETL is significantly below that of the HTL, so that electrons are confined in the ETL. The low hole mobility in the ETL causes a build up in hole density, and thus enhance the collision capture process. Furthermore, by spacing this interface at a sufficient distance from the contact, the probability of quenching near the metallic surface is greatly reduced.
The simple structure can be modified to a three-layer structure, in which an additional luminescent layer is introduced between the HTL and ETL to function primarily as the site for hole– electron recombination and thus electroluminescence. In this respect, the functions of the individual organic layers are distinct and can therefore be optimized independently. Thus, the luminescent or recombination layer can be chosen to have a desirable EL color as well as a high luminance efficiency. Likewise, the ETL and HTL can be optimized primarily for the carrier-transport property. The extremely thin organic EL medium offers reduced resistance, permitting higher current densities for a given level of electrical bias voltage. Since light emission is directly related to current density through the organic EL medium, the thin layers coupled with increased charge injection and transport efficiencies have allowed acceptable light emission to be achieved at low voltages. 2.2. Materials
The advantages of organic materials over inorganic materials are their excellent color gamut and high fluorescence efficiency. Light is produced in organic materials by the fast decay of excited molecular states, and the color of light depends on the energy difference between those excited states and the molecular ground level. Many materials show intense photoluminescence with near unity quantum yield, while the EL efficiency is limited by the probability of creating non-radiative triplet exited states in the electron–hole recombination.
Most organic materials support preferentially the transport of either electrons or holes with their mobilities in the range of 108to 102cm2/(V s). Electron mobility in organic materials is generally orders of magnitude lower than hole mobility. The most important electron-transport material is tris(8-hydroxyquinolinato)aluminum (Alq3) with its molecular structure shown in Fig. 2, and the
electron mobility in Alq3strongly depends on electric field with a value of approximately 106cm2/
(V s) at 4 105V/cm. Alq
3 is also used as an emissive material, which emits in the green with a
broad emission peaking at 530 nm. Other EL colors can be obtained by doping a small amount of specific guest molecules in Alq3or by choosing different organic fluorescent materials as emitters. In
some cases, doping also enhances luminance efficiency by reducing non-radiative decay.
There are several materials that have been preferred as hole-transport materials (HTMs). Among them, N,N0-diphenyl-N,N0-bis(3-methylphenyl)(1,10-biphenyl)-4,40-diamine (TPD) and N,N0 -bis(l-naphthyl)-N,N0-diphenyl-1,10-biphenyl-4,40-diamine (NPB) have been studied extensively. The hole-transport materials in Fig. 2 have a glass transition temperature below 100 8C and a hole mobility in the range of 103to 104cm2/(V s). In OLEDs, the hole current dominates the total current, owing to efficient hole-injection and sufficiently high hole mobilities.
A barrier for electron injection is commonly present at the metal–organic contact when the work function of the metal is larger than the LUMO of the organic materials, and thus the use of a low-work function metal is highly desirable to facilitate the injection of electrons. Mg alloyed with a small amount of silver is a commonly used cathode. Mg is a relatively stable metal with a work function of 3.66 eV, which is sufficiently low for it to be useful as an electron-injecting electrode. The small amount of Ag assists the Mg deposition by presumably providing nucleating sites on the alloyed film during co-sublimation.
The hole-injecting contact requires a metal of high work function to match with the HOMO of the organic material. All OLEDs rely on the transparent and conductive indium tin oxide (ITO) as the anodic material to facilitate hole-injection while permitting light to exit the device in an effective manner. The work function of ITO ranges from 4.5 to 5.0 eV, strongly depending on the methods of surface treatment. Treatments of ITO glass substrates using UV ozone or oxygen plasma substantially increase its work function and consequently enhance hole-injection from the ITO anode into the HTL. 2.3. Device preparation
Vacuum evaporation by resistive heating is most appropriate for depositing molecular materials. Organic vapor phase deposition (OVPD) has also been demonstrated to deposit organic materials on large substrates [7]. It is also common to deposit cathode materials using the same vacuum
evaporation from filaments. For the deposition of high-temperature metals one may employ e-beam evaporation or sputtering. The latter is particularly useful for large substrates and high throughput production. However, OLEDs are extremely sensitive to radiation, and special care needs to be taken. In e-beam deposition, a magnetic field is applied across the substrate to repel electrons and ions. In sputter deposition, a buffer layer is required to minimize the radiation damage inflicted on the OLED organic layer stack.
Typical device fabrication occurs by the following sequence: devices are grown on glass slides pre-coated with transparent ITO with a sheet resistance of 15–100 O/&. Substrates are ultrasonically cleaned in detergent solution, followed by thorough rinsing in deionized water. They are then cleaned in organic solvents and dried in pure nitrogen gas. After cleaning, the ITO glass is subject to an oxygen treatment either using UV ozone or oxygen plasma to enhance hole-injection. Single heterostructure devices (Fig. 2) are formed by sequential high vacuum (105 to 106Torr) vapor deposition of a hole-transport layer such as NPB, followed by an electron-transport layer of Alq3,
previously purified by temperature gradient sublimation. Deposition is carried out by thermal evaporation from a baffled Ta crucible at a nominal deposition rate of 0.2–0.4 nm/s. An electron-injecting electrode of approximately 10:1 Mg:Ag volume ratio is subsequently deposited by co-evaporation from separate Ta boats at a vacuum of 105Torr. The device preparation is completed with encapsulation in a dry argon box.
OLEDs are constructed using glassy and amorphous organic films and thus provide significant advantages in device fabrication and cost reduction. They are pronouncedly different in structures from inorganic LEDs consisting of epitaxial semiconductor thin films.
2.4. Analytical tools
2.4.1. Electrical and optical properties
Absorption, excitation and luminescence spectra are determined with a spectrophotometer. Current–voltage and luminance–current characteristics are measured by using a radiometer and a digital voltmeter. Typical current–voltage and luminance–current characteristics of a device with 75 nm of NPB, and 70 nm of Alq3 are shown in Fig. 3. The voltage required for light emission is
strongly dependent on the thickness of the Alq3.
In OLEDs we distinguish the quantum efficiency (Z) and the luminous efficiency (Zp). The
quantum efficiency Z is defined as the ratio of the number of emitted quanta to the number of charge carriers. Quantum efficiency is an important quantity, which reflects the comprehensive result of the EL process, but the luminous efficiency has a more technical significance, which is the ratio of the luminous flux emitted by the device and the consumed electric power.
2.4.2. Metal–organic interfaces
Ultraviolet photoelectron spectroscopy (UPS) and X-ray photoelectron spectroscopy (XPS) are commonly used to determine the nature of the metal–organic interfaces [8]. Samples are illuminated with light, and the kinetic energy distribution of photogenerated electrons is analyzed. With a relatively low photon energy of 21.2 eV, UPS is employed to measure the ionization potential of organic materials and the work function of metals. With relative high photon energy in the range of 1 keV, XPS is employed to determine the elements present in the near surface region and provides subtle information on chemical bonding.
2.4.3. Carrier mobilities
Both time-of-flight (TOF) and transient EL are utilized to measure carrier mobilities [9,10]. The TOF method determines the flight time, which is needed for single sign charges generated near one surface of a sample to move across the sample to the other side. A short pulse of strongly absorbed light generates the free charge carriers. The sample is sandwiched between two electrodes (one semitransparent) to allow application of a constant electric field. In transient EL analysis, EL from a testing OLED is investigated with the materials of interest as a carrier-transport layer. Using a voltage pulse as an excitation source, the delay time for the EL is measured and interpreted as the carrier transit time across the carrier-transport layer.
3. Molecular organic electroluminescent materials
The phenomenon of organic electroluminescence was first discovered by Pope in 1963 [11]. But, the development of organic light-emitting device (or diode) known as today’s OLED technology actually began in the Chemistry Division of Kodak Research Laboratories in the late 1970s by Tang and coworkers. Their research led eventually to the discovery of the first efficient multi-layered organic electroluminescent device based on the concept of the heterojunction architecture [12,13] which was followed by the disclosure of the doped emitter using the highly fluorescent organic dyes for color tuning and efficiency enhancement [14,15]. Since then, tremendous progress has been made in the field of organic electroluminescence, particularly in recent years and commercialization activity is well underway. One of the key enablers in the history of OLED advancement can be attributed to the continuing discovery of new and improved electroluminescent materials which were made possible by the dedication and ingenuity of many organic chemists who provide the design and skilled synthesis. Indeed, from small molecules, oligomers to conjugated polymers, intense research in both academia and industry has yielded OLEDs with remarkable color fidelity, device efficiencies and operational stability.
This part of the review is written primarily from an organic chemist’s perspective highlighting some of the most significant development of small molecular materials published in the last three years (through 2001). These will include the coverage of hole-injection, hole transport, electron transport and host emitting materials as well as RGB and miscellaneous fluorescent dopants. At the end of this section, a special section is devoted to the progress made on the triplet emitter harvesting
the organic phosphorescent emission. For complementary readings, the readers are referred to two recent review articles published in January and June of 2000 by Shirota [16] on Organic materials for electronic and optoelectronic devices and by Mitschke and Bauerle [17] on The electroluminescence of organic materials, respectively. For the sake of continuity, several representative reviews prior to 1999 are cited here for further reference [18–24]. For readers who are more interested in polymeric electroluminescent materials whose design features in certain respects are similar to those of small molecules, a fair number of excellent reviews are also included in the reference section [25–28].
3.1. Hole-injection materials
Oxidation of the ITO surface by O2plasma, CF4/O2plasma [29] or UV ozone treatment can reduce
the carrier injection energy barrier, remove residual organic contaminants and get its work function up to near 5 eV which is still about 0.5 eV lower than most of the HOMO of the hole-transport materials. A layer of hole-injection material (sometimes referred to as ITO or anode buffer layer) which reduces the energy barrier in-between ITO/HTL is therefore beneficial to enhancing charge injection at the interfaces and ultimately improving power efficiency of the device. Thus, hole-injection can be promoted by introduction of new hole-transport layers with optimized HOMO levels and by inserting a thin layer of copper phthalocyanine (CuPc) [30], starburst polyamines [31], polyaniline [32,33] and SiO2[34] between the ITO/HTL interface. In addition, HTLs doped with oxidizing agents such as FeCl3
[35], iodine [36], tetra(fluoro)-tetra(cyano)quinodimethane (TF-TCNQ) [37] and tris(4-bromopheny-l)aminium hexachloroantimonate (TBAHA) [38] have been reported as effective materials for hole-injection (see examples in Fig. 4). The latter can also be dispersed in a polymeric matrix to improve surface roughness as well as thermal stability [39]. One of the widely used polymers for promoting hole-injection is poly(3,4-ethylenedioxythiophene)–poly(styrene) known as PEDOT/PSS which has been found to be useful in a hybrid OLED architecture combining both the advantages of polymer LED (PLED) and multi-layered small molecule OLED [40]. PEDT/PSS is an aqueous gel dispersion. The hydrated gel particles are formed by PSS interlinked by PEDT chains. Like CuPc, PEDOT/PSS as hole-injection layer can smooth the ITO surface, reducing the probability of electrical shorts, decreasing the turn-on voltage and prolonging the operation lifetime of the device [41]. One of the potential drawbacks of using PEDOT/PSS is its acidity which could get as high pH as3. However, by modifying particle size distribution and increasing the materials resistivity, a new formulation promoted by Bayer AG is reported to be able to reduce crosstalk in passive-matrix OLEDs [42].
Nuesch et al. found that surface treatment using grafting of molecules [43] and adsorption of acids or bases [44] on ITO could also modify its work function. In the case of 2-chloroethylpho-sphonic acid, a self-assembled monolayer (SAM) can be readily formed on the ITO substrate which significantly reduced the threshold voltage of a standard OLED device of [ITO/TPD/Alq3/Al] [45].
Later, it was found by using polar adsorbate molecules with the dipole oriented outward from the surface, an artificial dipolar layer was formed and the work function is increased. With this method the threshold voltage for light emission could be reduced by 4 V and the maximum luminance increased by a factor of 3.5, giving an overall performance superior to that using the stable Mg:Ag counter electrode [46].
Recently, p-doped aromatic diamines have been found to be excellent injection materials such as SbCl5-doped N,N0-bis(m-tolyl)-1,10-biphenyl-4,40-diamine (TPD) thin film [47] as well as the
amorphous starburst amine, 4,40,400-tris(N,N-diphenylamino)triphenylamine (TDATA) doped with a very strong acceptor TF-TCNQ by controlled co-evaporation [48]. Multi-layered OLEDs consisting of [ITO/TF-TCNQ (2%):TDATA (100 nm)/TPD (10 nm)/Alq3 (65 nm)/LiF (1 nm)/Al] achieved a
very low operating voltage of 3.4 V for obtaining 100 cd/m2 at 9.1 mA/cm2. Another notable example of an anode buffer layer is a-sexithiophene (a-6T) which at an optimal thickness of 60 nm has been shown to improve EL efficiency and lower operating voltage [49]. Since the oxidation potential of oligothiophenes can be readily tuned by lengthening p-conjugation and by regiospecific substitution, e.g. DetDOc6T [50], the doped oligothiophenes can also be expected to be useful as injection materials for electroluminescence.
3.2. Hole-transport materials
Ever since the discovery of using tri-arylamines with a ‘‘bi-phenyl’’ center core as the hole-transport layer which greatly improved both EL efficiency and operational stability of OLED [51], most of the new HTM developed seemed to have all evolved around this theme. The creativity of synthetic chemists as well as material scientists throughout the world continue to provide the OLED device community with their ever improved products having superb properties and elegant design. By far, one of the most widely used HTM in OLED is still NPB. One of the reasons for its popularity is because sublimed NPB can be manufactured readily and is thus abundantly available even though its Tg at 98 8C is a trifle low which may affect its morphological stability at high operating
temperature. Therefore, studies on the design and synthesis of new HTMs have been continually focused on finding materials with high thermal and thin film morphological stabilities and on finding
ways to control and optimize carrier injection and transport. These approaches to molecular design can be roughly categorized into biphenyl diamine derivatives; starburst amorphous molecular glass; spiro-linked biphenyl diamines and miscellaneous examples as follows:
1. Biphenyl diamine derivatives: Heat treatment of organic multi-layers has been found to cause an interdiffusion between organic layers in OLED [52] which ultimately effects the stability of the device. Therefore, to increase the Tg of HTM is critical in obtaining a more thermally durable
display. Using thermodynamical consideration, Sato has proposed a molecular design rule according to which high Tgmaterials can be obtained by increasing the number of p-electrons and
by decreasing rotational moment by placing a heavy moiety at the center of the molecule [31]. Fig. 5 shows the various structures of HTMs with a biphenyl diamine related core along with its reported Tg (8C) and Ip (eV) values for comparison [53,54]. Another variation of NPB is to
substitute a-naphthyl with distyryl groups as in p-dmDPS [55] which has a Tg of 108 8C and
showed an enhanced luminance and improved I–V characteristics in the presence of CuPc as hole-injection layer in a device consisting of [ITO/CuPc (10 nm)/HTM (50 nm)/Alq3(50 nm)/LiF/Al].
2. Starburst amorphous molecules: The formation of pinhole free thin film morphology can be assured if the HTMs can form stable amorphous glasses with high Tgs. The guidelines for the
design of amorphous molecular materials as provided by Shirota et al. [56] are: (1) to increase the number of conformers together with non-planar molecular structure; (2) to introduce bulky and heavy substituents as to enlarge molecular size for attaining and maintaining the stability of the glassy state; (3) to increase Tg by the incorporation of a rigid moiety or an intermolecular
hydrogen bonding site into non-planar molecules and by increasing the molecular weight. Based on these guidelines, Fig. 6 shows the various general molecular structures that have been synthesized in the laboratory of Shirota. This can be exemplified by the glass of p-MTDAB which
tends to crystallize in several months on standing whereas the amorphous glasses of p-DPA-TDAB and MTBDAB are very stable without any crystallization even on heating above their Tgs.
A similar idea has been expanded by using a p-rich thiophene system coupled onto the benzene nucleus [57,58]. In Tao’s approach, e-rich thienyl ring [59] next to the nitrogen atom can effectively lower the oxidation potential of HTM which could be useful for mediating hole-injection in a double HTL device [60]. In hexakis[(diarylamino)thienyl]benzene (1) shown in Fig. 2, it is worth noting that it was synthesized in one scoop by a palladium-catalyzed six-fold arylation of hexabromobenzene using Stille’s cross-coupling reaction in good yield.
3. Spiro-linked molecules: The concept of enhancement of thermal stability of the amorphous state of HTMs and augmentation of Tgs has been introduced by Salbeck et al. [61] via a spiro center
with a 90 8C molecular architecture. It has been shown by TOF technique that spiro-linked HTM such as spiro-mTTB shown in Fig. 7 increased hole mobility [62] and improved EL performance [63] as compared to the parent non-spiro analogs. The advantage of Tgs and other thermal
properties and solid PL are compared in Table 1.
An unsymmetrical spiro-compound with a high Tg of 122 8C (shown as spiro2) [64] can be
synthesized by coupling 2,7-bis(diphenylamino)-9-fluorene with 2-lithio-4,40-di(t-butyl)biphenyl, followed by acid-catalyzed spirocylization of the resulting alcohol [65]. In a device structure of [ITO (160 nm)/HTM (60 nm)/Alq3 (60 nm)/LiF (0.5 nm)/Al (150 nm)], spiro2 reached a
luminance efficiency of 6.1 cd/A and 3.6 lm/W at a drive voltage around 6 V and a luminance of 300 cd/m2. The authors attributed the enhancement of its EL efficiency to the extreme non-planar conformation and steric hindrance which are effective in suppressing exciplex formation in the two-layer device.
4. Miscellaneous examples (Fig. 8): HTM can also serve as the emissive material in a two-layer OLED device if the exciton can be generated in the HTL. The fluorescent pyrene-substituted
carbazole derivative 2 (Tg184 8C), which can be synthesized in high yield by palladium coupling
reaction between N-(pyrenyl)aniline and 3,6-dibromocarbazole using tris(dibenzylideneaceto-ne)dipalladium (Pd(dba)2) catalyst [66] and a sterically hindered ligand P(t-Bu)3in the presence
of sodium t-butoxide in reluxing toluene displayed a green emission at 530 nm. Later it was found 2-(di-tert-butylphosphino)-o-biphenyl was a better ligand than P(t-Bu)3 to promote the
palladium-catalyzed amination of aryl halides and sulfonates for the synthesis of unsymmetrical triarylamines [67]. Their device had a two-layer structure of [ITO/HTM (50 nm)/TPBI (50 nm)/ Mg:Ag] in which TPBI [1,3,5-tris(N-phenylbenzimidazol-2-yl)benzene] served as electron-transport layer [68] as it is a more effective hole blocker than Alq3 to confine the charge
recombination in the HTL [69]. Efficient blue-green emission with lmax¼ 490 nm from hole
transporting dibenzochrysene derivatives (DBC) [70] has been observed in a device of [ITO/DBC (60 nm)/Alq3 (60 nm)/LiF (0.5 nm)/Al (150 nm)]. Closer examination of their reported EL
spectra in which there was a shoulder near 530 nm tailing into 600 nm revealed that the green portion of the emission could conceivably originate from charge recombination in the e-transport Alq3 layer. This is another piece of evidence of the bipolar nature of Alq3 which transports
Fig. 7. Example of spiro-linked hole-transport materials.
Table 1
Comparison of thermal and solid PL properties of spiro-HTMs with NPB
HTM NPB Spiro-NPB Spiro-TAD Molecular weight 588.8 1185.5 985.3 Tg(8C) 98 147 133 Tm(8C) 290 294 276 Tsublim(8C at 4 105mbar) 310 430 355 lmax(PL film) 440 450 405
electrons but does not appear to block holes as effectively as TPBI. In addition, Popovic and coworkers at Xerox [71] showed rather convincingly that holes injected into Alq3 which are
trapped to form radical cations [Alq3]þcould be one of the major causes for device degradation.
In their experiments [72], a dual-layer device with rubrene doping in both hole-transport NPB and electron-transport Alq3showed extrapolated life of over 50,000 h at the current density of 25 mA/
cm2 and an initial luminance of 100 cd/m2, that is 100 better than the device without doping. The same authors have recently confirmed that the increase in OLED stability by means of doping the HTL is found to be associated with changes in bulk HTL properties rather than interfacial properties, in agreement with the OLED degradation mechanism based on the instability of the Alq3 cationic species [73]. Further studies using time-resolved fluorescence
measurements [74] show that the decrease in the PL quantum efficiency is associated with a decrease in the lifetime of the Alq3 excited states, thus revealing the nature of the degradation
products as luminescence quenchers.
High TgHTMs based on 5,11-(diaryl)indolo[3,2-b]carbazoles (3) have been reported by Hu et al.
[75] with one of the best derivatives 2 having a Tgof 164 8C as shown in Fig. 3. Finally, it is worth
noting that TPD, a morphologically unstable HTM with a low Tg can be dispersed up in
concentrations to 75 wt.% into the high Tgtransparent polymers to form solid solutions which can be
spin-coated as thin films (80 nm) and used as thermally and morphologically stable HTL [76]. 3.3. Electron-transport and host emitting materials
To date, the most widely used electron-transport and host emitting material in OLEDs is still Alq3. This is because Alq3is thermally and morphologically stable to be evaporated into thin films,
easily synthesized and purified, molecularly shaped to avoid exciplex formation (e.g. with NPB at the interface), and green fluorescent to be a good host emitter. Arguably, it is still one of the most robust electron-transport backing layers in OLED, particularly with the help of the hole blocker to trap the hole carriers from injecting into Alq3[77] or doping with lithium or other alkali metals [78]
to assist electron injection to lower the drive voltage [79]. But, it has also many shortcomings such as quantum efficiency, mobility, bandgap and the ashing problem during sublimation. Studies toward the understanding of its molecular packing [80], photodegradation [81], electron drift mobility [82], transport phenomenon [83], excited state [84], fluorescence by ZINDO calculation [85] and electric field-dependent quenching of EL [86] are continuing. It has been found by the time-of-flight technique that the drift mobility of electrons in Alq3is increased by about two orders of magnitude
(to 104cm2/(V s)) as the deposition rate decreased from 0.7 to 0.2 nm/s. The electron drift mobility in Alq3is found to increase linearly with the square root of the applied electric field.
There were also attempts to improve the quantum efficiency, thermal stability and thin film morphology of Alq3 by structural modifications as shown in Fig. 9.
Tris(5-hydroxymethyl-8-quinolinolato)aluminum (AlOq) [87] was reported to form a more uniform amorphous thin film than Alq3 by slow vacuum deposition and the device [ITO/PVK/AlOq/Al] also emitted green light with
better efficiency. This improvement was attributed to the hydroxymethyl groups attached to the C-5 positions of quinoline which, due to intermolecular hydrogen bondings, could lead to the formation of super molecular structures in AlOq thin film. An unsymmetrically substituted bis(5,7-dichloro-8-quinolinolato)-(8-quinolinolato)aluminum (Alq(Clq)2) [88] was prepared by first reacting
8-hydroxyquinoline and triethylaluminum in refluxing benzene to form Alq(C2H5)2as a yellow solid
which could be further reacted with 2 eq. of 5,7-dichloro-8-quinoline to precipitate the product in 61% yield. The emission of Alq(Clq)2 is slightly bathochromicly shifted from that of Alq3.
Comparison of the EL lmaxfor Alq3, Alq(Clq)2and Al(Clq)3indicates that the extent of the red shift
increases nearly linearly as the number of Clq ligands increases. Another tridentate unsymmetrical (salicylidene-o-aminophenolato)(8-quinolinolato)aluminum (Al(Saph-q)) [89] could be prepared by simply reacting AlCl36H2O and 8-hydroxyquinoline (1 eq.) in ethanol in the presence of piperidine
followed by another equivalent of salicylidene-o-aminophenol. The glassy solid (Al(Saph-q)) was reported to have a higher Tg (226 8C) and to be thermally more stable than Alq3. It was however
difficult to understand why the intermediate of dichloro(8-quinolinolato)aluminum could simply be formed by mixing of AlCl36H2O and 8-hydroxyquinoline (1 eq.) without the contamination of Alq3
as usually unsymmetrical Alq2q could only be prepared using a hindered q-ligand as in
2-methylquinoline [90].
Other metal chelates which showed decent device performances and interesting fluorescent properties are shown in Fig. 10. Sano et al. at Sanyo [91] have prepared several kinds of 2:1 complexes with 8-hydroxyquinoline derivatives and a variety of metal ions, such as Be, Mg, Ca, Sr, Sc, Y, Cu or Zn. Amongst them, the beryllium complex (Beq2) was found to be the most fluorescent (520 nm) in the
green and the zinc complex (Znq2) was found to have a strong yellow fluorescence (556 nm). They
have also synthesized bis(10-hydroxybenzo[h]qinolinato)beryllium (Bebq2) which has a very strong
green fluorescence at lmax 515 nm, a high melting point (368oC), electron-transport property and
achieved a luminance efficiency of 6.1 cd/A in a device of [ITO/TPD (50 nm)/Bebq2(50 nm)/Mg:In
Fig. 9. Modified Alq3molecules.
(200 nm)]. The highly fluorescent bis[2-(2-hydroxyphenyl)pyridine]beryllium (Bepp2) [92] as
electron-transport emitter in [ITO/NPB (60 nm)/Bepp2 (50 nm)/LiF (1 nm)/Al (200 nm)] was found
to emit in the blue (445 nm) with a turn-on voltage of 3 V and a maximum luminance efficiency reaching 3.8 cd/A at 33 cd/m2and a maximum luminance of 15,000 cd/m2at drive voltage of 12 V. In their device, however, it was difficult to discern the origin of the blue emission which could come from both NPB and Bepp2layers. The diphenylboron analogs of Alq3(Ph2Bq2) have also been synthesized
and their solution PLs were about 15 nm shorter and their quantum efficiencies were higher than those of the corresponding Alq3 derivatives [93]. By combinatorial approach,
(8-hydroxyquinolinolato)-lithium (Liq) and (2-methyl-8-hydroxyquinolinolato)(8-hydroxyquinolinolato)-lithium (LiMeq) have been found to be useful emitter and electron injection/transport materials [94].
Recently, the 2-(2-hydroxyphenyl)-5-phenyloxadiazole (ODZ) and 1-phenyl-2-(2-hydroxyphe-nyl)benzimidazole (BIZ) have been found to be useful bidentate ligands for aluminum and zinc chelations [95]. The most interesting chelates are Zn(ODZ)2, Zn(BIZ)2 and Al(ODZ)3 as they all
could serve as host electron-transport emitters in the blue.
In addition to metal chelates, there were a number of innovative approaches to the design of n-type organic semiconductors for use in OLEDs as electron-transport materials. Not all of them have been proven to work in devices. Fig. 11 represents a structural glossary of what have been published lately.
One of the most widely used electron-transport and hole blocking materials is 2-biphenyl-4-yl-5-(4-t-butylphenyl)-1,3,4-oxadiazole (PBD) which has been branched, spiro-linked [96] and
starbursted to prevent from crystallization in thin films. In oligothiophene, it was shown that by substituting with electron-withdrawing and bulky dimesitylboryl groups as in BMB-3T [97,98] could produce a superior electron-transport emitter. This concept was further demonstrated in thiophene/ thienyl S,S-dioxide oligomer which has a high solid-state quantum efficiency (37%) and was used as electron-transport emitter in composite OLED [99] with substantial improvement of both the turn-on and operating voltages in comparison to conventional polythiophene-based OLEDs. Similar to electron-deficient oligothiophenes, perfluorinated oligo(p-phenylene)s have been shown to be efficient n-type semiconductors for OLED [100]. The perfluoro-2-naphthyl-substituted PF-6P which could be synthesized by organocopper cross-coupling reactions was reported to be a better electron-transport material than Alq3[101]. In a device of [ITO/TPTE (60 nm)/Alq3(40 nm)/PF-6P (20 nm)/
LiF (0.5 nm)/Al (160 nm)] the luminance and current density were higher than those of Alq3above
4.5 V and reached 19,970 cd/m2 at 10 V.
Tao et al. [102] revived 2,20,200-(benzene-1,3,5-triyl)-tris(1-phenyl-1H-benzimidazole) (TPBI) as an excellent electron-transport and hole blocker material in device [ITO/NPB/TPBI/Mg:Ag]. With an intermediate layer of CBP inserted between the NPB and TPBI layers, hole transport into the TPBI layer was assured [103]. The synthesis and characterization of tetraaryl-, tetraarylethynyl- and octaaryl-cyclooctatetraenes (COTs) as well as their use as electron-transport materials in OLEDs were reported by Thompson and coworkers [104]. These COTs are capable of forming highly stable glasses (Tg> 177 8C) and have a sufficiently wide energy bandgap (>3 eV) for fabricating
blue-emitting OLEDs. Another class of electron-transport materials for OLEDs is 2,5-diarylsiloles [105]. It was reported that the bipyridylsilole derivative, PyPySPyPy, showed higher electron-transporting abilities than that of Alq3 [106]. The efficiency of a device consisting of [ITO/TPD/Alq3/
PyPySPyPy/Mg:Ag] reached 2.2 lm/W at 3.4 V with a half-life 3 longer than that of the controlled device without PyPySPyPy under constant current drive.
3.4. Fluorescent dopants
One of the key developments in the advancement of OLED display technology can be attributed to the discovery of the guest–host doped emitter system [107]. This is because a single host material with optimized transport and luminescent properties may be used together with a variety of highly fluorescent guest dopants leading to EL of desirable hues with very high efficiencies. Another advantage of the doped emitter system in OLED is the enhancement of its operational stability by transferring the electrogenerated exciton to the highly emissive and stable dopant site thus minimizing its possibility for non-radiative decay [108]. This doping principle has recently been successfully extended to the exploitation of highly phosphorescent materials leading to nearly 100% internal EL efficiency [109]. We will discuss highlight of fluorescent RGB dopant developments in the following sub-sections and devote the last section to the review of the phosphorescent triplet emitter.
3.4.1. Green
The green fluorescent dopant was amongst the first to be successfully demonstrated in a commercial product and it is also by far the most efficient. One of the best green dopants is 10-(2- benzothiazolyl)-1,1,7,7-tetramethyl-2,3,6,7-tetrahydro-1H,5H,11H-[l]benzo-pyrano[6,7,8-ij]quinoli-zin-11-one, known as C-545T [110–112] which belongs to the highly fluorescent class of coumarin laser dyes. By virtue of its structural coplanarity, the julolidine donor situated at carbon #7 aligns its p-orbital of nitrogen to overlap with the p-orbitals of the phenyl ring for more effective conjugation which results in increasing its relative PL quantum yield (Z) to 90%. It is believed that the
enhancement of Z is obtained by diminishing non-radiative deactivation of the excited state by reducing the internal mobility and, since the single C–N bond often is the weakest linkage in such a molecule, the enhanced conjugation of the julolidine system is expected to also improve dye stability [113]. The four strategically positioned methyl groups in C-545T are important as steric spacers to minimize the dye–dye interaction at high concentration that leads to quenching of the fluorescence. Later, the Kodak group discovered that by substituting t-butyl groups at the benzothiazolyl ring as in C-545TB [114], the concentration quenching problem could be further suppressed and the thermal property was also greatly improved (Tgenhanced to 142 8C from 100 8C) without compromising its
emissive color. In addition, in the device of [ITO/CuPc (15 nm)/NPB (75 nm)/Alq3þ 1% dopant
(35 nm)/Alq3 (35 nm)/Mg:Ag (200 nm)], the luminance efficiency could be significantly increased
from 10.5 cd/A (C-545T) to 12.9 (C-545TB) at a drive current density of 20 mA/cm2 and with a 1931 CIE color coordinates of x¼ 0:30; y ¼ 0:65.
One of the more interesting results in subsequent research was found in C-545MT [115,116] where an extra methyl group was substituted at the C-4 position of C-545T. When doped in Alq3as
green emitter in OLED, C-545MT has the unusual property of resistance to concentration quenching and the sustaining of its EL luminance efficiency (7.8 cd/A) over a wide range of doping concentration from 2 to 12%, which is more than 10 times that of C-545T (Fig. 12). Single crystal XRD evidence (Fig. 13) suggested that the difference in doping behaviors could be linked to the
Fig. 12. Comparison of luminance efficiency vs. doping concentration of C-545T and C-545MT.
difference in molecular packing as well as volume of unit cell of the perspective dopant crystals. The twisted molecular geometry of C-545MT which is distorted by the steric effect of C(4)-methyl substituent could prevent the undesirable molecular aggregation from occurring at high concentration thus delaying the onset of quenching. Additional advantage of this particular series of coumarin dopants can be found from C-545TB in Fig. 14 with its nearly flat response of luminance efficiency (cd/A) with respect to a wide range of drive current conditions at an optimal doping concentration of about 1%. This is particularly desirable for the passive-matrix displays where the system would need to be capable of very high luminance at low voltage and have a ‘‘flat’’ cd/A response with respect to drive voltage.
In studying the effect of the location and width of doping region on efficiency using 3-(2-benzothiazolyl)-7-(diethylamino)-2H-1-benzopyran-2-one (C-6), it was found that [117] the recombination zone was 10 nm thick near the doped emitter/hole-transport layer interface and dopant molecules behaved as carrier traps and hopping sites in the Alq3hosted emitter layer. Other
notable green dopants exemplified in Fig. 15 are N,N0-diethylquinacridone (DEQ) [118] which was shown to be thermally durable in the doped device with temperature-independent quantum efficiency; imidazolidinone 4 derived from the green fluorescent protein of Aequorea [119]; 1-aryl-2-(5-dithoacetal-S,S-dioxide ketenyl-2-thienyl)-5-(2-thienyl)pyrrole (5) [120] and organolanthanide phosphor, (t-bu-PMP)3Tb(Ph3P=O) [121].
3.4.2. Red
Among the RGB dopants used in OLED, red emission, due to its low efficiency, remains to be the weakest link in realizing the full potential of an active-matrix [122] and passive dot matrix [123] full-colored OLED display. Research in search of a good red has established a target specification of luminance efficiency of over 4 cd/A, a saturated color with CIE coordinates approaching (x¼ 0:65; y¼ 0:35) and a device lifetime of over 10,000 h at an initial luminance of 300 cd/m2under constant
current drive. To date, few dopants in the red can meet all of these goals. One of the best that comes close is 4-(dicyanomethylene)-2-t-butyl-6-(1,1,7,7-tetramethyljulolidyl-9-enyl)-4H-pyran better
Fig. 14. Luminance Efficiency of C-545TB vs. current density.
known as DCJTB [124]. In 1999, the luminance efficiency of DCJTB as reported by Sanyo/Kodak was 1 cd/A for the 2.4 in. LTPS AMOLED. The luminance of red was a factor of 10 lower than the green. When a white emission was demanded, the red component occupied more than half of the power consumption. Later, it was found by adding rubrene as an emitting assist dopant, the energy transfer from Alq3to DCJTB could be achieved more efficiently leading to a desired saturation with
2.2 cd/A efficiency [125]. Further, by adding a carrier trap dopant in Alq3layer or as a separate layer
in the device to adjust the charge balance, Hamada [126] of Sanyo reportedly has achieved an EL efficiency of 2.8 cd/A for DCJTB doped emitter. The exact device structure disclosed was [ITO/NPB (150 nm)/2% DCJTBþ 6% NPB þ 5% rubrene in Alq3(37.5 nm)/Alq3(37.5 nm)/LiF/Al (200 nm)]
which produced a saturated red emission at lmax 632 nm with a CIEx;y ¼ ½0:65; 0:35 at a drive
voltage of 8.5 V with a luminous power efficiency of 1.03 lm/W [127]. Because of the extra NPB dopant in the emitter layer to trap the excess [Alq3]þ, the device was demonstrated to be very stable
attaining an operating half-life of8000 h. The EL efficiency of DCJTB doped Alq3 emitter has
been found to also decrease markedly with increasing current density—a phenomenon that is not observed in the corresponding green coumarin dopants. This phenomenon is attributed to the quenching of excited singlet [DCJTB] by cationic species, either [DCJTB]þ or possibly [Alq3]þ,
but not the corresponding anionic species [128]. Recently, a double heterostructure with narrow recombination zone [129] has been proposed to increase the red EL efficiency, but did little to improve that of DCJTB. It appears that there is still ample room for improvement of the EL efficiency of DCJTB once the quenching mechanism of this remarkable dopant is understood.
An alternative dopant in the same class is the i-propyl derivative DCJTI [130] which is easier to synthesize and more amenable to large scale production without compromising its EL efficiency and chromaticity. Recently, it was reported [131] that an 8-methoxy-substituted derivative (DCJMTB) doped in tris(8-hydroxyquinolinolato)gallium (Gaq3) host matrix at 1% in a device structure of [ITO/
TPD(80 nm)/DCJMTB%:Gaq3 (60 nm)/Gaq3 (20 nm)/Mg:Ag (200 nm)] achieved a luminance
efficiency of 2.64 cd/A at 20 mA/cm2 with a power efficiency of 0.72 lm/W and CIE color coordinates of (x¼ 0:63; y ¼ 0:36). The same dopant in Alq3 could only get 1.91 cd/A with the
same chromaticity. The improved performance in Gaq3 host was rationalized as due to the better
energy level matching between the host and the dopant which enhanced carrier injection and confinement from Gaq3to DCMTB. Similar dopants based on the chromene [132] were much less
efficient, however. Replacing the julolidyl donor in DCJTB with a lesser donating diphenylamino group as in 4-(dicyanomethylene)-2-t-butyl-6-(p-diphenylaminostyryl)-4H-pyran (DCTP) [133] which shifts the PL emission to yellow at lmax 564 nm with a quantum yield of 94%. Its EL
efficiency at 2% doping in Alq3reached 5.3 cd/A and 1.6 lm/W at 11 V and 20 mA/cm2with a CIEx,y
of 0.47; 0.51.
As shown in Fig. 16, other notable red dopants published in the last three years were: 3-(dicyanomethylene)-5,5-dimethyl-1-[(4-dimethylamino)styryl]cyclohexene (DCDDC) [134], 6-methyl-3-[3-(1,1,6,6-tetramethyl-10-oxo-2,3,5,6-tetrahydro-1H,4H,10H-11-oxa-3a-azabenzo[de]-anthracen-9-yl)acryloyl]pyran-2,4-dione (AAAP) [135], 6,13-diphenylpentacene (DPP) [136] and 3-(N-phenyl-N-p-tolylamino)-9-(N-p-styrylphenyl-N-p-tolylamino)perylene [(PPA)(PSA)Pe-1] [137]. DCDDC was claimed to exhibit an EL spectra with a full width at half maximum (FWHM) of 70 nm
when doped in [ITO/PVK:TPD/DCDDC:Alq3/Mg:Ag]. Judging by the EL spectra recorded at
various doping concentration, the energy transfer was rather poor. Dopant AAAP appeared to be a novel structure, but presumably due the presence of too many ketonic carbonyl groups, its luminance efficiency was too low to be attractive. The more interesting DPP was reported to exhibit a narrow emission spectrum when doped at 0.55 mol% into Alq3giving rise to a saturated red peak, centered
appears to rest in its relatively stable external EL quantum efficiency at around 1% over a wide range of drive current density of 1–100 mA/cm2 that is ideal for passive as well as active display applications. The disadvantage of this dopant seems to come from the intrinsic properties of its EL and solid PL emissions which contain several vibronic structures at 680 and 740 nm (shoulder) that are too close to the edge of the infrared spectrum. As a result, DPP’s luminance is compromised in the region where eyes are least sensitive. A new perylene derivative, 3-(N-phenyl-N-p-tolylamino)-9-(N-p-styrylphenyl-N-p-tolylamino)perylene ((PPA)(PSA)Pe-1) designed with separated functional-ities of luminescent center and carriers transport units was found to produce a red emission with 1.0 cd/A at 100 cd/m2 in a triple-layered device of [ITO/starburst amine (25 nm)/(PPA)(PSA)Pe-1 (60 nm)/Alq3/Mg:Ag (140 nm)]. The styryl substituent gives rise to a shoulder at 620 nm in addition
to its lmax emission peaking at 580 nm which shifts the CIEx,y coordinate from the usual orange
emission to a more saturated red at (x¼ 0:64; y ¼ 0:35). Like DPP, the EL efficiency of many of these perylene derivatives with styryl substituents are almost independent of the drive current density. It is reasoned that when the voltage applied to a device increases, the higher carrier mobility of the styrylphenyl-amino-substituted perylene makes the increase in electric field strength in the light-emitting layer smaller. As a result, there will be little change in the distribution of the charge recombination zone in a light-emitting layer with steady electric field strength.
One of the more exciting developments in red emitter was recently disclosed by Sony [138]. The 1,10-dicyano-substituted bis-styrylnaphthalene derivative (BSN) has a film absorption at lmax
507 nm and a strong PL at lmax 630 nm with a quantum efficiency of 0.80. It has good thermal
properties with Tg¼ 115 8C, Tc¼ 161 8C and Tm ¼ 271 8C and was reported to form a good
amorphous thin film on evaporation. Solid-state photo-ionization measurements place its LUMO at 2.93 eV and HOMO at 5.38 eV which favors the efficient recombination of electrons and holes to take place within the emitter layer. In a device of [ITO/2-TNATA/NPB/BSN/Alq3/Li2O/Al], BSN as
a red emitter without dopants displays an impressive luminous efficiency of 2.8 cd/A at 500 cd/m2 with CIE x; y¼ 0:63; 0:37. This red emission apparently can be further improved to the saturated
color of CIE x; y¼ 0:66; 0:34 by Sony’s proprietary top emission adaptive current drive structure (TAC) demonstrated in their 13 in. LTPSi active-matrix TOLED prototype display. It was disclosed further that the color of this device was apparently not affected by the drive current. Thus, BSN rivals the best of doped red fluorescent emitters of today. A similar red electroluminescent dye D-CN based on the bis(styryl) nucleus and synthesized by the bis-Knoevenagel condensation of 1,4-phenylenediacetonitrile and 2 eq. of 4-formyl-40-methoxytriphenylamine has also been reported recently [139]. A trinuclear zinc(II) complex with six 2-[2-(4-cyanophenyl)ethenyl]-8-quinolinolato bidentate ligands (Zn3(2-CEQ)6) [140] was found to have a strong solid PL at lmax 600 nm and a
quantum efficiency of 0.32 with an excited state lifetime of 2.40 ns. This bulky metal complex, if thermally sublimable, could be an interesting host material for use as a red doped emitter.
There were many studies on getting sharp and saturated emission from europium complexes by varying the ligand designs. These dopants however usually have low efficiencies and brightness. Representative examples are: europium tris(dibenzoylmethide)(triphenylphosphine oxide) (Eu(DBM)3(TPPO)) [141,142], europium tris(thienoyl-trifluoroacetone)(phenanthroline)
(Eu(TTFA)3(phen)) 121 and europium
tris(dibenzoylmethide)(1-ethyl-2-(2-pyridyl)benzimida-zole) (Eu(DBM)3(EPBM)) [144]. The fabrication of an OLED device using an ytterbium (III)
complex, tris(dibenzoylmethanato)(monobathophenanthroline)Yb(III) [145] as a near-IR emitting material around 900–1100 nm region for application in optical fiber communications has also been reported.
3.4.3. Blue
The blue doped emitter in OLED often necessitates the judicious selection or design of an appropriate blue host material which has a wide enough bandgap energy and a set of matching LUMO/HOMO level to effect the sensitization. Several examples are shown in Fig. 17. For full-color OLED displays [106], the target to shoot for is around 4–5 cd/A depending upon the color purity which can be from the CIE coordinates of (x¼ 0:14–0.16) and (y ¼ 0:11–0.15). It appears that there are a number of blue dopant/host systems that can achieve this efficiency with CIE x¼ 0:15; y¼ 0:15. But, the status of their device operational lifetimes leaves much to be improved [146]. In commercial circles, one of the best blue emitters used in OLED is believed to have been patented by Idemitsu Kosan Co. based on the basic structure of distyrylarylene (DSA) host doped with a hole transporting amine-substituted DSA dopant [147,148]. This doped blue emitter reportedly could achieve a luminous efficiency of >6 lm/W at 100 cd/m2 with a lifetime of continuous DC operation of >30,000 h and the shelf storage stability is >500 h at 85 8C [149]. The corresponding CIE coordinates were unfortunately not disclosed and the exact nature of the structures remained a heavily guarded trade secret. However, most people in the OLED materials research community should have realized by now that the best dopant/host molecules are probably not what were disclosed in the open literature (e.g. BCzVBi/DPVBi) [150]. TDK in efforts of developing the white OLED emission has also disclosed a blue host material based on phenylanthracene derivatives (PAD). The details of these structures are also sketchy [151]. Kodak also disclosed its doped blue emitter based on 9,10-di(2-naphthyl)anthracene (b-DNA) and tetrakis(t-butyl)perylene (TBPe) [152] which in the device structure of [ITO/CuPc (15 nm)/NPB (60 nm)/DNAþ 2% TBPe (30 nm)/Alq3
(35 nm)/Mg:Ag (200 nm)] reached a luminance efficiency of 3.2 cd/A at 20 mA/cm2with a cyanish blue emission of CIE coordinates of x¼ 0:154; y ¼ 0:232. Since DNA could effectively transport holes due to its relatively high HOMO, it was not clear whether some of the emission might have come from some of the leakage of holes that recombined in the Alq3 electron-transport layer. By
replacing Alq3with a hole blocking, electron-transport material TPBI, the same device achieved an
CIE x¼ 0:137; y ¼ 0:203. The bulky TBPe was reinvestigated by doping in bis(2-methyl-8-quinolinolato)(p-phenylphenolato)aluminium (III) (B-Alq) and found to be more resistant to concentration quenching than the parent perylene. As a result, the constancy of EL color and efficiency are preserved by using the stericly hindered TBPe as a blue dopant [153]. The non-planar blue-emitting TPBI host also could prevent the exciplex formation of dopant BMB-2T with TPD in the device of [ITO/TPD/TPBI:BMB-2T/TPBI/Mg:Ag] [154] where a pure blue emission peaking at 446 nm with CIE coordinates of x¼ 0:15; y ¼ 0:12 was obtained with an efficiency of <1 cd/A.
Many hole transporting materials such as NPB, PPD, TTNND described previously are capable of emitting in the blue. Here, in order to prevent holes from recombination elsewhere, insertion of a hole blocking layer (HBL) such as Salq [53] and TPBI [155] in OLED often becomes necessary to enhance the EL efficiency and maintaining the purity of the blue emission. There were a number of novel pyrazoline derivatives [156] that have been used as emitters in blue OLED, but their performance was not remarkable. Other notable examples of blue-emitting amorphous materials as shown in Figs. 17 and 18 are: 2,5-bis{4-[bis-(9,9-dimethyl-2-fluoreyl)amino]phenyl}thiophene (BFA-1T) [157], spiro-oligo(p-phenylene) [96], 9,10-bis[(200,700-di-t-butyl)-90900 -spirobifluorenyl]an-thracene (TBSA) [158], terfluorene [159] and the triphenylsilyl-substituted derivative PhTPAOXD [160]. Many of these novel materials which have very high glass transition temperatures (with the spiro-terfluorene championing at Tg¼ 296 8C) can act as efficient blue emitters without needing a
dopant. The bis(spirobifluorenyl)anthracene (TBSA) is reported to achieve a luminous efficiency of 1.22 lm/W (3 cd/A) at a drive voltage of 7.7 V and a brightness of 300 cd/m2in the device of [ITO/ CuPc (20 nm)/NPB (50 nm)/TBSA (20 nm)/Alq3 (30 nm)/LiF (1 nm)/Al]. The CIE color of this
emission is x¼ 0:15; y ¼ 0:11 which is claimed to be the nearest to the NTSC standard ever reported for a blue OLED.
Finally, a blue fluorescent boron complex of the dianion of 1,6-bis(2-hydroxy-5-methylphe-nyl)pyridine, (mdppy)BF (solid PL lmax 450 nm) has recently been reported to produce highly
efficient white organic EL from a double-layered device [161]. The ligand (H2mdppy) was prepared
by the reaction of 2,6-dibromopyridine and the Grignard reagent from 2-bromo-4-methylanisole in THF with [NiCl2(dppe)] as catalyst (dppe¼ Ph2PCH2CH2PPh2) followed by demethylation in
molten pyridinium chloride. Reaction of H2mdppy with 1 eq. of BF3 in benzene gave the boron
complex, (mdppy)BF. The white EL, whose spectrum is very broad spanning the entire visible spectrum from 400 to 700 nm, is said to have been caused by exciplex emissions formed at the interface between NPB and (mdppy)BF. It is interesting to note that this simple device [ITO/NPB (60 nm)/(mdppy)BF (60 nm)/LiF (1 nm)/Al (200 nm)] with only one emitting material could achieve an efficiency of 1.5 lm/W (3.3 cd/A) and a luminance of 620 cd/m2at a current density of 20 mA/ cm2and a driving voltage of 7 V with a white color of CIEx;y ¼ ½0:30; 0:36 .
Enhanced blue electroluminescence from vapor deposited p-sexiphenyl (6p) layer has been observed in [ITO/TPD (60 nm)/6p (50 nm)/Mg:Ag] device structure [162]. p-Sexiphenyl was reported to be 5 times more electron-transporting but 10 times worse fluorescent than Alq3.
4-(p-Methoxyphenyl)-3-methyl-1-phenylpyrazol-5-one (MeOPAQ) [163] doped in TPBI in the presence of a hole blocker, N,N-bis(carbazolyl)-4,40-biphenyl (CBP), in the device of [ITO/NPB/CBP/ TPBI:MeOPAQ/TPBI/Mg:Ag] reached a luminance efficiency of 3.8 cd/A at 25 mA/cm2and CIE of x¼ 0:19; y ¼ 0:16.
An efficient blue emitter based on bis(2-methyl-8-quinolinolato)aluminum(III) hydroxide (AlMq2OH) [164] was observed without dopant in a device consisting of [ITO (72 O/&)/CuPc
(10 nm)/NPB (50 nm)/AlMq2OH (80 nm)/LiF (0.7 nm)/Al (80 nm)] which achieved a luminance
efficiency of 5.03 cd/A at 80 mA/cm2. The EL emission centered at 485 nm with FWHM of 80 nm might appear greenish. AlMq2OH which was made by mixing a 1:2.2 equivalent ratio of aluminum
acetate in DI water to 8-hydroxyquinaldine in absolute alcohol (80% yield) was surprisingly quite
stable with an onset degradation temperature of 424 8C and could be train-sublimed at 106Torr. Another tridentate aluminum chelates, tris(2,3-methyl-8-hydroxyquinolinolato)-aluminum(III) (Aldmq2) [165] was reported to emit blue light in a device consisting of [ITO/TPD/Aldmq2/
Mg:Ag] with a EL emission lmax 470 nm and a FWHM of 90 nm. The luminance efficiency was
however, low and judging by the EL curve shape, the emission looked greenish. Finally, an elegantly designed and meticulously synthesized shape-persistent polyphenylene dendrimer 6 from Mullen’s laboratory [166] deserves citation for possible application in the OLED as blue host material. 3.5. Triplet emitting materials
One of the key developments in the advance of modern OLED sciences and technology is the discovery of electrophosphorescence which lifts the upper limit of the internal quantum efficiency of the usual fluorescent dopant-based devices from 25% to nearly 100%. Phosphorescence is inherently a slower and less efficient process, but triplet states constitute the majority of electrogenerated excited states (75%), so the successful utilization of the triplet manifold to produce light should undoubtedly increase the overall luminance. The design and synthesis of triplet emitting materials containing heavy-metal complexes, where strong spin-orbit coupling leads to singlet–triplet state mixing which removes the spin-forbidden nature of the radiative relaxation of the triplet state, are therefore particularly important in achieving high-efficiency electrophosphorescence in OLEDs. Several examples are shown in Fig. 19. One of the first examples of triplet emitting device is based on the red 2,3,7,8,12,13,17,18-octaethyl-12H,23H-porhine platinum(II) (PtOEP) [167] which achieved an external quantum yield of 5.6% doped in CBP [168]. But, the relatively long
phosphorescence lifetime (50 ms) of PtOEP tends to result in triplet–triplet annihilation at high current [109]. Lanthanide europium (Eu) complexes also show triplet emission and have been used in red electroluminescent devices [169]. Later it was found that the red triplet phosphor, bis(2-(20 -benzo[4,5-a]thienyl)pyridinato-N,C30) iridium(acetylactonate) (Btp2Ir(acac)) [170] with a shorter
phosphorescence lifetime (4 ms) leads to a significant improvement in Zextreaching 2.5% at high
current density of 100 mA/cm2 as compared to PtOEP. The red EL spectrum of Btp2Ir(acac) has a
maximum at lmax¼ 616 nm with additional intensity peaks at 670 and 745 nm and the CIE
coordinates are x¼ 0:68; y ¼ 0:32.
An even shorter lifetime (1.8 ms) has recently been reported for Os complexes which were used as red electrophosphorescent dopants blended in poly(N-vinylcarbazole) and 2-t-butylphenyl-5-biphenyl-1,3,4-oxadiazole as the emitting layer in a polymer LED [171]. The emission peaks of the reported Os complexes, ranging from 620 to 650 nm, can be tuned by changing the structures of the ligands because the emission originates from triplet metal-to-ligand charge-transfer excited state. The osmium complexes trap both electrons and holes, which facilitates the direct recombination of holes and electrons on the dopant sites. The peak external quantum efficiency and brightness achieved from the complexes were 0.82% and 970 cd/m2 with CIEx;y ¼ 0:65, 0.33.
One of the best electrophosphorescent green materials is fac-tris(2-phenylpyridine)iridium (Ir(ppy)3) [172] which when doped into CBP host exhibits peak external quantum and power
efficiencies of 28 cd/A and 31 lm/W, respectively. At 100 cd/m2, the external quantum and power efficiencies of 26 cd/A and 19 lm/W were reached at a drive voltage of 4.3 V. By further optimization of materials and device architecture, the efficiencies of the green triplet emitter based on Ir(ppy)3can be improved to nearly double (38 lm/W) [173] at the voltage of 4 V and a luminance
of 105 cd/m2. A tris-ortho-cyclometalated iridium complex based on pinene-substituted 2-phenylpyridine (Ir(mppy)3) has recently been found to reduce concentration quenching because
the sterically hindered pinene spacer in the phosphor molecule minimizes molecular aggregation or interaction [174].
One of the keys to highly efficient phosphorescent emission in OLEDs is to confine the triplet excitons generated within the emitting layer. Thus, by employing starburst perfluorinated phenylene (C60F42) as both hole and exciton blocking layer, and a starburst hole-transport material of 4,40,400
-tris(N-carbazolyl)triphenylamine as a host matrix for the Ir(ppy)3dopant in the emitter, a maximum
external quantum efficiency reaches to 19.2%, and the efficiency is sustained over 15% even at high current densities of 10–20 mA/cm2 which is better than the brightness of fluorescent tubes for lighting. The onset voltage of the electroluminescence is as low as 2.4 V and the peak power efficiency reaches 72 lm/W which fulfills the promise of OLED as a potential low-power display device [175]. This performance can be attributed to the efficient transfer of both singlet and triplet excited states of the host to Ir(ppy)3, leading to high internal efficiency. Another advantage of
Ir(ppy)3 is that it has a short phosphorescent decay time of <1 ms which reduces saturation of the
phosphor at high drive current conditions. In addition to emission from the iridium dopant, it is possible to transfer the exciton energy also to a fluorescent dye by Fo¨rster energy transfer. The Ir dopant in this case acts as a sensitizer, utilizing both singlet and triplet excitons to efficiently pump a fluorescent dye [109] and thus avoiding the quenching of phosphorescence by triplet–triplet annihilation. Based on Ir(ppy)3-sensitized DCJ fluorescence in CBP host matrix, high-efficiency
yellow double-doped OLEDs [176] with external quantum efficiency of 4.1% and EL efficiency of 11 cd/A (3.1 lm/W) have been achieved at a drive current of 10 mA/cm2.
The latest news on the operational lifetime of the green Ir(ppy)3and the red Btp2Ir(acac) doped
emitters under constant DC drive as defined by T1/2 (50% of initial brightness) are 10,000 h
improvement of operational lifetime of the red Btp2Ir(acac) to 10,000 h is possible by optimizing
the architecture of a mixed-layer comprising a hole transporter, an electron transporter and a phosphorescent dopant as the emissive layer [177].
All of the iridium chelates can be synthesized by cyclometalation using IrCl3nH2O onto an
appropriate ligand such as phenylbenzoxazole, phenylbenzothiazole, phenylpyridine and 2-phenylquinoline [178]. These reactions produce chloride-bridged dimers, [C^N2]Ir(m-Cl)2[Ir C^N2]
which can be transformed into a variety of different octahedral iridium(III) complexes [C^N2]Ir(LX).
Judicious choice of the cyclometalating ligand [C^N] can lead to a variety of colors of emission that range from green to red. Recently, DuPont researchers have developed a new synthesis of organometallic iridium complexes that can be prepared in one step by heating IrCl3nH2O in excess
[C^N] ligand in the presence of a catalytic amount of silver trifluoroacetate as promoter [179]. By this method, they have synthesized a variety of fluorinated and trifluoromethylated derivatives with emissions ranging from lmax 506 to 595 nm. It was claimed that by replacing C–H bonds to C–F
bonds by fluorination, there are several potential benefits: (1) the C–H bond is an effective promoter for radiationless decay of an excited state. Replacing it with a C–F bond of lower vibrational frequency can reduce the rate of radiationless deactivation and enhance the PL efficiency; (2) fluorinated compound usually can be sublimed better for thin film deposition; (3) introduction of C– F bond or CF3group can alter the molecular packing and minimize the self-quenching behavior; (4)
fluorination can enhance the electron mobility; (5) HOMO/LUMO levels can be modified by fluorination and thus allow the optimization of carrier injection and the tuning of EL color. As a matter of fact, one of the bluest emissive triplet complexes reported is fac-tris[2-(4,50 -difluorophenyl)pyridine-C02,N] iridium(III) known as Firppy [180] which has lmax¼ 500 nm in
polystyrene with a quantum yield of22% and a triplet lifetime of 4.5 ms. It was also shown for the first time that the fac-tris[5-fluoro-2-(5-trifluoromethyl-2-pyridinyl)phenyl-C,N]iridium (Ir-2h) could be fabricated as a triplet host emitter without the use of CBP [181]. Thus, using Al as the cathode, Ir-2h as the luminescent layer, 4,7-diphenyl-1,10-phenanthroline (DPA) as the electron-transport layer and bis[4-(N,Ndiethylamino)-2-methylphenyl]-(4-methylphenyl)methane (MPMP) as hole-transport layer on top of ITO/glass, intense electroluminescence at 525 nm with an efficiency of 20 cd/A and a maximum radiance of 4800 cd/m2was achieved.
To date, the bluest phosphorescent iridium complex is Firpic [182] which was reported by Thompson and his associates at Universal Display Corp. in the recent ICEL-3 conference (Fig. 20). In their device of [ITO/NPB/CBPþ Firpic/BAlq/Mg:Ag], the external quantum efficiency was reported to reach 5.5% with EL efficiencies of 5 lm/W and 12 cd/A at 100 cd/m2. Its EL emission peaks at 470 nm with a CIE coordinates of x¼ 0:14; y ¼ 0:30 which is still somewhat cyan in color, but definitely bluer than those of Firppy. In the poster, it was also disclosed that a t-butyl-isocyanide
(t-BuNC) ligated iridium complex (Fir(tBuNC)Cl) actually had a triplet blue photoluminescence peak at 450 nm in methylene chloride. But, unfortunately it was too unstable to be fabricated into a device. Both of these new blue iridium complexes were isolated as a mixture of facial and meridonal isomers.
Finally, by controlling exciton diffusion in the multi-layered OLED devices using 6% Firpic/ CBP as the blue emitter, 8% Bt2Ir(acac)/CBP as the yellow and 8% Btp2Ir(acac)/CBP as the red
emitters, high-efficiency white phosphorescent OLEDs have been reported just before submission of this review [183]. One of these white OLED devices which attained a maximum external quantum efficiency of 5:2 0:5%, maximum EL efficiency of 11 1 cd/A, a maximum luminous efficiency of 6:4 0:6 lm/W with CIEx;y ¼ ½0:37; 0:40 at 10 mA/cm2and a maximum luminance of 31,000 cd/
m2 at 14 V, is among the best ever reported to date [184–186].
It should also be mentioned that there appears to be growing interest and research effort in the academia to develop dendrimers which contain site-isolated chromophores [187] that can control charge-transport, exciton formation, fluorescence and intermolecular interaction [188]. These giant dendritic molecules synthesized by building convergently from a highly functionalized core through several generations have been demonstrated to form single-layered devices by solution processing. Although much work still needs to be done before this idea can be materialized and proven device worthy, the best performance of such dendritic OLED reported to date is based on the green phosphorescent fac-Ir(ppy)3cored dendrimers [189] shown in Fig. 21. A first generation dendrimer
doped into the wide-gap 4,40-bis(N-carbazolyl)biphenyl (CBP) host was shown to display a peak external quantum efficiency of 8.1% (28 cd/A) at a brightness of 3450 cd/m2 and 5 mA/cm2.
4. Charge injection and transport
In OLEDs, both operating voltage and luminance efficiency of the devices strongly depend on effective charge injection from the electrodes to the organic medium and charge transport in the organic materials. In general, to achieve the lowest possible voltage it is necessary to have Ohmic interfaces between the organic layers and the charge-injecting contacts and to maximize the drift mobility of both types of carriers. Furthermore, charge injection and charge transport also play an important role in optimizing the device efficiency of an OLED. An unbalanced injection results in an excess of one carrier type that does not contribute to light emission, and it can also result in an enhanced non-radiative recombination because of interactions of excitons with the charge carriers. In general, the two processes are difficult to disentangle on the basis of the electrical characteristics of OLEDs. The charge-injection contacts are studied with three methods in the
