WO1998007202A1

WO1998007202A1 - Gallium nitride based cathodes for organic electroluminescent devices and displays

Info

Publication number: WO1998007202A1
Application number: PCT/IB1996/000780
Authority: WO
Inventors: Supratik Guha; Richard Alan Haight; Samuel Clagett Strite; Ronald Roy Troutman
Original assignee: International Business Machines Corp
Current assignee: International Business Machines Corp
Priority date: 1996-08-08
Filing date: 1996-08-08
Publication date: 1998-02-19
Anticipated expiration: 1999-02-08

Abstract

An organic light emitting device is provided which comprises a substrate (60), an anode electrode (64), a cathode (61), and an organic region (62, 63) in which electroluminescence takes place if a voltage is applied between said anode (64) and cathode (61). The cathode (61) comprises Gallium Nitride (GaN).

Description

DESCRIPTION

Gallium Nitride Based Cathodes for Organic Electroluminescent

Devices and Displays

TECHNICAL FIELD

The present invention pertains to organic electroluminescent devices, arrays, displays and methods for making the same.

BACKGROUND OF THE INVENTION

Organic electroluminescence (EL) has been studied extensively because of its possible applications in discrete light emitting devices, arrays and displays. Organic materials investigated so far can potentially replace conventional inorganic materials in many applications and enable wholly new applications The ease of fabrication and extremely high degrees of freedom in organic EL device synthesis promises even more efficient and durable materials in the near future which can capitalize on further improvements in device architecture.

Organic EL at low efficiency was observed many years ago in metal/organic/metal structures as, for example, reported in Pope et al., Journal Chem. Phys., Vol. 38, 1963, pp. 2024, and in "Recombination Radiation in Anthracene Crystals", Helfrich et al., Physical Review Letters, Vol. 14, No. 7, 1965, pp. 229-231 . Recent developments have been spurred largely by two reports of high efficiency organic EL. These are C.W. Tang et al , "Organic electroluminescent diodes", Applied Physics Letters, Vol. 51 , No. 12, 1987, pp. 913-915, and by a group from Cambridge University in Burroughs et al., Nature, Vol. 347, 1990, pp. 539. Tang et al. made two-layer organic light emitting devices using vacuum deposited molecular dye compounds, while Burroughs used spin coated poly(p-ρhenylenevinylene) (PPV), a polymer.

The advances described by Tang and in subsequent work by the Cambridge group, for example in "Efficient LEDs based on polymers with high electron affinities", N. Greenham et al., Nature, Vol. 365, 1993, pp. 628-630, were achieved mainly through improvements in the design of EL devices derived from the selection of appropriate organic multilayers and contact metals.

Organic EL light emitting devices (OLEDs) function much like inorganic LEDs, except that light is commonly extracted through a transparent electrode deposited on a transparent glass substrate. The simplest possible structure, schematically illustrated in Figure 1A, consists of an organic emission layer 10 sandwiched between two electrodes 1 1 and 12 which inject electrons (e ) and holes (h ¹ ), respectively. Such a structure has been described in the above mentioned paper of Burroughs et al., for example. The electrons and holes meet in the organic layer 10 and recombine producing light. It has been shown in many laboratories, see for example: "Conjugated polymer electroluminescence", D. D. C. Bradley. Synthetic Metals, Vol. 54, 1993, pp. 401-405, and "Carrier tunneling and device characteristics in polymer LEDs", I. D. Parker, Journal of Applied Physics, Vol. 75, No. 3, 1994, pp. 1656-1666, that improved performance can be achieved when the electrode materials are chosen to match the respective molecular orbitals of the organic material forming the organic layer 10. Such an improved structure is shown in Figure 1 B. By choosing the optimized electrode materials 13 and 14, the energy barriers to injection of carriers are reduced, as illustrated. Still, such simple structures perform poorly because little stops electrons from traversing the organic layer 10 and reaching the anode 14, or the holes from reaching the cathode 15.

Figure 2A illustrates a device with a large electron barrier 16, such that few electrons are injected, leaving the holes no option but to recombine in the cathode 15. A second problem, illustrated in Figure 2B, is that the mobilities of electrons and holes in most known organic materials, especially conductive ones, differ strongly. Figure 2B illustrates an example where holes injected from the anode 18 quickly traverse the organic layer 19, while the injected electrons move much slower, resulting in recombination near the cathode 17. If the electron mobility in the organic layer 19 were larger than the holes', recombination would occur near the anode 18. Recombination near a metal contact is strongly quenched by the contact which limits the efficiency of such flawed devices.

Tang, as shown in Figure 3, separated electron and hole transport functions between separate organic layers, an electron transport layer 20 (ETL) and a hole transport layer (HTL) 21 , mainly to overcome the problems described above. In "Electroluminescence of doped organic thin films", C.W. Tang et al., Journal of Applied Physics, Vol. 65, No. 9, 1989, pp. 3610-3616, it is described that higher carrier mobility was achieved in the two-layer design, which led to reduced device series resistance enabling equal light output at lower operating voltage. The contact metals 22, 23 could be chosen individually to match to the ETL 20 and HTL 21 molecular orbitals, respectively, while recombination occurred at the interface 24 between the organic layers 20 and 21 , far from either electrode 22, 23. As electrodes, Tang used a MgAg alloy cathode and transparent Indium-Tin-Oxide (ITO) as the anode. Egusa et al. in "Carrier injection characteristics of organic electroluminescent devices", Japanese Journal of Applied Physics, Vol. 33, No. 5A, 1994, pp. 2741-2745 have shown experimentally that the proper selection of the organic multilayer can lead to a blocking of both electrons and holes at an organic interface remote from either electrode. This effect is illustrated by the structure of Figure 3 which blocks electrons from entering the HTL 21 and holes from entering the ETL by a clever choice of HTL and ETL materials. This feature eliminates non-radiative recombination at the metal contacts as described in Figure 1A and also promotes a high density of electrons and holes in the same volume leading to enhanced radiative recombination. With multilayer device architectures now well understood and commonly used, the major performance limitation of OLEDs is the lack of ideal contact electrodes. The main figure of merit for electrode materials is the position of the bands relative to those of the organic materials (see Bradley and Parker above for detailed discussion). In some applications it is also desirable for the electrode material to be either transparent or highly reflective. The electrode should also be chemically inert and capable of forming a dense uniform film to effectively encapsulate the OLED. It is also desirable that the electrode not strongly quench organic EL.

The choice of a good cathode electrode is particularly difficult. No material has yet been identified which is transparent, conductive, chemically stable, and a good electron injector for OLEDs. Good electron transporting organic materials which are known have their lowest unoccupied molecular orbitals (LUMO) matched only with metals having very low work functions. A low work function in a metal is tantamount to high chemical reactivity. While, e.g., Ca is very well matched in energy to an Alq3 (tris(8-hydroxyquinoline aluminum)) electron transport layer, a Ca cathode survives intact only a short time in air, leading to rapid device degradation. It is also likely that such highly reactive metals undergo a chemical reaction with the nearby organic materials which also could have negative effects on device performance. Such a mechanism has been proposed by Parker in the above cited reference to explain why Sm or Yb cathode OLEDs have poorer performance than Ca cathode OLEDs despite the lower work function of Sm and Yb compared to Ca. A low work function cathode metal approach requires careful handling of the device to avoid contamination of the cathode metal, and immediate, high quality encapsulation of the device if operation in a normal atmosphere is desired. Even well encapsulated low work function metal contacts are subject to degradation resulting from naturally evolved gases from the organic LED materials caused by heating during device operation. On the other hand, the choice of a stable metal having a higher work function, e.g. Al, dictates that the device can only be operated at high voltages. High voltage is necessary because electron injection from Al into, e.g., Alq3 is field assisted. The high operating voltage reduces device efficiency due to increased ohmic losses. In addition, the higher electrical fields present at increased voltages also are likely to degrade the device materials more rapidly by driving interdiffusion or exciting parasitic chemical reactions or recombination processes. Al contacts, of lesser reactivity compared to Mg or Ca, have still been observed to degrade during OLED operation, see e.g. L. M. Do et al., "Observation of degradation processes of Al electrodes in organic EL devices by electroluminescence microscopy, atomic force microscopy, and Auger electron microscopy", Journal of Applied Physics, Vol. 76, No. 9, 1994, pp. 5118-5121 .

Many approaches have been attempted in order to solve the problem of cathode electrode instability, degradation and high injection voltage. A common approach is the use of a low work function metal subsequently buried under a thicker metal coating. In this case, pinholes in the metal still provide ample pathways for oxygen and water to reach the reactive metal below, as is described in: Y. Sato et al., "Stability of organic electroluminescent diodes", Molecular Crystals and Liquid Crystals, Vol. 253, 1994, pp. 143-150 and J. Kido et al., "Bright organic electroluminescent devices with double-layer cathode", IEEE Transactions on Electron Devices, Vol. 40, No. 7, 1993, pp. 1342-1344. Furthermore, such contacts are degraded by evolved gases from and reactions with the OLED constituent materials. The overall lifetime of OLEDs using this approach is limited and extensive encapsulation is required.

I. D. Parker and H. H. Kim, "Fabrication of polymer light-emitting diodes using doped silicon substrates", Applied Physics Letters, Vol. 64, No. 14, 1994, pp. 1774-1776, showed that, depending on the semiconductor doping, Si/Si02 is capable of either hole or electron injection. Si electrodes, due to the small bandgap and moderate work function of Si, have both a barrier to electron and hole injection into organic LED materials, and therefore represent no marked improvement over conventional metals. Parker and Kim avoided this by adding a Si02 interlayer between the Si contact and OLED. While the voltage drop across the Si02 insulator permitted the Si bands to line up with their organic molecular orbital counterpart, electrons were not directly injected, rather forced to tunnel through the Si02 insulator. Such OLEDs had turn-on voltages of > 10 V, too high for efficient device operation.

On the anode side, ITO, which has been preferred primarily for its high work function and its properties as a transparent conductor, is also not ideal. Parker, in the reference cited above, has shown that the replacement of ITO with higher work function Au yields a two-fold improvement in device efficiency. This work reveals that an undesirable barrier to hole injection exists between ITO and preferred HTL materials. There are also questions about the chemical reactivity of ITO contacts. Oxidation of, and In diffusion into the organic layers, arising from the proximity of the ITO contact has been speculated to contribute to OLED degradation. ITO is also polycrystalline, and its numerous grain boundaries provide ample pathways for impurity diffusion.

The lack of inert, stable, energetically matched, and transparent electrode materials for low voltage, efficient and stable OLED operation remains a major obstacle to OLED development.

Organic LEDs have great potential to outperform conventional inorganic LEDs in many applications. One important advantage of OLEDs and devices based thereon is price since they can be deposited on large, inexpensive glass substrates, or a wide range of other inexpensive transparent, semitransparent or even opaque crystalline or non-crystalline substrates at low temperature, rather than on expensive crystalline substrates of limited area at comparatively higher growth temperatures (as is the case for inorganic LEDs). The substrates may even be flexible enabling pliant OLEDs and new types of displays. To dale, the performance of OLEDs and devices based thereon is inferior to inorganic ones for several reasons:

1. High operating voltage: Organic devices require more voltage to inject and transport the charge to the active region (emission layer) which in turn lowers the power efficiency of such devices. High voltage results from the need for high electric fields to inject carriers over energy barriers at the electrode/organic interfaces, and from the low mobility of the carriers in the organic transport layers (ETL and HTL) which leads to a large ohmic voltage drop and power dissipation.

2. Low brightness: Today's OLEDs can produce nearly as many photons per electron as common inorganic LEDs, i.e. their quantum efficiency is good. OLEDs lag inorganic LEDs in brightness mainly because comparatively little charge can be conducted through the resistive transport layers (HTL or ETL). This well known effect is referred to as Space Charge Limited Current. Simply put, due to the low mobility of carriers in organic materials, a traffic jam develops which restricts the flux of electrons and holes reaching the emission layer. Better emitter materials cannot offer greatly improved brightness until high conductance transport layers are also available.

3. Reliability: Organic LEDs degrade in air and during operation. Several problems are known to contribute.

A) Efficient low field electron injection requires low work function cathode metals like Mg, Ca, Li etc. which are all highly reactive in oxygen and water. Ambient gases, and gases coming out of the organic materials during ohmic heating degrade the contacts. B) Conventional AgMg and ITO contacts still have a significant barrier to carrier injection in preferred ETL and HTL materials, respectively. Therefore, a high electric field is needed to produce significant injection current. Stress from the high field and ohmic heating at the resistive electrode/organic interface contribute to device degradation.

C) The high resistance of carrier transport layers heats the device under operation. D) Thermal stability of most OLED materials is poor making them sensitive to heating. Upon heating, many amorphous organic materials crystallize into grains. The crystallites have less volume and pack less uniformly than the amorphous solid. The resulting gaps and odd shapes of the crystallites make conduction from one crystallite to the next difficult, increasing resistance and heating in a positive feedback loop, while opening further channels for gaseous contaminants to penetrate, or for neighboring materials to diffuse.

4. Poor chemical stability: Organic materials commonly used in OLEDs are vulnerable to degradation caused by reaction with and diffusion of contact electrode materials and the ambient atmosphere.

OLEDs are mainly limited by their contacts and transport layers, and feedback from the transport layer heating. It is thus highly desirable to replace the low work function metal based cathodes with stable, possibly transparent cathode characterized by barrierless charge injection into OLEDs. A transparent cathode provides the additional advantage of allowing conventional ITO anodes to be replaced with improved anodes.

However, present day solutions inhibit performance and degrade device reliability. The price of distancing the active layer from the metal contacts for higher recombination efficiency are ohmic voltage drops across the HTL/ETL, leading to heating and power consumption. Low work function metals are unstable and unreliable.

As can be seen from the above examples and the description of the state of the art the contact materials need to be improved to realize OLEDs and displays based thereon with superior characteristics. It is an object of the present invention to provide new and improved organic EL devices, arrays and displays based thereon.

It is a further purpose of the present invention to provide new and improved organic EL devices, arrays and displays based thereon with improved efficiency, lower operating voltage, and increased reliability.

It is a further object to provide a method for making the present new and improved organic EL devices, arrays and displays.

SUMMARY OF THE INVENTION

The above objects have been accomplished by providing an OLED having a cathode comprising gallium nitride (GaN) wide bandgap semiconductor.

The inventive approach capitalizes primarily on the favorable conduction band energy of GaN, as well as its good conductivity, transparency in the visible spectrum, chemical inertness, hardness, and ability to be deposited in the amorphous state at extremely low temperatures on glass, organic thin films, or other amorphous or crystalline substrates. Our experiments have shown that GaN is conductive, even when deposited at room temperature in the amorphous state, highly transparent in the visible spectrum, and has a favorable conduction band alignment for electron injection into the LUMOs of preferred OLED materials. Furthermore. GaN is an excellent encapsulant for OLEDs due to the extremely low diffusivity of impurities in GaN and the nearly amorphous state of material deposited at low temperature. In addition, the considerable strength of the Ga-N chemical bond makes GaN chemically inert, even capable of resisting the attack of highly corrosive acids and bases. We have experimentally observed that GaN - having all of the above favorable properties - can be deposited onto glass, or even directly onto an OLED multilayer structure, and produces a device with improved performance and stability.

In one embodiment of the present invention, a single or multi-layer OLED structure having a GaN cathode directly in contact with the corresponding organic layer, and a conventional opposite contact electrode is envisioned .

In another embodiment of the present invention, an OLED structure having an (AI,ln)GaN cathode directly in contact with the corresponding organic layer, and a conventional opposite contact electrode is envisioned. The purpose of alloying the GaN cathode with small amounts ( < 20%) of AIN is to finely tune the position of the cathode conduction band to the corresponding organic material molecular orbital The purpose of alloying the GaN with InN is to improve the ohmic contacts between the cathode and the outer cathode layer(s).

In yet another embodiment of the present invention, an OLED in which a GaN cathode, preferably an AIGalnN cathode, is separated from the nearest organic layer by a thin metal interlayer, is envisioned. The metal can be selected for its transparency, work function, or properties as a diffusion barrier between the organic materials and GaN, and serves the purpose of further improving the stability or electron injection of the GaN/organic interface.

In yet another embodiment of the present invention, an OLED in which a GaN cathode, preferably an AIGalnN cathode, is in direct contact to the nearest organic layer, but has a thin embedded metal interlayer very near to the GaN/organic interface. The metal can be selected for its work function, properties as a diffusion barrier between the organic materials and GaN, or transparency, and serves the purpose of further improving the stability or electron injection of the GaN/organic interface.

The introduction of a cathode comprising GaN leads to the following advantages:

1. Low voltage carrier injection is realized through the highly favorable band energies of GaN with respect to preferred OLED materials.

2. GaN is highly transparent to visible light. This adds flexibility in the choice of anode designs, e.g. the anode can be opaque.

3. GaN is chemically inert and thermally stable and therefore has no undesirable solid state interactions with the organic layers with which it is in contact or close proximity. 4. GaN is an outstanding encapsulant and mechanical protectant material for OLEDs, due to its nearly amorphous state, hardness and low impurity diffusion constants.

5. GaN can be deposited at conditions required for OLED formation (e.g. low temperature, amorphous substrates, minimum damage to the growth surface) in a conductive state.

6. GaN, as a semiconductor, quenches optical recombination in nearby organic layers less strongly than metals enabling reduced transport layer thicknesses.

There are four preferred GaN-based cathodes for OLEDs. In each case, the device enjoys all of the advantages tabulated above.

1. GaN/Organic Cathode: The GaN is in direct contact with the electron transporting OLED layer. The GaN may be deposited onto the organic material or the organic material may be deposited onto the GaN.

2. InGaN/GaN/AIGaN/organic Cathode: The AIGaN or GaN is in direct contact with the electron transporting OLED layer. Small amounts ( < 20%) of AIN tune the (AI,ln)GaN conduction band to precisely match the organic material . InGaN or GaN is furthest from the organic material and facilitates lateral conduction or ohmic contact formation to the GaN based cathode. The cathode may be deposited onto the organic material or the organic material may be deposited onto the cathode.

3. (AI,ln)GaN/Metal/Organic Cathode: The GaN or (Al .ln)GaN is separated from the electron transporting OLED layer by a thin metal layer. The thickness of the metal layer can range between a partial monolayer to 20 nm. Either the cathode or the organic layer can be deposited first. 4. (AI,ln)GaN/Metal/(AI,ln)GaN Organic Cathode: The GaN or (AI,ln)GaN is in direct contact with the electron transporting OLED layer, and encapsulates a thin metal layer a short distance ( < 50 nm) from the cathode/organic interface. The thickness of the metal layer can range between a partial monolayer and 50 nm. Either the cathode or the organic layer can be deposited first.

DESCRIPTION OF THE DRAWINGS

The invention is described in detail below with reference to the following schematic drawings:

FIG. 1A shows a known OLED having an emission layer and two electrodes.

FIG. 1 B shows another known OLED having an emission layer and two metal electrodes, with work functions chosen such that the energy barriers for carrier injection are reduced .

FIG. 2A shows another known OLED having an emission layer and two metal electrodes, the work function of the anode being chosen such that the energy barrier for hole injection is low, whereas the work function of the cathode poorly matches the emission layer yielding little electron injection and little radiative recombination in said emission layer.

FIG. 2B shows another known OLED having an emission layer with lower electron mobility than hole mobility such that the recombination occurs close to the cathode where it is quenched.

FIG. 3 shows another known OLED having an electron transport layer and hole transport layer.

FIG. 4 shows an optical absorption spectrum of a GaN/glass thin film.

The onset of absorption occurs around 360nm in the near ultraviolet, which is responsible for the high transparency of

GaN grown at low substrate temperatures on all types of substrates. FIG. 5 shows the band structure measured by ultraviolet photoemission spectroscopy for a GaN/Alq3 heterojunction. Experimental error is included in the Alq3 bands as drawn. To within experimental error, the lowest unoccupied molecular orbital (LUMO) of Alq3 lies at equal or lower energy compared to the GaN conduction band minimum meaning that electron injection from GaN to Alq3 is barrierless.

FIG. 6 shows a cross section of the first embodiment of the present invention in which a GaN cathode is deposited on the substrate before the organic layer stack. Light is emitted through the transparent cathode and glass substrate into the plane below the substrate.

FIG. 7 shows a cross section of the second embodiment of the present invention in which a GaN cathode is deposited on top of the organic layer stack. Light is emitted through the transparent cathode into the plane above the substrate.

FIG. 8 shows a cross section of the third embodiment of the present invention in which an AIGalnN cathode is deposited on top of the organic layer stack. Light is emitted through the ITO anode and glass substrate into the plane below the substrate. If a transparent outer cathode metal other than Al would be chosen, the cathode would also be transparent, and light would be emitted into both planes above and below the substrate.

FIG. 9 shows a cross section of the fourth embodiment of the present invention in which a GaN cathode deposited on top of the organic layer stack is separated by a thin metal interlayer. The metal serves to either improve injection, protect the organic during GaN deposition, or act as a diffusion barrier. Light is emitted through the transparent cathode plane above the substrate.

FIG. 10 shows a cross section of the fifth embodiment of the present invention in which a GaN cathode encapsulates a thin metal layer which is situated very near to the organic layer stack. The thin GaN layer permits electrons to pass freely from the metal to the organic layer, either by tunneling or thermionic emission, while chemically and physically isolating the metal from the organic. Light is emitted into both the planes above and below the substrate.

FIG. 11 shows a cross section of a display or array according to the present invention. The transparent GaN based top cathode permits an opaque Si substrate to be chosen. The Si substrate can be an integrated circuit providing the display function to the OLED array deposited on top of it.

FIG. 12 shows a cross section of a display or array according to the present invention in which a GaN based cathode is deposited onto an opaque Si substrate. The GaN cathode provides a stable cathode which can withstand typical handling and processing of the Si wafer.

FIG. 13 shows a cross section of a display or array according to the present invention in which a transparent GaN based cathode down geometry is used on a glass substrate. Light is emitted in the plane below the glass substrate. GENERAL DESCRIPTION

GaN is an ideal material for electron injection into OLEDs, based both on what is known of the material in the literature, and what we have discovered in our laboratory. The physical properties of GaN are catalogued in: S. Strife and H. Morkoc, "GaN, AIN and InN: a review", Journal of Vacuum Science and Technology B, Vol. 10, 1992, pp. 1237-1266 and "Properties of Group III Nitrides", edited by James H. Edgar, (The Institution of Electrical Engineers, London 1994). The properties of GaN as a transparent conductor have been described by H. Sato et al., "Transparent and conductive GaN thin films prepared by an electron cyclotron resonance plasma metalorganic chemical vapor deposition method", Journal of Vacuum Science and Technology A, Vol. 11 , No. 4, 1993, pp. 1422-1425. As reviewed above, an ideal contact electrode material should be characterized by:

1. Depositability onto organic layers, amorphous, crystalline or polycrystalline substrates at low temperatures with little damage to the underlying material.

2. Favorable energy band levels for injection of charge into preferred OLED materials.

3. Sufficient electrical conductivity so that total device series resistance is unaffected by the electrode.

4. High transparency in the visible spectrum (for low quenching and flexibility in light extraction).

5. Chemical inertness

6. Low diffusivity of impurities

7. Mechanical hardness GaN can be deposited at low temperatures via magnetron sputtering, laser ablation, plasma enhanced molecular beam deposition (PEMBD) or other related techniques in which the energy required to create a reactive nitrogen radical is supplied from some external stimulus, and not thermal energy at the substrate. PEMBD, in which thermally evaporated Ga atoms react with low kinetic energy nitrogen radicals at the substrate surface, is our preferred method due to the small amount of chemical and/or kinetic damage endured by the substrate and the large supply of reactive nitrogen species created by the plasma. Ga can be supplied from thermally evaporated elemental metal, or from a Ga containing compound or gas. Active N is supplied by the plasma excitation which excites and cracks a N bearing gas, typically N₂, NH₃, hydrazine. Using this method, undoped GaN has a resistivity of roughly 10 Ohm-cm when grown at room temperature. OLEDs incorporating a GaN cathode grown onto the organic layer stack show comparable properties to inverted structures in which the GaN cathode is grown first, demonstrating that GaN can be grown by the PEMBD technique without damaging the organic layer below. The temperature at which the GaN is grown should be below 150 degree centigrade.

We have studied the band structure of GaN deposited onto glass at low temperature. Figure 4 is a transmission spectrum of one such thin film. These data indicate that low temperature amorphous GaN, much like crystalline GaN, has a wide bandgap energy of 3.3 - 3.4 eV, making it highly transparent to visible light. In order to determine the energy position of the GaN conduction and valence bands relative to common OLED materials, a clean GaN surface was prepared onto which a thin layer of Alq3 was vacuum deposited. In this arrangement, Ultraviolet Photoemission Spectroscopy can resolve the relative positions in energy of the GaN valence band and the Alq3 highest occupied molecular orbital (HOMO). These data, depicted in Figure 5 and combined with our knowledge of the GaN and Alq3 bandgap, indicate that the GaN conduction band is positioned at equal or higher energy with respect to the Alq3 lowest unoccupied molecular orbital (LUMO), meaning that electron injection from GaN into electron transporting Alq3 can proceed with no or even a negative intermediate barrier.

Further confirmation of the favorable energy positioning of the GaN conduction band comes from measurements of actual OLED device structures, e.g. a GaN cathode structure having the following layer sequence from the glass substrate up: Glass/ITO/CuPc/NPB/Alq3/GaN/lnGaN/AI.

Please note that the abbreviation NPB means:

N,N'-diphenyl-N,N'-bis-(4-phenylphenyl)-1 , -biphenyl-4,4'-diamine. This device performed better than conventional Glass/ITO/CuPc/NPB/Alq3/MgAg

OLED devices, both in terms of operating voltage and external efficiency.

InGaN was added to facilitate ohmic contact from the Al to the GaN .

Polymer based OLEDs also show improved device performance when GaN cathodes are formed, which is expected since injection occurs via the same mechanisms in polymers as molecular organic compounds.

Our experimental results have shown that GaN fulfills the first four points of the above list describing an ideal contact electrode. That GaN fulfills the next three points ( 5-7) is apparent from the technical literature on GaN which is readily available, e.g. in the above referenced review paper by Strite and Morkoc or the book entitled "Properties of Group III Nitrides."

In order to overcome the problems of convenlional metal cathodes for discrete light emitters, light emitting arrays and display applications. improved structures which capitalize on the favorable physical properties of GaN, as illustrated in Figures 6-10, are provided, enabling new array and display applications as illustrated in Figures 1 1-13.

Several embodiments of improved OLEDs incorporating GaN contact electrodes are now detailed in connection with Figures 6-10.

The simplest embodiment of a GaN cathode OLED, already improved with respect to the state of the art is depicted in Figure 6. From the substrate up, listed in the order of deposition, is a glass/GaN/ETL/HTL/Metal OLED structure. In addition to the lower barrier to electron injection afforded by the cathode 61 comprising GaN formed on the glass substrate 60, the ETL 62 thickness may be reduced as a result of reduced optical quenching and the conventional ITO anode can be replaced by a higher work function metal since the anode 64 must no longer serve as the transparent contact. We note here that the structure depicted in Figure 6 might also benefit from the addition of an additional layer 61.1 (e.g. ITO) between the GaN 61 .2 and glass 60 lo lower the lateral sheet resistance of the cathode 61. Finally, any substrate, even an opaque one, can replace the glass substrate 60 depicted. In this case, the preferred embodiment would have a transparent top contact 64, e.g. ITO in the case of a cathode 61 comprising GaN. The organic region 65 of the first embodiment comprises an ETL 62 and HTL 63. It is to be noted that the present Figure and all other Figures are not drawn to scale.

Table 1 : Exemplary details of the first embodiment

Layer No. Material Width present example

substrate 60 glass 0.1 mm-5mm 1 mm

outer cathode 61.1 ITO 10-300 nm 50 nm

cathode 61.2 GaN 10-300nm 50nm

ETL and EL 62 Alq3 20-1000nm 80nm

HTL 63 TAD 5-500nm 50nm

anode 64 Au 10-2000nm 50nm

A second embodiment of a GaN cathode device is depicted in Figure 7. From the substrate 70 up, listed in the order of deposition, is a giass/Metal/HTL/EL/ETL/GaN OLED structure. The major difference between Figure 7 and Figure 6 is that the cathode 75 comprising GaN is deposited last on top of the organic layer stack 72-74, which in this case includes a separate emission layer 73 (EL) as is sometimes practiced in the art. We note that structures in which the GaN layer 75.1 being part of the cathode 75 is directly deposited onto the OLED stack 72-74 have similarly improved performance in comparison to structures such as that depicted in Figure 6. ITO might be substituted for the metal anode 71 of Figure 7 if that contact needs to be transparent. Also, the cathode 75 might comprise additional layer or layers, e.g. an InGaN layer 75.2, an ITO layer 75.3, might be grown on top of the GaN layer 75.1 in order to reduce the lateral sheet resistance of the contact and improve electron injection from the ITO 75.3 into the GAN 75 1 . Any substrate other than glass can be chosen, even an opaque one. In the latter case, the cathode 75 is preferably designed to be fully transparent for ease of light extraction. The organic region 75 of the second embodiment comprises an ETL 74, a layer 73 suited for electroluminescence (EL) and HTL 72.

Table 2: Exemplary details of the second embodiment

Layer No. Material Width present example

substrate 60 glass 0.1 mm-5mm 1 mm

anode 71 Ni 10-300 nm 50 nm

HTL 72 TAD 5-500nm 50nm

EL 73 coumarine- doped Alq3 20-1000 nm 70 nm

ETL 74 Alq3 M OOnm 20nm

cathode contact 75.1 GaN 10-2000nm 50nm

ohmic contact layer 75.2 InGaN 10-2000nm 50nm

outer cathode 75.3 ITO 10-2000nm 50nm

A third embodiment of the present invention is depicted in Figure 8. From the substrate 80 up, listed in the order of deposition, this embodiment comprises glass/ITO/HTL/ETL/(AI,ln)GaN OLED structure. In this case, the GaN has been alloyed with AIN to match the cathode conduction band to the LUMO of the ETL. The improved alignment may be necessary if an ETL is selected which has a higher LUMO energy than the GaN CB. We note that these concepts are equally valid for cathode down structures in which the GaN cathode is grown prior to the organic layers and top anode. GaN has been combined with ITO to improve ohmic contact between the cathode 84.1 and outer cathode 84.3. The organic region 85 of the third embodiment comprises a combined ETL/EL layer 83 and a HTL 82. Table 3: Exemplary details of the third embodiment

Layer No. Material Width present example

substrate 80 glass 0.1 mm-5mm 1 mm

anode 81 ITO 10-300nm 50nm

HTL 82 TAD 5-500nm 50nm

ETL and EL 83 Oxadiazole 50-1000nm 70nm

cathode 84.1 AI₀ ,Ga_{0 9}N 10-2000nm 50nm

ohmic contact layer 84.2 InGaN 10-2000nm 50nm

outer cathode 84.3 Al 10-2000nm 50nm

A fourth embodiment of the present invention is depicted in Figure 9. From the substrate 90 up, listed in the order of deposition, is a glass/ITO/HTL/ETL/TM/GaN OLED structure. In this case, the thin metal (TM) 94.1 is chosen either for its transparency, low work function or properties as a chemical, diffusion or protective barrier between the organic ETL and GaN 94.2. The TM may be helpful in improving the alignment of the GaN 94.2 CB to the LUMO of the ETL 93, in which case GaN 94.2 doubles as both an electron injector and an encapsulant for the OLED and TM. The TM 94.1 may improve reliability by inhibiting chemical reactions or diffusion between the organic and the GaN layer 94.2 during or after growth, or as a protective barrier to shield the ETL 93 from degradation caused by the GaN 94.2 deposition We note that these concepts are equally valid for cathode down structures in which the GaN then the TM are grown prior to the organic layers and top contact, and also for (Al.ln)GaN alloys according lo the third embodiment. The organic region 95 of the fourth embodiment comprises a combined ETL/EL layer 93 and a HTL 92.

Table 4: Exemplary details of the fourth embodiment

Layer No. Material Width present example

substrate 90 glass 0.1 mm-5mm 1 mm

anode 91 Pt 10-300nm 50nm

HTL 92 TAD 5-500nm 50nm

ETL and EL 93 Alq3 20-1000nm 70nm

thin metal 94.1 Ca O.OMOnm 1nm

cathode 94.2 GaN 10-2000nm 50nm

ohmic contact layer 94.3 InGaN 10-2000nm 50nm

outer cathode 94.4 ITO 10-2000nm 50nm

Finally, Figure 10 depicts an OLED structure in which a GaN or (AI,ln)GaN cathode encapsulates an embedded TM interlayer near the cathode/organic interface. The thin GaN layer between the TM and the organic stack serves as a chemical and diffusion barrier lo prevent mixing and reactions between the metal and organic materials. The GaN layer between the TM and the organic is thin enough to permit electrons to easily pass from the TM to the organic region by either tunneling or thermionic emission. We note that this concept is equally valid for both cathode down and up architectures, as well as for (AI,ln)GaN cathodes. The organic region 105 of the fourth embodiment comprises a combined ETL/EL layer 103 and a HTL 102.

Table 5: Exemplary details of the fifth embodiment

Layer No. Material Width present example

substrate 100 glass 0.1 mm-5mm 1 mm

sub-anode 101 ITO 10-300nm 50nm

HTL 102 TAD 5-500nm 50nm

ETL and EL 103 Alq3 20-1000nm 70nm

cathode interlayer 104.1 GaN 1-50nm 10nm

TM 104.2 Li 1-20nm 2nm

cathode 104.3 GaN 10-200nm 50nm

ohmic contact 104.4 InGaN 10-200nm 50nm

outer cathode 104.5 ITO 10-2000nm 50nm

In the following, some display embodiments, based on and enabled by the present invention, are disclosed.

It would be advantageous if one could integrate OLEDs onto Si substrates because prior to OLED deposition, the substrate could be fabricated to contain active Si devices, such as for example an active matrix, drivers, memory and so forth. Such a structure can be a very inexpensive small area organic display with high resolution and performance realized in the Si An OLED, OLED arrays or an OLED display may either by grown directly on such a Si substrate carrying Si devices, or it may be fabricated separately and flipped onto the Si substrate later. A problem is the Si metallization. Traditional OLED cathode metals are not stable in Si processes or air. Another problem is that a transparent top contact is needed because Si is not a transparent substrate. The present invention offers a solution to these problems. The disclosed GaN based cathode permits a stable, low voltage contact to be formed on top of the standard Si process metallizations, and are therefore compatible with OLED technology.

An organic array or display structure formed on a Si substrate is illustrated in Figure 11 and described in Ihe following. This display comprises a Si substrate 1 10 which has integrated circuits comprising active and/or passive devices such as memory cells, drivers, capacitors, transistor etc. (these devices are not shown). On top ot the Si integrated circuit, a stable OLED anode (e.g. ITO, Au, Pt, Ni, Cr) material 11 1 is patterned to connect the Si devices to the OLEDs 112. An OLED, in the cathode-up geometry is deposited on the patterned anodes 111 and Si substrate 1 10. Finally, a GaN-based cathode 113 is provided. II is to be noted that no details of the OLED(s) are shown for sake of simplicity, but the OLED may be any color including blue or white. For the examples of blue and white OLEDs respectively, full color might be realized by patterning a color conversion dye array or color filter array on top of the cathode.

For example, an Al-metallized Si chip 110 on which ITO anodes 11 1 are patterned may serve as substrate for an OLED array or display 1 12. One such OLED comprises (from the bottom to the top)- a stable anode layer, e.g. ITO 1 11 , a HTL, an organic doped or undoped active region, an ETL, and a cathode 1 13 which comprises GaN. This cathode 113 may for example be composed of the following stack of layers': TM/GaN/lnGaN/ITO. The organic region of the present devices may - in addition to charge transport layers if needed at all - either comprise:

• a stack of more than one organic emission layers (EL), or

• an organic compound doped with one or more impurities, organic or inorganic, chosen to dominate and enhance the electroluminescence, or

• a stack of more than one organic emission layer, some of which may be doped to dominate or enhance the electroluminescence of that particular organic emission layers, or

• a stack of more than one organic layer, in which the role of one or more of said organic layers is to electrically confine one or more carrier types to improve the emission of an adjacent organic layer.

Another array or display embodiment, where the OLEDs 122 have the anode up, is illustrated in Figure 12. In this Figure, OLEDs 122 on top of a Si substrate 120 are schematically shown. In this case, the Si substrate 120 is partially covered by Al metal electrodes 121 .1 which inject charge into the GaN-based cathodes 121 . Other areas 120.1 do not conduct current. In addition, the Si IC substrate 120 could be planarized during the back end of the Si processing. This approach lowers processing cost because a blanket GaN-based cathode 121 can be deposited immediately before OLED deposition, and does not require additional patterning. As discussed above, this is possible because the intended vertical current must traverse a distance much smaller than the spacings between Al contact pads 121.1. Simple geometry insures that little crosstalk will occur, even when the anode 123 is common to all devices, as is shown in Figure 12. The top anode 123 must be transparent since the Si substrate 120 is opaque to most visible light. Color could also be conveniently incorporated into the embodiment shown in Figure 12, as with Figure 1 1 also, by means of color filters and/or converters 124 which are deposited or patterned onto the transparent top contact.

The anode up embodiment on Si of Figure 12 may have advantages compared to the cathode up version of Figure 1 1 which arise from the generally higher hole mobilities in preferred HTL layers compared to electron mobilities in preferred ETL layers. If any damage to the upper organic layer occurs during electrode deposition, or contamination diffuses through the electrode and degrades the HTL, it could still have a higher mobility than the buried and ungraded ETL, and therefore not be the limiting factor in overall current conduction. Simply put, since the HTL initially outperforms the ETL in known OLED devices, the device is less sensitive to the initial stages of degradation of the HTL than the ETL.

Another possible display embodiment, illustrated in Figure 13, is described below. This display comprises a transparent substrate 130 on top of which amorphous-Si or poly-Si structures are formed using the same technology developed for active matrix liquid crystal displays. Usually the Si is structured to provide thin-film-transistors 131 (TFTs) and other devices, to produce an active matrix. The Si devices 131 formed may then be covered or planarized by special layers 134. Color filters or color converters 132 can be provided, in addition, if the OLEDs 135 emit white or blue light, respectively. The Si devices 131 include structured GaN-based cathodes 133, for example, onto which the OLEDs 135 can be deposited. An advantage of this approach is that entrenched active matrix liquid crystal display (AMLCD) technology can be leveraged in combination with OLEDs to realize inexpensive, high performance AM displays over large areas. Furthermore, clever design permits light to be emitted through the glass substrate 130 so no transparent top contact (anode 136) is needed. The anode 136 may be covered by a cap layer 137.

In the following some examples of the different organic materials which can be used are given. Electron transport/Emitting materials:

Alq3, Gaq3, Inq3, Scq3, BAIq3 (q means 8-hydroxyquinoline) and other 8-hydroxyquinoline metal complexes such as Znq2, Beq2, Mgq2, ZnMq2, BeMq2, and AIPrq3, for example. These materials can be used as ETL or emission layer. Other classes of electron transporting materials are deficient nitrogen containing systems, for example oxadiazoles like PBD (any many derivatives), triazoles, for example TAZ (1 ,2,4-triazole). These functional groups can also be incorporated in polymers, starburst and spiro compounds. Further classes are materials containing pyridine, pyriimidine, pyrazine and pyridazine functionalities. Finally, materials containing quinoline, quinoxaline, cinnoline, phthalazine and quinaziline functionalities are well known for their electron transport capabilities. Other materials are cyano-substituted polymers, didecyl sexithiophene (DPS6T), bis-triisopropylsilyl sexithiophene (2D6T), Azomethin-zinc complexes, pyrazine (e.g. BNVP), strylanthracent derivatives (e.g. BSA-1 , BSA-2), non-planar distyrylarylene derivatives, for example DPVBi (see C. Hosokawa and T. Kusumoto, International Symposium on Inorganic and Organic Electroluminescence 1994, Hamamatsu, 42), cyano PPV (PPV means poly(p-phenylenevinylene)) and cyano PPV derivatives.

The following materials are particularly well suited as

Emission layers and Dopants:

Anthracene, phyridine derivatives (e.g. ATP), Azomethin-zinc complexes, pyrazine (e.g. BNVP), strylanthracent derivatives (e.g. BSA-1 , BSA-2), Coronene (also suited as dopant), Coumarin (also suited as dopant), DCM compounds (DCM1 , DCM2; both also suited as dopants), distyryl arylene derivatives (DSA), alkyl-substituted distyrylbenzene derivatives (DSB), benzimidazole derivatives (e.g. NBI), naphthostyrylamine derivatives (e.g. NSD), oxadiazole derivatives (e.g. OXD, OXD-1 , OXD-7), N,N,N',N^'-tetrakis(m-methylphenul)-1 ,3-diaminobenzene (PDA), Perylene and Perylene derivatives, phenyl-substituted cyclopentadiene derivatives, 12-phthaloperinone derivatives (PP), squarilium dye (Sq), 1 ,1 ,4,4-tetraphenyl-1 ,3-butadiene (TPBD), sexithiophene (6T), poly(2,4-bis(cholestanoxyl)-1 ,4--phenylene-vinylene (BCHA-PPV),

Polythiophenes, quinacridones (QA) (see T. Wakimoto et al., International Symposium on Inorganic and Organic Electroluminescence, 1994, Hamamatsu, 77), and substituted quinacridones (MQA), Rubrene, DCJT (see for example: C. Tang, SID Conference San Diego; Proceedings, 1996, 181 ), conjugated and non-conjugated polymers, for example PPV and PPV derivatives (soluble precursor), MEH-PPV

(poly(2-methoxyl,5-(2'ethyl-hexoxy)-1 ,4-phenylene-vinylene), dialkoxy and dialkyl PPV derivatives, segmented PPVs (see for example: E. Staring in International Symposium on Inorganic and Organic Electroluminescence. 1994, Hamamatsu, 48. and T. Oshino et al . in Sumitomo Chemicals, 1995 monthly report).

Hole transport layers and Hole injection layers:

The following materials as suited as hole injection layers and hole transport layers. Materials containing aromatic amino groups, like TDP, NPB (see C. Tang, SID Meeting San Diego, 1996, and C. Adachi et al. Applied Physics Letters, Vol. 66, p. 2679, 1995), TPA, NIPC, TPM, DEH (for the abbreviations see for example: P. Borsenberger and D.S. Weiss, Organic Photoreceptors for Imaging Systems. Marcel Dekker, 1993). These aromatic groups can also be incorporated polymers, starburst (for example: TCTA, m-MTDATA, see Y. Kuwabara et al ., Advanced Materials, 6, p. 677, 1994, Y. Shirota et al., Applied Physics Letters, Vol. 65, p. 807, 1994) and spiro compounds. Further examples are: Cu(ll) phtalocyanine (CuPc), NPB

(N,N'-diphenyl-N,N'-bis-(4-phenylphenyl)-1 , 1 ' biphenyl-4,4 -diamine) distyryl arylene derivatives (DSA), naphthalene, naphthostyrylamine derivatives (e.g. NSD), Quinacridone (QA; also suited as dopant), poly(3-methylthiophene) family (P3MT), Perylene and Perylene derivatives, polythiophene (PT), 3,4,9, 10-perylenetetracarboxylic dianhydride (PTCDA) (also suited as isolator), tetra phenyldiaminodiphenyl (TPD-1 , TPD-2, or TAD), PPV and some PPV derivatives, poly(2-methoxyl,5-(2'ethyl-hexoxy)-1 ,4-phenylene-vinylene (MEH-PPV), poly(9-vinylcarbazole) (PVK), discotic liquid crystal materials (HPT).

There are many other organic materials known as being good light emitters, and many more will be discovered. These materials can be used as well for making light emitting structures according to the present invention. Examples of such materials are given in the publications cited in the introductory portion of the present description. The contents of these publications is herewith incorporated by means of reference.

Additionally, blend (i.e. guest host) systems containing active groups in a polymeric binder are also possible. The concepts employed in the design of organic materials for OLED applications are to a large exlent derived from the extensive existing experience in organic photoreceptors. A brief overview of some organic materials used in the fabrication of organic photoreceptors is found in the above mentioned publication of P. Brosenberger and D.S. Weiss, and in Teltech, Technology Dossier Service, Organic Electroluminescence (1995), as well as in the primary literalur.

OLEDs have been demonstrated using polymeric, oligomeric and small organic molecules. The devices formed from each type of molecules are similar in function, although the deposition of the layers varies widely. The present invention is equally valid in all forms described above for polymeric and oligomeric organic light emitting devices.

Small Molecule devices are routinely made by vacuum evaporation. This is extremely compatible with PEMBD of GaN. Evaporation can be performed in a Bell jar type chamber with independently controlled resistive and electron-beam heating of sources. It can also be performed in a Molecular Beam Deposition System incorporating multiple effusion cells and electron-beam evaporators. In each case, GaN deposition can occur in the same chamber, a vacuum connected chamber, or even a separate chamber if some atmospheric contamination is tolerable. Oligomeric and Polymeric organics can also be deposited by evaporation of their monomeric components with later polymerization via heating or plasma excitation at the substrate. It is therefore possible to alloy these by co-evaporation also, and they are fully compatible with monomeric compounds.

In general, polymer containing devices (single layer, multilayer, polymer blend systems, etc.) are made by dissolving the polymer in a solvent and spreading it over the substrate either by spin coating or the doctor blade technique. After coating the substrate, the solvent is removed by heating or otherwise. This method allows the fabrication of well defined multilayer structures, provided that the respective solvents for each subsequent layer do not dissolve previously deposited layers. Additionally, hybrid devices containing both polymeric and evaporated small organic molecules are possible. In this case, the polymer film is generally deposited first, since evaporated small molecule layers cannot withstand much processing. More interesting is the possibility of making a polymer/inorganic transport layer on top of which monomeric layers are evaporated, possibly also incorporating alloys. If the polymer is handled in an inert atmosphere prior to introduction to vacuum, sufficient cleanliness for device fabrication is maintained. In any case, the chemical inertness of GaN and other n-d WBS makes it highly tolerant of polymer OLED processing.

To summarize, all inventions here are fully compatible to polymeric, oligomeric, and small molecule OLED designs, or any hybrid design thereof.

Claims

1. Organic light emitting device having a) a substrate (60; 70; 80; 90; 100: 1 10; 120; 130), b) an anode electrode (64; 71 ; 81 ; 91 ; 101 ; 1 1 1 ; 123; 136), c) a cathode (61 ; 75; 84; 94; 104; 113; 121 , 133), and d) an organic region (65; 76; 85; 95; 105; 1 12; 122: 135) in which electroluminescence takes place if a voltage is applied between said anode and cathode, said device being characterized in that said cathode comprises Gallium

Nitride (GaN).

2. The light emitting device of claim 1 , wherein the sequence of layers is: substrate/cathode/organic region/anode.

The light emitting device of claim 2, wherein light generated by said electroluminescence is either emitted from said organic region through said anode, or from said organic region through said cathode and substrate, or both.

The light emitting device of claim 1 , wherein the sequence of layers is: substrate/anode/organic region/cathode.

5. The light emitting device of claim 4, wherein light generated by said electroluminescence is either emitted from said organic region through said cathode, or from said organic region through said anode and substrate, or both.

6. The light emitting device of claim 1 , wherein said organic region comprises a single organic layer or a stack of organic layers. ι

7. The light emitting device of claim 1 , wherein said Gallium Nitride is alloyed with Aluminum, Indium, Arsenic, Phosphorous, Antimony, or some combination of any two or more.

5 8. The light emitting device of claim 1 , wherein said cathode (94; 104) comprises an interlayer (94.1 ; 104.2).

9. The light emitting device of claim 8, wherein said interlayer is either in direct contact with said organic region, or included in said cathode

10 near to the interface between said organic region and said cathode.

10. The light emitting device of claim 8, wherein said interlayer (94.1 ; 104.2) comprises a metal.

15 1 1. The light emitting device of claim 1 , wherein said organic region (65; 76; 85: 95; 105) comprises an electron transport layer (62; 74; 83; 93; 103) being arranged such that said cathode (61 ; 75: 84; 94; 104) is in direct contact with said electron transport layer.

2 12. The light emitting device of claim 1 , wherein said substrate (70; 80; 90; 100; 130) is transparent or semitransparent.

13. The light emitting device of claim 1 , wherein said substrate consists of Silicon, glass or plastic.

25

14. The light emitting device of claim 1 , wherein said substrate is flexible.

15. The light emitting device of claim 1 , wherein said substrate is a Silicon substrate (110; 120) comprising integrated circuits.

30

16. The light emitting device of claim 1 , wherein said Gallium Nitride is either polycrystalline or amorphous.

17. The light emitting device of claim 2, wherein said cathode comprises an Indium-Tin-Oxide (ITO) layer separating said Gallium Nitride from said substrate.

18. The light emitting device of claim 1 , wherein said organic region comprises either

• a stack of more than one organic emission layers (EL), or

19. Organic light emitting array or display comprising more than one light emitting device pursuant to any of the preceding claims.

20. The organic light emitting array or display of claim 19, wherein said substrate is a Silicon substrate comprising devices and/or circuits and/or electrical connections.

21. The organic light emitting array or display of claim 20. wherein said devices and/or circuits and/or electrical connections are designed for driving and controlling at least one of said light emitting devices.

22. The organic light emitting array or display of claim 20, comprising color filters ( 132) providing for the emission of light at different wavelengths, or color conversion layers (132), providing for the conversion of the organic light emitting device's emissive wavelength. ι

23. The organic light emitting array or display of claim 20, wherein said light emitting devices are deposited cathode first onto said Silicon substrate, and wherein

• said cathode comprising Gallium Nitride injects electrons 5 efficiently into the organic region of said light emitting devices, and

• the anode of said light emitting devices is transparent or semitransparent, arranged such that light emitted by said light emitting devices is ιo emitted into the half space above the Silicon substrate plane.

24. The organic light emitting array or display of claim 20, wherein said light emitting devices are deposited anode first onto said Silicon substrate, and wherein said cathode comprising Gallium Nitride is

15 transparent or semitransparent such that light emitted by said light emitting devices is emitted into the half space above the Silicon substrate plane.

25. The organic light emitting array or display of claim 21 , wherein said 20 light emitting devices are formed such that light is emitted from said organic region through said cathode and substrate into the half space below said substrate plane.

26. The organic light emitting array or display of claim 19. wherein the 25 substrate and the entire array or display are flexible.

27. Method for making organic light emitting devices having a substrate, an anode electrode, a cathode, and an organic region in which electroluminescence takes place if a voltage is applied between said

3 anode and cathode, comprising the steps of: a) forming said cathode such that it comprises Gallium Nitride (GaN).

28. The method of claim 27, whereby said cathode is formed using • an electron cyclotron resonance or radio-frequency (RF) plasma, or

• metalorganic chemical vapor deposition, or

• magnetron sputtering, or • laser ablation, or

• a plasma enhanced molecular beam deposition.

29. The method of claim 27, whereby said cathode is formed at low temperature below 150 degree centigrade, and preferably at a temperature in the range of the room temperature.