WO2016207233A1 - Organic light emitting device comprising polar matrix, metal dopant and silver cathode - Google Patents

Abstract

The present invention relates to an electronic device comprising at least one light emitting layer between an anode and a substantially silver cathode, the device further comprising between the cathode and the anode at least one mixed layer comprising (i) at least one substantially covalent electron transport matrix compound comprising at least one polar group selected from phosphine oxide group or diazole group, and (ii) in substantially elemental form, an electropositive element selected from substantially non-radioactive alkali metals, alkaline earth metals, rare earth metals, and transition metals of the fourth period of the Periodic table having proton numbers 22, 23, 24, 25, 26, 27, 28, 29.

Description

Organic light emitting device comprising polar matrix, metal dopant and silver cathode

The present invention concerns organic light emitting device with improved electrical properties, particularly the device comprising, in combination with a silver cathode, an improved electron transporting and/or electron injecting layer and or an OLED stack comprising the improved transporting and/or electron injecting layer and, optionally, an improved charge generating layer.

I. BACKGROUND OF THE INVENTION

Among the electronic devices comprising at least a part based on material provided by organic chemistry, organic light emitting diodes (OLEDs) have a prominent position. Since the demonstration of efficient OLEDs by Tang et al. in 1987 (C.W. Tang et al., Appl. Phys. Lett, 51 (12), 913 (1987)), OLEDs developed from promising candidates to high-end commercial displays. An OLED comprises a sequence of thin layers substantially made of organic materials. The layers typically have a thickness in the range of 1 nm to 5 μπι. The layers are usually formed either by means of vacuum deposition or from a solution, for example by means of spin coating or jet printing.

OLEDs emit light after the injection of charge carriers in the form of electrons from the cathode and in form of holes from the anode into organic layers arranged in between. The charge carrier injection is effected on the basis of an applied external voltage, the subsequent formation of excitons in a light emitting zone and the radiative recombination, of those excitons. At least one of the electrodes is transparent or semitransparent, in the majority of cases in the form of a transparent oxide, such as indium tin oxide (ITO), or a thin metal layer.

Among the matrix compounds used in OLED light emitting layers (LELs) or electron transporting layers (ETLs), important position have the compounds that comprise at least one polar group selected from phosphine oxide and diazole. The reason why such polar group often significantly improves the electron injecting and/or electron transporting properties of the semiconducting material is not yet fully understood. It is believed that the high dipole moment of the polar group plays somehow the positive role. Especially recommended for this use are triaryl phosphine oxides comprising at least one condensed aromatic or heteroaromatic group attached directly to the phosphine oxide group, see e.g. JP 4 876 333 B2. Among diazole groups, especially phenylbenzimidazole groups have been widely utilized for the design of Hew electron transport matrix compounds, for example TPBI described in US 5 645 948, and some compounds comprising benzimidazolyl structural moieties linked with other structural moieties comprising delocalized pi-electrons in two or more aromatic or heteroaromatic rings nowadays are considered to be industrial standard, e.g. compound LG- 201 (for example US 6 878 469)

Electrical doping of charge transporting semiconducting materials for improving their electrical properties, especially conductivity, is known since 1990s, e.g. from US 5 093 698 A. An especially simple method for n-doping in ETLs prepared by the thermal vacuum deposition, which is currently the standard method most frequently used, e.g. in industrial manufacture of displays, is evaporation of a matrix compound from one evaporation source and of a highly electropositive metal from another evaporation source and their co-deposition on a solid substrate. As useful n-dopants in triaryl phosphine oxide matrix compounds, alkali metals and alkaline earth metals were recommended in JP 4 725 056 B2, with caesium as the dopant successfully used in the given examples. Indeed, caesium as the most electropositive metal offers the broadest freedom in the choice of a matrix material, and it is likely the reason why solely caesium was the n-doping metal of choice in the cited document.

For an industrial use, caesium as a dopant has several serious drawbacks. First, it is very reactive, moisture and highly air sensitive material that renders any handling difficult and incurs significant additional costs for mitigating the high safety and fire hazard unavoidably linked with its use. Second, its quite low normal boiling point (678 °C) indicates that it may be highly volatile under high vacuum conditions. Indeed, at pressures below 10^"2 Pa used in industrial equipment for vacuum thermal evaporation (VTE), caesium metal evaporates significantly already at slightly elevated temperature. Taking into account that the evaporation temperatures for typical matrix compounds used in organic semiconducting materials at pressures below 10^" Pa are typically between 150-400 °C, avoiding an uncontrolled caesium evaporation, resulting in its undesired deposition contaminating the colder parts of the whole equipment (e.g. the parts that are shielded against heat radiation from the organic matrix evaporation source), is a really challenging task.

Several methods for overcoming these drawbacks and enabling industrial applicability of caesium for n-doping in organic electronic devices have been published. For safe handling, caesium may be supplied in hermetic shells that open just inside the evacuated evaporation source, preferably during heating to the operational temperature. Such technical solution was provided e.g. in WO 2007/065685, however, it does not solve the probiem of caesium high volatility.

US 7 507 694 B2 and EP 1 648 042 Bl offer another solution in form of caesium alloys that melt at low temperature and show significantly decreased caesium vapour pressure in comparison with the pure metal. Bismuth alloys of WO2007/109815 that release caesium vapours at pressures of the order 10⁴ Pa and temperatures up to about 450 °C represent another alternative. Yet, all these alloys are still highly air and moisture sensitive. Moreover, this solution has further drawback ie the fact that the vapour pressure over the alloy changes with the decreasing caesium concentration during the evaporation. That creates new problem of an appropriate deposition rate control, e.g. by programming the temperature of the evaporation source. So far, quality assurance (QA) concerns regarding robustness of such process on an industrial scale hamper a wider application of this technical solution in mass production processes.

A viable alternative to Cs doping represent highly electropositive transition metal complexes like W₂(hpp) that have ionisation potentials comparably low as caesium and volatilities comparable with volatilities of usual organic matrices. Indeed, these complexes disclosed as electrical dopants first in WO2005/086251 are very efficient for most electron transporting matrices except some hydrocarbon matrices. Despite their high air and moisture sensitivity, these metal complexes provide satisfactory n-doping solution for an industrial use, if supplied in the shells according to WO 2007/065685. Their main disadvantage is their high price caused by relative chemical complexity of comprised ligands and necessity of a multistep synthesis of the final complex, as well as additional costs incurred by necessity of using the protective shells and/or by the QA and logistic issues linked with shell recycling and refilling. Another alternative represent strong n-dopants created in situ in the doped matrix from relatively stable precursors by an additional energy supplied e.g. in form of ultraviolet (UV) or visible light of an appropriate wavelength. Appropriate compounds for this solution were provided e.g. in WO2007/107306 Al. Nevertheless, state-of-the-art industrial evaporation sources require materials with very high thermal stability, allowing their heating to the operational temperature of the evaporation source without any decomposition during the whole operating cycle (e.g., for a week at 300 °C) of the source loaded with the material to be evaporated. Providing organic n-dopants or n-dopant precursors with such long-term thermal stability is a real technical challenge so far. Moreover, the complicated arrangement of the production equipment that must ensure a defined and reproducible additional energy supply for achieving reproducibly the desired doping level (through the in situ activation of the dopant precursor deposited in the matrix) represents an additional technical challenge and a potential source of additional CA issues in mass production.

Yook et al (Advanced Functional Materials 2010, 20, 1797-1802) successfully used caesium azide in laboratory as an air-stable Cs precursor. This compound is known to decompose under heating above 300 °C to caesium metal and elemental nitrogen. This process is, however, hardly applicable in contemporary industrial VTE sources, due to difficult control of such heterogeneous decomposition reaction in a larger scale. Moreover, release of nitrogen gas as a by-product in this reaction brings a high risk that, especially at higher deposition rates desired in the mass production, the expanding gas will expel solid caesium azide particles from the evaporation source, causing thus high defect counts in the deposited layers of doped semiconducting materials.

Another alternative approach for electrical n-doping in electron transporting matrices is doping with metal salts or metal complexes. The most frequently used example of such dopant is lithium 8-hydroxy-quinolinolate (LiQ). It is especially advantageous in matrices comprising a phosphine oxide group, see e.g. WO 2012/173370 A2. The main disadvantage of metals salt dopants is that they improve basically only electron injection to the adjacent layers and do not increase the conductivity of doped layers. Their utilization for decreasing the operational voltage in electronic devices is thus limited on quite thin electron injecting or electron transporting layers and does hardly allow e.g. an optical cavity tuning by using ETLs thicker than approximately 25 nm, what is well possible with redox-doped ETLs having high conductivity. Furthermore, metal salts typically fail as electrical dopants in cases wherein creation of new charge carriers in the doped layer is crucial, e.g. in charge generating layers (CGL, called also p-n junctions) that are necessary for the function of tandem OIJEDs.

For the above reasons, and especially for electrical doping in ETLs thicker than approximately 30 run, the current technical practice prefers lithium as an industrial redox n- dopant (see e.g. US 6 013 384 B2). This metal is relatively cheap and differs from other alkali metals by its somewhat lower reactivity and, especially, by its significantly lower volatility (normal boiling point about 1340 °C), allowing its evaporation in the VTE equipment at temperatures between 350-550 °C.

Nevertheless, quite in accordance with its high n-doping power allowing Li to dope majority of usual types of electron transporting matrices, this metal possesses also a high degree of reactivity. It reacts under ambient temperature even with dry nitrogen and for its use in a highly reproducible manufacturing process complying with contemporary industrial QA standards, it must be stored and handled exclusively under high purity noble gases. Moreover, if Li is co-evaporated with matrix compounds that have evaporation temperatures in the range 150-300 °C, its significantly higher evaporation temperature in comparison with the matrix evaporation temperature already causes cross-contamination problems in the VTE equipment.

Many documents suggest as alternative n-dopants almost any known metallic element including weakly reductive and highly volatile Zn, Cd, Hg, weakly reductive Al, Ga, In, Tl, Bi, Sn, Pb, Fe, Co, Ni, or even noble metals like Ru, Eh, IT and/or refractory metals with highest known boiling points like Mo, W, Nb, Zr (see e.g. JP 2009/076508 or WO 2009/106068). Unfortunately, not only in these two documents cited here as examples but throughout the scientific and patent literature overall, there is in fact lack of any evidence that some of these suggestions have ever been experimentally tested.

To be more specific, even WO 2009/106068 that does not merely mention all imaginable dopants but really strives to claim all the named metalloid elements as n-dopants in organic electronic devices due their alleged applicability through a high-temperature decomposition of a gaseous precursor compound in a heated nozzle, does not bring any single numeric value documenting the physical parameters of allegedly prepared doped materials and/or technical performance of allegedly prepared devices.

On the other hand, US2005/0042548 published before the date of priority of WO 2009/106068 teaches in paragraph 0069 (see namely the last two lines of the left column and first three lines of the right column on page 7) that iron pentacarbonyl can be used for n- doping in organic ETMs if the compound is activated by UV radiation which splits off a carbon monoxide ligand. The coordinatively unsaturated iron compound then reacts with the matrix, what results in. the observed doping effects. In the light of this previous art showing that the metal carbonyls that were used in the alleged working example of WO 2009/106068 are known n-dopants in organic matrices if activated by supply of additional energy, it seems quite likely that if the applicants of WO2009/106068 really obtained with their jet of iron pentacarbonyl flowing through a ceramic nozzle electrically heated to a white glow (see the last paragraph of the German text on page 12 of the cited PCT application) any doping effect in the target bathocuprom layer, this effect was caused rather by the same coordinatively unsaturated iron carbonyl complex as produced by UV irradiation in US2OO5/ 042548, than with elemental iron as they suggest. This suspicion is further supported by the fourth paragraph on page 13 of the cited PCT application which teaches that the same result can be obtained with a cold nozzle, if the stream of iron pentacarbonyl is irradiated with an infrared laser having the wavelength fitting with the absorption frequency of the CO groups in the iron pentacarbonyl complex. Here, it is even more likely that the laser activation resulted not in naked metal atoms or clusters of metal atoms but in a reactive, coordinately unsaturated iron complex still bearing some carbonyl ligand s, analogously to the reactive complex formed by activation with the UV light.

Despite metals with strongly negative standard redox potentials like alkali earth metals or lanthanides are recited as alternative n-dopants besides alkali metals basically in each document dealing with redox n -doping, the record of the proven n-doping with any metal different from alkali metals is very scarce.

Magnesium is in comparison with alkaline metals much less reactive. It reacts even with liquid water at the ordinary temperature very slowly and in air it keeps its metallic luster and does not gain weight for months. It may be thus considered as practically air-stable. Moreover, it has low normal boiling point (about 1100 °C), very promising for its VTE processing in an optimum temperature range for co-evaporation with organic matrices.

On the other hand, the authors of the present application confirmed in a screening done with dozens of state-of-the-art ETMs that Mg does not possess a sufficient doping strength for common ETMs which are free of strongly polar groups like phosphine oxide group. The only favourable result has been achieved in OLEDs comprising thin electron injection layers consisting of a specific kind of triaryl phosphine oxide matrix (comprising a special fcris- pyridyl unit designed for chelating metals), doped with magnesium, as shown in EP 2 452 946 Al. Despite the structural specifity and very favourable (in terms of its LUMO level which is quite deep under the vacuum level in the absolute energy scale) dopability of the exemplary matrix tested with magnesium in EP 2 452 946 Al , the positive results achieved with this n- doped semiconducting material encouraged further research focused on n-doping with substantially air stable metals.

It is an object of the invention to overcome the drawbacks of the prior art and to provide organic light emitting diodes with better characteristics, especially with low voltage and, more specifically, OLEDs with low voltage and high efficiency, preferably utilizing substantially air stable metals as n-dopants.

II. SUMMARY OF THE INVENTION

The object is achieved by an electronic device comprising at least one light emitting layer between an anode and substantially silver cathode, the device further comprising between the cathode and the anode at least one mixed layer (which may also be referred to as substantially organic layer) comprising

(i) at least one substantially covalent electron transport matrix compound comprising at least one polar group selected from phosphine oxide group or diazole group, and

(ii) in substantially elemental form, an electropositive element selected from

substantially non-radioactive alkali metals, alkaline earth metals, rare earth metals, and transition metals of the fourth period of the Periodic table having proton numbers 22, 23, 24, 25, 26, 27, 28, 29.

It is to be understood that "substantially silver" means silver content at least 99 weight %. Preferably, the silver content in the cathode is at least 99.5 weight %, more preferably at least 99.8 weight %, even more preferably at least 99.9 weight %, most preferably at least 99.95 weight %.

If the polar group is the phosphine oxide group, the reduction potential of the electron transport matrix compound, if measured by cyclic voltammetry under the same conditions, has preferably the value which is more negative than the value obtained for tris(2- benzo[d]thiazol-2-yl)phenoxya]uminum, preferably more negative than for 9,9', 10,10'- tetraphenyl-2,2'-bianthracene or 2,9-di([l , 1 '-biphenyl] -4-yl)-4,7-diphenyl- 1,10- phenanthroline, more preferably more negative than for 2,4,7,9-tetraphenyl-l , 10- phenanthroline, even more preferably more negative than for 9, 10-di(naphthalen-2-yl)-2- phenylanthracene, even more preferably more negative than for 2,9-bis(2-methoxyphenyl)- 4,7-diphenyl-l ,10-phenanthroIine, even more preferably more negative than for 9,9'- spirobi[fluorene]-2,7-diylbis(diphenylphosphine oxide), even more preferably more negative than for 4,7-diphenyl -1,10-phenanthroline, even more preferably more negative than for 1 ,3,5-tris(l -phenyl- lH-benzimidazol-2-yl)benzene, most preferably more negative than for pyrene and still preferably more negative than the value obtained for

[1 ,1 '-binaphthalen]-2_>2'-diylbis(diphenylphosphine oxide).

If the polar group is the imidazole group, the reduction potential of the electron transport matrix compound, if measured by cyclic voltammetry under the same conditions, has preferably the value which is more negative than the value obtained for 4,7-diphenyl- 1,10- phenanthroline, preferably more negative than for (9-phenyl-9H-carbazole-2,7- diyl)bis(diphenylphosphine oxide), more preferably more negative than for (9,9-dihexyl- 9H-fluorene-2,7-diyl)bis(diphenylphosphine oxide), highly preferably more negative than for 1 ,3 ,5-tri s( 1 -phenyl- 1 H-benzimidazol-2-yl)benzene, even more preferably more negative than for 3-phenyl-3H-benzo[b]dinaphto[2, 1 -d: 1 2^,-f|phosphepine-3-oxide, most preferably more negative than for pyrene and still preferably more negative than for [Ι ,Γ- binaphthalenj-2^2'-diylbis(diphenylphosphine oxide).

It is to be understood that "substantially covalent" means compounds comprising elements bound together mostly by covalent bonds. Examples of substantiall covalent molecular structures may be organic compounds, organometallic compounds, metal complexes comprising polyatomic ligands, metal salts of organic acids. In this sense, the term "substantially organic layer" is to be understood as a layer comprising a substantially covalent electron transport matrix compound.

It is further understood that the term "substantially covalent compound" encompasses materials which can be processed by usual techniques and equipment for manufacturing organic electronic devices, like vacuum thermal evaporation or solution processing. It is clear that purely inorganic crystalline or glassy semiconducting materials like silicon or germanium which cannot be, due to their extreme high vaporization temperatures and insolubility in solvents, prepared in equipment for organic electronic devices, are not encompassed by the term "substantially covalent compound. Preferably, the phosphine oxide polar group is part of a substantially covalent structure comprising at least three carbon atoms directly attached to the phosphorus atom of the phosphine oxide group and having overall count of covalently bound atoms, preferably selected from C, H, B, Si, N, P, O, S, F, CI, Br and I, in the range 16-250, more preferably in the range 32-220, even more preferably in the range 48-190, most preferably in the range 64-

160.

Also preferably, the electron transport matrix compound comprises a conjugated system of at least ten delocalized electrons.

Examples of conjugated systems of delocalized electrons are systems of alternating pi- and sigma bonds. Optionally, one or more two-atom structural units having the pi-bond between its atoms can be replaced by an atom bearing at least one lone electron pair, typically by a divalent atom selected from O, S, Se, Te or by a trivalent atom selected from N, P, As, Sb, Bi. Preferably, the conjugated system of delocalized electrons comprises at least one aromatic ring adhering to the Hiickel rule.

More preferably, the substantially covalent electron transport matrix compound comprises at least two aromatic or heteroaromatic rings which are either linked by a covalent bond or condensed. Also preferably, the phosphine oxide polar group is selected from phosphine oxide substituted with three monovalent hydrocarbyl groups or one divalent hydrocarbylene group forming with the phosphorus atom a ring and one monovalent hydrocarbyl group, and the overall count of carbon atoms in the hydrocarbyl groups and the hydrocarbylene group is 8-80, preferably 14-72, more preferably 20-66, even more preferably 26-60, most preferably 32-54. In another preferred embodiment, there are two substantially covalent compounds comprised in the semiconducting material, the first compound comprising the polar group selected from the phosphine oxide group and the diazole group, wherein the first compound is free of a conjugated system of delocalized electrons, or does comprise a conjugated system of less than 10 delocalized electrons; and the second compound that does comprise a conjugated system of at least 10 delocalized electrons. More preferably, the second compound is free of a polar group selected from the phosphine oxide group and/or the diazole group.

Examples of compounds comprising a condensed aromatic skeleton comprising at least 10 delocalized electrons are e.g. naphthalene, anthracene, phenanthrene, pyrene, quinoline, indole or carbazole. The conjugated system of delocalized electrons may also consist of at least two directly attached aromatic rings, the simplest examples of such systems being biphenyl, bithienyl, phenylthiophene, phenylpyridine and like. On the other hand, it is known that a pentavalent phosphorus atom, for example the phosphorus atom in the phosphine oxide group, does not participate in conjugation in systems of delocalized electrons attached to the pentavalent P atom, and in this sense resembles sp hybridized carbon, for example the carbon atom in a methylene group. Most preferably, the conjugated system of at least 10 delocalized electrons comprised in the second compound is comprised in a Cn-Cso-arene or a Cg-Cso heteroarene structural moiety, wherein the overall count of carbon atoms comprises also possible substituents. It is in the spirit of the present invention that the count, topology and spatial arrangement of the substituents on the structural moiety comprising the conjugated system of delocalized electrons is not decisive for the function of the invention. Preferred heteroatoms in the heteroarene structural moiety are B, O, N and S. In the second compound, both the core atoms bearing the comprised system of delocalized electrons as well as the polyvalent atoms like C, Si, B, which preferably form the peripheral substituents attached to the core atoms, may be substituted with terminal atoms of elements which are typically monovalent in organic compounds and which are more preferably selected from H, F, CI, Br and I. The conjugated system of at least 10 delocalized electrons is preferably comprised in a C[4-C5o-arene or a Cg-Cso heteroarene structural moiety. Also preferably, the diazole polar group is an imidazole group.

In one of preferred embodiments, the electronic device further comprises a metal salt additive consisting of at least one metal cation and at least one anion. Preferably, the metal cation is Li⁺ or Mg²*. Also preferably, the metal salt additive is selected from metal complexes comprising a 5-, 6- or 7- membered ring that contains a nitrogen atom and an oxygen atom attached to the metal cation and from complexes having the structure according to formula

(Π)

Formula (II), wherein A¹ is a C₆-C₃o arylene or C2-C30 heteroarylene comprising at least one atom selected from O, S and N in an aromatic ring and each of A and A is independently selected from a Cg-Cso aryl and C2-C30 heteroaryl comprising at least one atom selected from O, S and N in an aromatic ring. Equally preferably, the anion is selected from the group consisting of phenolate substituted with a phosphine oxide group, 8-hydroxyquinolinolate and pyrazolylborate. The metal salt additive preferably works as a second electrical n-dopant, more preferably, it works synergistically with the metallic element present in the elemental form and works as the first electrical n-dopant.

It is preferred that the electropositive element is selected from Li, Na, K, Be, Mg, Ca, Sr, Ba, Sc, Y, La, Sm, En, Tb, Yb, Lu, Τί, V and Mn, more preferably, from Li, Mg, Ca, Sr, Ba, Sm, Eu, Tb, Yb, most preferably from Li, Mg and Yb. Preferably, the molar ratio of the electropositive element to the first compound is lower than 0.5, preferably lower than 0.4, more preferably lower than 0.33, even more preferably lower than 0.25, even more preferably lower than 0.20, even more preferably lower than 0.17, most preferably lower than 0.15, still preferably lower than 0.13, still but less preferably lower than 0.10.

It is further preferred that the molar ratio of the electropositive element to the first compound is higher than 0.01, preferably higher than 0.02, more preferably higher than 0.03, even more preferably higher than 0.05, most preferably higher than 0.08.

Preferably, the electropositive element is comprised in an electron transporting, electron injecting, or charge generating layer. More preferably, the electron transporting or electron injecting layer is adjacent to a layer consisting of compounds that have their reduction potentials, if measured by cyclic voltammetry under the same conditions, more negative than the electron transport matrix compounds of the adjacent electron transporting or electron injecting layer. In one of preferred embodiments, the layer adjacent to the layer made of inventive semiconducting material is the emitting layer.

It is further preferred that the light emitting layer emits blue or white light. In one of preferred embodiments, the light emitting layer comprises at least one polymer. More preferably, the polymer is a blue light emitting polymer.

Also preferably, the electron transporting or electron injecting layer is thicker than 5 ran, preferably thicker than 10 nm, more preferably thicker than 15 ran, even more preferably thicker than 20 nm, most preferably thicker than 25 nm, still preferably thicker than 50 nm, and still preferably thicker than 100 nm.

In one of preferred embodiments, the electron transporting or electron injecting layer is adjacent to the cathode. Still another embodiment of the invention is a tandem OLED stack comprising a metal-doped pn-junction comprising

(i) in substantially elemental form, an electropositive element selected from substantially non-radioactive alkali metals, alkaline earth metals, rare earth metals, and transition metals of the fourth period of the Periodic table having proton numbers 22, 23, 24, 25, 26, 27, 28, 29, and

(ii) at least one substantially covalent electron transport matrix compound comprising at least one polar group selected from the phosphine oxide group and from the diazole group. The device of the invention is preferably prepared by process comprising, at least one step wherein

(i) an electropositive element selected from substantially non-radioactive alkali metals, alkaline earth metals, rare earth metals, and transition metals of the fourth period of the Periodic table having proton numbers 22, 23, 24, 25, 26, 27, 28, 29, and

(ii) at least one substantially covalent electron transport matrix compound

comprising at least one polar group selected from the phosphine oxide group and from the diazole group are coevaporated and codeposited under reduced pressure to form the substantially organic layer, and the electropositive element is deposited in its elemental or substantially elemental form. Preferably, the electropositive element is evaporated from its elemental or substantially elemental form, more preferably from a substantially air stable elemental or substantially elemental form. Also preferably, the pressure is lower than 10^" Pa, more preferably lower than 10 ³ Pa, most preferably lower than 10^'4 Pa.

Preferably, the electropositive element has normal boiling point lower than 3000 °C, more preferably lower than 2200 °C, even more preferably lower than 1800 °C, most preferably lower than 1500°C. Under normal boiling point, it is to be understood the boiling point at normal atmospheric pressure (101.325 kPa).

It is to be understood that the term "substantially air stable" refers to metals and their substantially elemental form (e.g. alloys with other metals) which reacts under ambient conditions with atmospheric gases and moisture slowly enough to avoid quality assurance issues if it is handled under ambient conditions in an industrial process. More specifically, a form of a metal shall be assigned as substantially air stable for the purposes of this application, if a sample of this form having the weight at least 1 g and a surface exposed to air at least 1 cm² may be kept under standard temperature 25 °C, pressure 101 325 Pa and relative humidity 80 % for at least an hour, preferably for at least 4 hours, more preferably for at least

24 hours, and most preferably for at least 240 hours, without showing a statistically significant weight gain, provided that the accuracy of weighing is at least 0.1 mg.

Most preferably, the electropositive element is evaporated from a linear evaporation source. In one of preferred embodiments, the inventive electronic device is an. inverted device.

In another preferred embodiment, the device is a top-emitting device with the cathode having thickness less than 30 nm, preferably less than 25 nm, more preferably less than 20 nm, most preferably less than 15 nm.

In one of preferred embodiments, the silver cathode is corrugated and/or combined with a light-scattering layer.

III. BRIEF DESCRIPTION OF DRAWINGS

FIG. 1 shows a schematic illustration, of a device in which the present invention can be incorporated.

FIG. 2 shows a schematic illustration of a device in which the present invention can be incorporated.

FIG. 3 shows absorbancc curves of two n-doped semiconducting materials;

circles stand for comparative matrix compound CIO doped with 10 wt% of compound Fl that forms strongly reducing radicals, triangles stand for compound E10 doped with 5 wt% Mg.

IV. DETAILED DESCRIPTION OF THE INVENTION

Device Architecture

Fig. 1 shows a stack of anode (10), organic semiconducting layer (11) comprising the light emitting layer, electron transporting layer (ETL) (12), and cathode (13). Other layers can be inserted between those depicted, as explained herein. Fig. 2 shows a stack of an anode (20), a hole injecting and transporting layer (21), a hole transporting layer (22) which can also aggregate the function of electron blocking, a light emitting layer (23), an ETL (24), and a cathode (25). Other layers can be inserted between those depicted, as explained herein.

The wording "device" comprises the organic light emitting diode. Material properties - energy levels

A method to determine the ionization potentials (IP) is the ultraviolet photo spectroscopy (UPS). It is usual to measure the ionization potential for solid state materials; however, it is also possible to measure the IP in the gas phase. Both values are differentiated by their solid state effects, which are, for example the polarization energy of the holes that are created during the photo ionization process. A typical value for the polarization energy is approximately 1 eV, but larger discrepancies of the values can also occur. The IP is related to onset of the photoemission spectra in the region of the large kinetic energy of the photoelectrons, i.e. the energy of the most weakly bounded electrons. A related method to UPS, the inverted photo electron spectroscopy (IPES) can be used to determine the electron affinity (EA). However, this method is less common. Electrochemical measurements in solution are an alternative to the determination of solid state oxidation (E_ox) and reduction (E,ed) potential. An adequate method is, for example, cyclic voltammetry. To avoid confusion, the claimed energy levels are defined in terms of comparison with reference compounds having well defined redox potentials in cyclic voltammetry, when measured by a standardized procedure. A simple rule is very often used for the conversion of redox potentials into electron affinities and ionization potential: IP (in eV) = 4.8 eV + e*E_0X (wherein E_ox is given in volts vs. ferrocenium/ferrocenc (FCVFC)) and EA (in eV) = 4.8 eV + e*E_ieij (E_red is given in volts vs. FcVFc) respectively (see B.W. D'Andrade, Org. Electron. 6, 11-20 (2005)), e* is the elemental charge. Conversion factors for recalculation of the electrochemical potentials in the case other reference electrodes or other reference redox pairs are known (see A.J. Bard, L.R. Faulkner,„Electrochemical Methods: Fundamentals and Applications", Wiley, 2. Ausgabe 2000). The information about the influence of the solution used can be found in N.G. Connelly et al., Chem. Rev. 96, 877 (1996). It is usual, even if not exactly correct, to use the terms„energy of the HOMO" E(HOMO) arid "energy of the LUMO" E_(LUMO). respectively, as synonyms for the ionization energy and electron affinity (Koopmans Theorem). It has to be taken into consideration that the ionization potentials and the electron affinities are usually reported in such a way that a larger value represents a stronger binding of a released or of an absorbed electron, respectively. The energy scale of the frontier molecular orbitals (HOMO, LUMO) is opposed to this. Therefore, in a rough approximation, the following equations are valid: IP = -E(HOMO) and EA = E(LU O) (the zero energy is assigned to the vacuum).

For the chosen reference compounds, the inventors obtained following values of the reduction potential by standardized cyclic voltammetry in tetrahydroruran (THF) solution vs. FCVFC:

tris(2-benzo[d]thiazol-2-yl)phenoxyaluminum, CAS 1269508-

14-6, -2.21 V, BO;

9,9', 10, 10'-tetraphenyl-2,2'-bianthracene (TPBA), CAS 172285-72-

2, -2.28 V, B 1 ;

2,9-di([l , 1 '-biphenyl]-4-yl)-4,7-diphenyl-l ,10-phenanthroline,

CAS 338734-83-1, -2.29 V, B2;

2,4,7,9-tetraphenyl- 1 ,10-phenanthroline, CAS 51786-73-3, -2.33 V, B3;

9, 10-di(naphthalen-2-yl)-2-phenylanthracene (PADN), CAS 865435-

20-7, -2.37 V, B4;

,10-phenanlJiroline, CAS

, -2.40 V, B5;

(SPP013), CAS 1234510-13-4, -2.41 V, B6;

spirobi[nuorene]-2,2',7,7'-tetraamine (Spiro TAD), CAS 189363-47-1, -3.10 V, B7

biphenyl]-4,4'-diamine (alpha- PD), CAS 123847-85-8, -2.96 V, B9; 4,4'-di(9H-carbazol-9-yI)-1 J '-biphenyl (CBP), CAS 58328-31-

7, -2.91 V, BIO;

oxide (BCPO),

CAS 1233407-28-7, -2.86, Bl 1;

3-([l,l'-biphenyl]-4-yl)-5-(4-(tert-butyl)phenyl 4-phenyl-4H- 1,2,4-triazole (TAZ), -2.76 V, B12;

CAS 129-00-0, -2.64 V, B13.

Examples of matrix compounds for state-of-the-artelectrically doped semiconducting materials based on matrix compounds comprising phosphinc oxide group and a conjugated system of at least 10 delocalized electrons are

oxide) (PP027), CAS 1299463-56-1, -2.51 V, El ;

[1,1 -binaphthalen] -2,2'-diylbis(diphenylphosphine oxide) (BINAPO), CAS 86632-33-9, -2.69 V, E2;

spiro [dibenzo[cJh]xanthene-7,9'-fluorene] -2',7- diylbis(diphenylphosphine oxide), -2.36 V, E3;

naphtalene-2,6-diylbis(diphenylphosphine oxide)

fl j': 4Μ "-ί^βφΙι^βηνΙ]-3.5-ίΙϊν]Κ8(ίΙί Η^βη 'ί 1ΐ08 Ηΐπβ oxide). -2.51 V,

E5;

3 -phenyl-3H-benzo [b]dinaphto[2, 1 -d: 1 2'-f]phosphepine-3-oxide, CAS 597578-38-6, -Z62 V. E6;

diphenyl(4-{9-phenyl-9H-carba2ol-3 -yl)phenylphosphine oxide,

-2.S1 ¥. E7:

(9,9-dihexyl-9H-fluorene-2,7-diyl)bis(diphenylphosphine oxide), -2.52 V, E8;

(3 -(3,11 -dimethoxydibenzo[c_sh]acridin-7- yl)phenyl)diphenylphosphine oxide (described in WO2013/079217 Al), -2.29 V» E9;

(3 -(2, 12-dimethoxydibenzo [c,h] acridin-7- yl)phenyl)diphenylphosphine oxide (described in WO2013/079217 Al), -2.24 V, E10;

diphenyl(5 -(pyren- 1 -yl)pyridine-2-yl)phosphinc oxide, described in WO2014/ 167020, -2.34 V, E1 1 ;

diphenyl(4-(pyren-l-yl)phenyl)phosphine oxide, described in PCT/EP2014/071659, -2.43 V, El 2.

(3-(anthracen-9-yl)phenyl)diphenylphosphine oxide,

E13.

diphenyl-(3 -(spiro [fluorene-9,9'-xanthen] -2- yl)phenyl)phosphine oxide, -2.71 V. E14

diphenyl-(4-(spiro[fluorene-9,9^,-xanthen]-2- yl)phenyl)phosphine oxide, -2,65 V, El 5. As comparative compounds were used

(4-(di benzo [c,h] acridin-7-yl)phenyl)diphenylphosphine

(described in WO2011/154131 Al), -2.20 V, CI;

(6,6'-(l-(pyridin-2-yl)ethane- 1,1- diyl)bis(pyridine-6_r2-diyl))bis(diphenylphosphinc oxide), described in EP 2 452 946, -2.21 V, C2;

2-(4-(9, 10-di(naphthalen-2-yl)anthracen-2- yl)phenyl)-l-phenyl-l H-benzold]imidazolc (LG-201), CAS 561064-11-7, -2.32 V, C3;

7-(4'-(l -phenyl- 1 H-benzo[d]imidazol-2-yl-[ 1 , '-biphenyl J-4- yl)dibenzo[c,h]acridine (described in WO2011/154131 Al), -2.24 V, C4;

7-(4'-(l -phenyl-1 H-benzo[d]imidazol-2- yl)phenyl)dibenzo[c,h]acridine (described in WQ2011/154131 Al), -2.22 V, C5;

1 ,3 ,5-tris( 1 -phenyl- i H-benzimidazol-2-yl)benzene (TPBI) CAS 192198-85-9, -2.58 V, C6;

4,7-diphenyl- 1 , 10-phenanthroline (Bphen) CAS 1662- 01-7, -2.47 V, C7;

1 ,3 -bis [2-{2 ,2'-bipyridine-6-yl)-

1 ,3,4-oxadiazol-5-ylJbenzene (Bpy-OXD), -2.28 V, C8;

(9, 10-di(naphthalen-2-yl)anthracen-2-yl)diphenylphosphine oxide, CAS 1416242-45-9, -2.19 V, C9;

4-(naphtalen- 1 -yl)-2,7,9-triphenylpyrido[3,2- h]quinazoline, according to EP 1 971 371, -2.18 V, CIO;

1 ,3 ~bis(9-phenyl- 1 , 10-phenanthrolin-2-yl)benzene, CAS

721969-84-4, -2.45 V, C11.

Substrate

It can be flexible or rigid, transparent, opaque, reflective, or translucent. The substrate should be transparent or translucent if the light generated by the OLED is to be transmitted through the substrate (bottom emitting). The substrate may be opaque if the light generated by the OLED is to be emitted in the direction opposite of the substrate, the so called top-emitting type. The OLED can also be transparent. The substrate can be either arranged adjacent to the cathode or anode.

Electrodes The electrodes are the anode and the cathode, they must provide a certain amount of conductivity, being preferentially conductors. Preferentially the "first electrode" is the cathode. At least one of the electrodes must be semi-transparent or transparent to enable the light transmission to the outside of the device. Typical electrodes are layers or a stack of layer, comprising metal and/or transparent conductive oxide. Other possible electrodes are made of thin busbars (e.g. a thin metal grid) wherein the space between the busbars is filled (coated) with a transparent material having certain, conductivity, such as graphene, carbon nanotubes, doped organic semiconductors, etc. hi one embodiment, the anode is the electrode closest to the substrate, which is called non- inverted structure. In another mode, the cathode is the electrode closest to the substrate, which is called inverted structure.

Typical materials for the Anode are ITO and Ag.

The inventors found that especially advantageous is the use of cathodes prepared substantially of metallic silver, because neat silver provides the best reflectivity, and thus best efficiency, specifically e.g. in bottom emitting devices built on a transparent substrate and having an a transparent conductive oxide anode. Alternatively, for top emitting devices having transparent cathodes, electropositive metals like aluminium are basically inapplicable due to high reflectivity/absorptivity even in very thin layers, and other metals with high work function which can form transparent thin layers, like gold, also afford ver poor electron injection. Therefore, substantially silver cathodes could be so far hardly used in air stable devices, e.g. in devices having undoped ETLs or ETLs doped with metal salt additives, because such devices showed high operational voltages and low efficiencies due to poor electron injection from neat silver.

It is equally well possible that the cathode is pre-formed on a substrate (then the device is an inverted device), or the cathode in a non-inverted device is formed by vacuum deposition of a metal or by sputtering. Thin silver cathodes having thickness less than 30 nm, preferably less than 25 nm, more preferably less than 20 nm, most preferably less than 15 nm are transparent and therefore advantageously used in top emitting devices. One embodiment of the invention is a transparent device comprising the transparent silver cathode in a combination with a transparent anode. Both thick (light reflecting) as well as thin (translucent or transparent) cathodes can be advantageously combined with a light-scattering layer and/or be corrugated, as described in WO2011/115738 or in WO2013/083712.

Hole-Transporting Layer (HTL)

The HTL is a layer comprising a large gap semiconductor responsible to transport holes from the anode or holes from a CGL to the light emitting layer (LEL). The HTL is comprised between the anode and the LEL or between the hole generating side of a CGL and the LEL. The HTL can be mixed with another material, for example a p-dopant, in which case it is said the HTL is p-doped. The HTL can be comprised by several layers, which can have different compositions. P-doping of the HTL lowers its resistivity and avoids the respective power loss due to the otherwise high resistivity of the undoped semiconductor. The doped HTL can also be used as optical spacer, because it can be made very thick, up to 1000 nm or more without significant increase in resistivity.

Suitable hole transport matrices (HTM) can be, for instance compounds from the diamine class, where a delocalized pi-electron system conjugated with lone electron pairs on the nitrogen atoms is provided at least between the two nitrogen atoms of the diamine molecule. Examples are N4,N4'-di(naphthalen- 1 -yl)-N4,N4'-diphenyl- [1,1 '-biphenyl] -4,4 '-diamine (HTM1), N4,N4₅ N4^»',N4"-tetra([l , 1 ^*-biphenyl]-4-yl)-[ 1 , 1 ':4',1 "-terphenyl]-4,4"-diamine (HTM2). The synthesis of diamines is well described in literature; many diamine HTMs are readily commercially available.

Hole-Injecting Layer (HTL)

The H1L is a layer which facilitates the injection of holes from the anode or from the hole generating side of a CGL into an adjacent HTL. Typically, the HIL is a very thin layer (<10 nm). The hole injection layer can be a pure layer of p-dopant and can be about 1 nm thick. When the HTL is doped, an HIL may not be necessary, since the injection function is already- provided by the HTL.

Light-Emitting Layer (LEL)

The light emitting layer must comprise at least one emission material and can optionally comprise additional layers. If the LEL comprises a mixture of two or more materials the charge carrier injection can occur in different materials for instance in a material which is not the emitter, or the charge carrier injection can also occur directly into the emitter. Many different energy transfer processes can occur inside the LEL or adjacent LELs leading to different types of emission. For instance excitons can be formed in a host material and then be transferred as singlet or triplet excitons to an emitter material which can be singlet or triplet emitter which then emits light. A mixture of different types of emitter can be provided for higher efficiency. White light can be realized by using emission from an emitter host and an emitter dopant. In one of preferred embodiments of the invention, the light emitting layer comprises at least one polymer.

Blocking layers can be used to improve the confinement of charge carriers in the LEL, these blocking layers are further explained in US 7,074,500 B2.

Electron-Transporting Layer (ETL)

The ETL is a layer comprising a large gap semiconductor responsible for electron transport from the cathode or electrons from a CGL or EIL (see below) to the LEL. The ETL is comprised between the cathode and the LEL or between the electron generating side of a CGL and the LEL. The ETL can be mixed with an electrical n-dopant, in which case it is said the ETL is n-dopcd. The ETL can be comprised by several layers, which can have different compositions. Electrical n-doping the ETL lowers its resistivity and/or improves its ability to inject electrons into an adjacent layer and avoids the respective power loss due to the otherwise high resistivity (and/or bad injection ability) of the undoped semiconductor. If the used electrical doping creates new charge carriers in the extent that substantially increases conductivity of the doped semiconducting material in comparison with the undoped ETM, then the doped ETL can also be used as optical spacer, because it can be made very thick, up to 1000 nm or more without significant increase in the operational voltage of the device comprising such doped ETL. The preferable mode of electrical doping that is supposed to create new charge carriers is so called redox doping. In case of n-doping, the redox doping corresponds to the transfer of an electron from the dopant to a matrix molecule.

In case of electrical n-doping with metals used as dopants in their substantially elemental form, it is supposed that the electron transfer from the metal atom to the matrix molecule results in a metal cation and an anion radical of the matrix molecule. Hopping of the single electron from the anion radical to an adjacent neutral matrix molecule is the currently supposed mechanism of charge transport in redox n-doped semiconductors. It is, however, hard to understand all properties of semiconductors n-doped with metals and, specifically, of semiconducting materials of present invention, in terms of electrical redox doping. It is therefore supposed that semiconducting materials of present invention advantageously combine redox doping with yet unknown favourable effects of mixing ETMs with metal atoms and/or their clusters. It is supposed that semiconducting materials of present invention contain a significant part of the added electropositive element in its substantially elemental form. The term "substantially elemental" shall be understood as a form that is, in terms of electronic states and their energies, closer to the state of a free atom or to the state of a cluster of metal atoms than to the state of a metal cation or to the state of a positively charged cluster of metal atoms.

Without being limited by theory, it can be supposed that there is an important difference between the n-doped organic semiconducting materials of previous art and the n-doped semiconducting materials of the present invention. In common organic ETMs of previous art (having reduction potentials roughly in the range between -2.0 and -3.0 V vs. Fc⁺ Fc and comprising a conjugated system of at least ten delocalized electrons), the strong redox n- dopants like alkali metals or W₂(hpp)4 are supposed to create the amounts of charge carriers that are commensurate to the number of individual atoms or molecules of the added dopant, and there is indeed an experience that increasing the amount of such strong dopant in conventional matrices above certain level does not bring any substantial gain in electrical properties of the doped material.

On the other hand, it is difficult to speculate which role shall the n-doping strength of the electropositive element play in the matrices of the present invention comprising primarily the polar groups but only very small or none conjugated systems of delocalized electrons.

It may be perhaps be supposed that in such matrices, even for the strongest electropositive element like alkali metals, only part of added atoms of the electropositive element added as the n-dopant reacts with matrix molecules by the redox mechanism under formation corresponding metal cations. It is rather supposed that even in high dilution, when the amount of the matrix is substantially higher than the amount of added metallic element, a substantial part of the metallic element is present in a substantially elemental form. It is further supposed that if the metallic element of the present invention is mixed with matrix of the present invention in a comparable amount, the majority of the added metallic element is present in the resulting doped semiconducting material in the substantially elemental form. This hypothesis seems to provide a reasonable explanation as to why the metallic elements of the present invention can be effectively used in significantly broader range of ratios to the doped matrix than the stronger dopants of previous art, even though they are weaker dopants. The applicable content of the metallic element in the doped semiconducting material of the present invention is roughly in the range from 0.5 weight % up to 25 weight %, preferably in the range from 1 to 20 weight %, more preferably in the range from 2 to 15 weight %, most preferably in the range from 3 to 10 weight %.

Hole blocking layers and electron blocking layers can be employed as usual.

Other layers with different functions can be included, and the device architecture can be adapted as known by the skilled in the art. For example, an Electron-Injecting Layer (EEL) made of metal., metal complex or metal salt can be used between the cathode and the ETL.

Charge generation layer (CGL)

The OLED can comprise a CGL which can be used in conjunction with an electrode as inversion contact, or as connecting unit in stacked OLEDs. A CGL can have the most different configurations and names, examples are pn-j unction, connecting unit, tunnel junction, etc. Best examples are pn-j unctions as disclosed in US 2009/0045728 Al, US 2010/0288362 Al . Metal layers and or insulating layers can also be used.

Stacked OLEDs

When the OLED comprises two or more LELs separated by CGLs, the OLED is called a stacked OLED, otherwise it is called a single unit OLED. The group of layers between two closest CGLs or between one of the electrodes and the closest CGL is called a electroluminescent unit (ELU). Therefore, a stacked OLED can be described as anode/ELU i / { CGLx/^'ELU i +χ } χ/cathode, wherein x is a positive integer and each CGLx or each ELUi+x can be equal or different. The CGL can also be formed by the adjacent layers of two ELUs as disclosed in US2009/0009072 Al. Further stacked OLEDs are described e.g. in US 2009/0045728 Al , US 2010/0288362 Al, and references therein.

Deposition of Organic Layers

Any organic semiconducting layers of the inventive display can be deposited by known techniques, such as vacuum thermal evaporation (VIE), organic vapour phase deposition, laser induced thermal transfer, spin coating, blade coating, slot dye coating, inkjet printing, etc. A preferred method for preparing the OLED according to the invention is vacuum thermal evaporation. Polymeric materials are preferably processed by coating techniques from solutions in appropriate solvents.

Preferably, the ETL is formed by evaporation. When using an additional material in the ETL, it is preferred that the ETL is formed by co-evaporation of the electron transporting matrix (ETM) and the additional material. The additional materia! may be mixed homogeneously in the ETL. In one mode of the invention, the additional material has a concentration variation in the ETL, wherein the conceatration changes in the direction of the thickness of the stack of layers. It is also foreseen that the ETL is structured in sub-layers, wherein some but not all of these sub-layers comprise the additional material.

In is supposed that semiconductiHg material of present invention contains a significant part of the added electropositive element in its substantially elemental form. Consequently, the process of the present invention requires that the electropositive element is vaporized from its elemental or substantially elemental form. In this context, the term "substantially elemental" shall be understood as a form that is, in terms of electronic states and their energies and in terms of chemical bonds, closer to the form of an elemental metal, of a free metal atom or to the form, of a cluster of metal atoms, than to tie form of a metal salt, of a covaleet metal compound, or to the form of a coordination compound of a metal. Typically, metal vapour release from metal alloys according to EP 1 648 042 Bl or WO2007/109815 is understood as the evaporation from a substantially elemental for of the evaporated metal.

Electrical doping

The most reliable and, at the same time, efficient OLEDs are OLEDs comprising electrically doped layers. Generally, the electrical doping means improving of electrical properties, especially the conductivity and/or injection ability of a doped layer in comparison with neat charge-transporting matrix without a dopant. In the narrower sense, which is usually called redox doping or charge transfer doping, hole transport layers are doped with a suitable acceptor material (p-doping) or electron transport layers with a donor material (n-doping), respectively. Through redox doping, the density of charge carriers in organic solids (and therefore the conductivity) can be increased substantially. In other words, the redox doping increases the density of charge carriers of a semiconducting matrix in comparison with the charge carrier density of the undoped matrix. The use of doped charge-carrier transport layers (p-doping of the hole transport layer by admixture of acceptor-like molecules, n-doping of the electron transport layer by admixture of donor-like molecules) in organic light-emitting diodes is, e.g., described in US 2008/203406 and US 5,093,698.

US2008227979 discloses in detail the charge-transfer doping of organic transport materials, with inorganic and with organic dopants. Basically, an effective electron transfer occurs from the dopant to the matrix increasing the Fermi level of the matrix. For an efficient transfer in a p-doping case, the LUMO energy level of the dopant is preferably more negative than the HOMO energy level of the matrix or at least not more than slightly more positive, preferably not more than 0.5 eV more positive than the HOMO energy level of the matrix. For the n- doping case, the HOMO energy level of the dopant is preferably more positive than the LUMO energy level of the matrix or at least not more than slightly more negative, preferably not more than 0.5 eV lower compared to the LUMO energy level of the matrix. It is furthermore desired that the energy level difference for energy transfer from dopant to matrix is smaller than + 0.3 eV.

Typical examples of known redox doped hole transport materials are: copper phthalocyanine (CuPc), which HOMO level is approximately -5.2 eV, doped with tetrafluoro- tetracyanoquinonedimethane (F4TC Q), which LUMO level is about -5.2 eV; zinc phthalocyanine (ZnPc) (HOMO = -5.2 eV) doped with F4TCNQ; a-NPD (Ν,Ν'- Bis(naphthalen- 1 -ylj-N^-bisCphenyl^benzidine) doped with F4TCNQ. a-NPD doped with 2,2'-(perfluoronaphthalene-2,6-diylidene) dimalononitrile (PD1). a-NPD doped with 2,2',2"- (cyclopropane-l,2,3-triylidene)tris(2-(p-cyanotetrafluorophenyl)acetorntrile) (PD2). All p- doping in the device examples of the present application was done with 3 mol% of PD2.

Typical examples of known redox doped electron transport materials are: fullerene C60 doped with acridine orange base (AOB); perylene-3 ,4,9, 1 O-tetracarboxylic-3,4,9, 10-dianhydride (PTCDA) doped with leuco crystal violet; 2,9 - di (phenanthren-9-yl) - 4,7 - diphenyl - 1,10 - phenanthroline doped with tetrakis ( 1 ,3,4,6,7,8— hexahydro - 2H - pyrimido [ 1,2 - a] pyrimidinato) ditungsten (II) (W2(fapp)₄); naphthalene tetracarbox lic acid di-anhydride (NTCDA) doped with 3,6-bis-(dimethyl amino)-acridme; NTCDA doped with bis(ethylene- dithio) tetrathiafulvalene (BEDT-TTF).

Besides the redox dopants, certain metal salts can be alternatively used for electrical n-doping resulting in lowering operational voltage in devices comprising the doped layers in

comparison with the same device without metal salt. True mechanis how these metal salts, sometimes called "electrically doping additives", contribute to the lowering of the voltage in electronic devices, is not yet known. It is believed that they change potential barriers on the interfaces between adjacent layers rather than conductivities of the doped layers, because their positive effect on operational voltages is achieved only if layers doped with these additives are very thin. Usually, the electrically undoped or additive doped layers are thinner than 50 nm, preferably thinner than 40 run, more preferably thinner than 30 nm, even more preferably thinner than 20 nm, most preferably thinner than 15 nm. If the manufacturing process is precise enough, the additive doped layers can be advantageously made thinner than 10 nm or even thinner than 5 nm.

Typical representatives of metal salts which are effective as second electrical dopants in the present invention are salts comprising metal cations bearing one or two elementary charges. Favourably, salts of alkali metals or alkaline earth metals are used. The anion of the salt is preferably an anion providing the salt with sufficient volatility, allowing its deposition under high vacuum conditions, especially in the temperature and pressure range which is

comparable with the temperature and pressure range suitable for the deposition of the electron transporting matrix.

Example of such anion is 8-hy drox yquinolinolate anion. Its metal salts, for example lithium 8- hydroxyquinolinolate (LiQ) represented by the formula Dl

are well known as electrically doping additives.

Another class of metal salts useful as electrical dopants in electron transporting matrices of the present invention represent compounds disclosed in the application PCT/EP2012/074127 (WO2013/079678), having general formula (II)

wherein A is a C6-C20 arylene and each of A -A is independently selected from a Q,-C2o aryl, wherein the aryl or arylene may be unsubstituted or substituted with groups comprising C arid H or with a further LiO group, provided that the given C count in an aryl or arylene group includes also all substitucnts present on the said group. It is to be understood that the term substituted or unsubstituted arylene stands for a divalent radical derived from substituted or unsubstituted arene, wherein the both adjacent structural moieties (in formula (I), the OLi group and the diary! prosphine oxide group) are attached directly to an aromatic ring of the arylene group. In examples of the present application, this class of dopants is represented by compound D2

wherein Ph is phenyl

Yet another class of metal salts useful as electrical dopants in electron, transporting matrices of the present invention represent compounds disclosed in the application PCT/EP2012/074125

, having general formula (III)

" Formula (III),

wherein M is a metal ion, each of A⁴- A⁷ is independently selected from H, substituted or unsubstituted Cg-Cio aryl and substituted or unsubstituted C2-C2₀ heteroaryl and n is valence of the metal ion. In examples of the present application, this class of dopants is represented by compound D3

V. ADVANTAGEOUS EFFECT OF THE INVENTIO

The favourable effects of the inventive electrically doped semiconducting materials are shown in comparison with comparative devices comprising instead of the inventive combination of electron transporting matrices and dopants other combinations of matrices and dopants known in the art. The used devices are described in detail in examples. la the first screening phase, there were in device of example! tested 32 matrix compounds with 5 wt% Mg as dopant. Electron transport matrices comprising phosphine oxide matrices and having their LUMO level expressed in terms of their reduction potential vs. Fc⁺ Fc (measured by cyclic voltammetry in THF) higher than compound BO (-2.21 V under standardized conditions used) performed better than CI and C2, in terms of operational voltage and/or quantum efficiency of the device, and sigmficantly better than matrices lacking the phosphine oxide group, irrespective of their LUMO level. These observations were confirmed also for several other divalent metals, namely Ca, Sr, Ba, Sm and Yb.

The results are summarized in Table 1, wherein relative change of voltage and efficiency (both measured at current density 10 mA/cm²) is calculated against the C2/Mg system of previous art take as the reference. The overall score is calculated by subtraction of relative voltage change from relative change of efficiency.

Table 1

ETL ETL wt% U E CEQB-EQEr_f)/ BQEjt_f score

.matrix dopant dopant (V) (%) (%) (%)

El Li 0.5 3.255 -39 5.340 -5 +34

El Mg 5 3.2 -40 5.1.5 -9 +31

El Ca 1 3.497 -35 5.415 -4 +31

El Ca 5 3.633 -32 5.235 -7 +25

El Ba 1 3.577 -33 6.090 +41

El Ba 5 3.491 -35 5.560 -1 +34

E2 0.5 3.264 -39 5.345 -5 +34

E2 1 5 3,56 -34 5.33 -5 +29

E2 1 3.245 -39 5.750 +41.

E2 Ca 5 3..83 -29 5.83 33

E2 Ba 1 6.104 +13 6.245 +14 +1

Ba 5 3.293 -39 6.055 +47

Mg ~ 5 1 3.68 ~3 i 8 +14 14 Mg 5 -33 3.9

E4 2 0 -35 5.900 +5 +40

Ba 2 4.020 -25 6.150 +9 +34

E4 1 2 3.806 -29 )0 0 +29

Yb 2 3.844 -28 5,390 -4 +24

E5 Mg 5 3.29 -29 5 -3 +26

E6 Mg 5 3.53 7.73 +38 +72

E6 Ca 2 3.350 -37 5.650 +37

E6 Ba 2 3.710 ^r -31 6.320 + 12 +43 i Sm 2 3.429 6.040 +7 +43

E6 Yb 2 3.427 -36 5.965 +6 -w

E7 Mg 5 5.2 -3 6.48 + 15 +18

ES Mg 5 3.36 -37 5.6 0 +37

ES Ca 2 3.26 -39 5.24 -7 +32

ES Ba 2 3.329 -38 5.990 +6 +44

E9 Mg 5 4.51 -16 7.5 +33 +49

5 1 5 3. SI -29 4.7 - 17 +12

El l. Mg 5 3.88 28 4.53 -20 +8

Ei l Sr 1 3.642 -32 5.500 -2 +30

El l Sr 3 3, 653 -32 5.075 - 10 +22

El l Sm 2 4.1 13 -23 5.365 -5 +18

Ei l 1 5 4.067 -24 4.435 -21 +3

El l Yb 2 3.693 -31 5.485 -3 +28

Ei l Yb 5 3.796 -29 5.105 -9 20

B6 Mg 5 3.44 -36 4.00 -29 +7

: Mg 5 5.67 +5 0 66 -89 -94

B4 Ca 2 7.549 +40 0.49 -92 -132

B4 Ba 2 9.784 +82 Ϊ0 60 -142

14 Sm 7.993 +48 1.400 -75 -123

In the second phase of the research, various metals were tested in device 2 in matrices El, E2 and CI _f with two different ETL thicknesses 40 nm (Ui and ¾) and 80 nm (L½ and U₄) and with two different doping concentrations 5 wt% (Uj and ¾) and 25 wt% (U3 and U₄), all for current density 10 mA/cm².

The results summarized in Table 2 led to a prelimmary conclusion that metals that are able to form stable compounds in oxidation state II are especially appropriate for n-doping in phosphine oxide matrices despite their significantly lower reactivity and higher air stability in comparison with the least reactive alkali metal (Li). From the divalent metals tested, only zinc having extremely high sum of the first and second ionization potential failed as n-dopant, whereas aluminium, with typical oxidation state III gave reasonably low operational voltages only if present in the doped ETL in the high 25 wt% concentration that afforded ETLs with impractically high light absorption. Transrnittance assigned as "OD" that stands for "optical density" is reported in Table 2 only for 25 wt% doping concentration (O ₃ for layer thickness 40 nm and OD₄ for layer thickness 80 urn), as the measurements for lower doping

concentrations suffered from bad reproducibility.

The typically trivalent bismuth failed as n-dopant completely, despite its ionization potential does not differ much, e.g. from manganese that showed, quite surprisingly, good doping action at least in EL

Low values of differences U1-U2 and U3-U4 can be assigned to doped materials having high conductivity (voltage of the device depends only weakly 011 the thickness of the doped layer).

Table 2

E₂ Mg 3.189 3.232 0.043 3.464 3.419 0.025 68 72

Ca 4.426 m 0.078 3.91 1 4.501 0.590 64 50

E₂ Sr 3.842 32 -0.010 - - - - -

E₂ Ba 29 2.935 m 97 3.397 30 74 71

¾ Sm 3.610 .3.894 0.284 53 39 1.887 72 63

Em - - - 4.516 4.838 0.322 - -

¾ Yb 2.932 » 0.001 5.442 6.625 73 65

E₂ Mn 6.02 8.09 0.99 - . - - -

E₂ Zn 7.898 >10 na 7.000 >10 na 66 71

E₂ 8.650 >10 7ia 3.203 3.196 -0.007 39 27

E₂ Bi 7.814 >10 na 7.173 >10 na 64 61

C I Li 6.997 >10 na » 8.314 2.105 72 48

CI Na - - - 4.417 4.455 0.037 56 31

CI Mg 4. I SO 4.178 -0.002 4.174 4.167 -0.007 62 57

CI Ca 4,031 4.104 0.074 3.619 3.61 -0.004 38 21

CI Sr 4.033 4.071 0.0038 - - - - -

CI Ba 3.916 3.909 -0.006 59 m 0.636 63 39

CI Sm 4.208 4.207 0.000 4.106 4.104 -0.002 63 48

Eu 3,972 3.984 0.012 - - - - - ci Yb 4.017 4.167 ^■0.003 4.148 4.173 0.025 33 29

CI Mn 4.27 4.26 -0.01 - . - - -

Zn 5.084 7.758 2.674 99 32 1.703 57 5£

CI Al 4, 152 4.949 0.797 3.135 3.123 -0.01 1 45 2f

C! Bi 4.842 6.355 1.513 4.306 4.603 0.297 59 68

It has been observed that in matrices with deep LUMO, like CI, the operational, voltage is often surprisingly higher than in devices comprising matrices with the LUMO levels in the range according to invention, despite good conductivity of many doped semiconducting materials based on CL Apparently, the good conductivity of a semiconducting material is not a sufficient condition for low operational voltage of the device comprising it. Based on this finding, it is supposed that doped semiconducting materials according to this invention enable, besides the reasonable conductivity, also efficient charge injection from the doped layer in the adjacent layer.

In the third research phase, the observed effects were confirmed in OLEDs of example 3 comprising alternative emitter systems and further embodiments of the invention described in examples 4- 7 were implemented. The achieved results summarized in the Table 3 confirmed the surprising superiority of phosphine oxide ETL matrices having higher LUMO levels (closer to vacuum level), despite these matrices should be more difficult to dope with the relatively weakly reducing metals used in the present invention in comparison with the phosphine oxide matrices of the previous art (like CI) which were thought to be dopable with Mg owing to their deeper LUMO (further away from vacuum level) and specific structure comprising the metal complexing unit

This series of experiments confirmed that also with other emitters, the matrix compounds like El and E2 having rather high LUMO energ levels perform better than other phospMne oxide matrix compounds, and much better in comparison with matrices lacking the phosphine oxide group.

These results showed that if combined with phosphine oxide matrices having sufficiently high LUMO levels, even substantially air stable metals, possessing moreover further technically advantageous features like good evaporability, can afford electrically doped semiconductive materials and devices that perform equally well or even better than devices available in the art.

Table 3

LEL ETL ETL 4 U EQE (EQE-EQE_ref) EQE_ref score matrix dopant dopant (V) (%)

(%) (%)

ABH- El I 5 3.498 -35 6.640 +18 +53 1 12/ E2 Mg 5 3.751 -30 5.975 +6 +36

NTJBD- CI Mg 5 4.545 - 15 3.905 -30 -15

369

C I O I 5 - - 0 no light -

Two - El Mg 5 3.480 -35 7.660 +36 +71 colour E2 Ϊ 5 3.83 -29 6.67 f , «< -48 fluoresc. CI I 5 0 -8 4.470 -20 - 12 white* CIO I 5 6.950 ^! +21 0.820 413

Spiro- El Mg 5 3.33 1 -38 6.19 +10 +4S Pye/

El Ca 5 3.31 1 -38 4.46 *^~ +18

BCzVB

El Ba 12 3.017 -42 3.44 +3

EI 8M 5 3.318 38 4.53 -19 +19

E2 Mg 5 0 -35 6.08 +1 +43

E2 Ca 5 3.497 ^■35 3.56 -37 +1

E2 Ba 5 3.090 -42 3,59 +6

C7 Mg 5 3,679 -31 2 -94 -63

C7 Ca 5 3.647 -33 0,52 -90 -57

* ABH-112/ UBD-369 + ABH-036/N D 129 (Sun Fine Chemicals)

Spiro-Pye BCzVB

Finally, the focus returned to the starting point. The remaining goal was to make the redox potential of the comparative compound C2 sigmficantly more negative, and test whether the corresponding LUMO level approaching closer to the zero in absolute energy scale has in phosphinc oxide compounds comprising a small conjugated system of less than 10 delocalized electrons similarly positive effects as in the matrix compounds described above. This task was accomplished relatively easily with commercially available compounds A1-A4. All these compounds have redox potentials that are difficult to measure by the standard procedure using THF as a solvent, because their values are more negative than for compound B7.

methylenebis(diphenylphosphme oxide), Al thane- 1 ,2-diylbis(diphenylphosphine oxide), A2

propane- 1 ,3-diylbis(diphenylphosphine oxide), A3

butane- 1 ,4-diylbis(diphenylphosphine oxide), A4.

Based oe mostly discouragieg results of previous experiments with C2 analogs and salt additives, rather negative results were expected from metal doping in matrices which are tree of a conjugated system of delocalized electrons or comprising less than 10 delocalized electrons in a conjugated system. Contrary to these expectations, it was surprisingly found that sufficiently strong dopants can provide semiconducting materials of good quality even with compounds having conjugated systems of delocalized electrons that are as small as mere six-electron Htlckel systems in isolated aromatic or heteroaromatic rings, provided that the compound still comprises polar groups selected from phosphine oxide and diazole. Subsequently, it was confirmed that such simple two-component semiconducting materials consisting of metal and the first compound having only six delocalized electrons in a conjugated system can be advantageously mixed with a second compound comprising larger conjugated systems of delocalized electrons without loss of performance, even though the second compound does not comprise any polar group at all. It was further surprisingly found that, contrary to previous experience with electron transport compounds comprising both polar group and the conjugated system of at least ten delocalized electrons, in matrices comprising polar group and a conjugated system of less than 10 delocalized electrons or free of the conjugated system of delocalized electrons, there is no significant difference anymore between the doping performance of divalent electropositive metals on one hand and other electropositive metals like alkali metals and typically trivalent rare earth metals like Sc, Y, La or Lu on the other hand. In this sense, it seems unnecessary to make any speculations about the extent of charge transfer between the doping electropositive element and the matrix and try to divide the electropositive elements used as n-dopants in semiconducting materials according to invention in groups of "strong" and "weak" dopants. It is, therefore, supposed that the electropositive element is still at least partially present in its substantially elemental form in the inventive semiconducting materials, and that the favourable results observed in matrices comprising polar phosph ne oxide or diazole groups are to be linked to specific interactions of these polar groups with atoms or atom clusters of the electropositive element.

Finally, subsequent experiments with matrix compounds comprising larger systems of delocalized electrons confirmed that with matrices having more negative reduction potentials than 4,7-diphenyl- 1 , 10-phenanthroline and with electropositive elements selected from Li, Na, K, Be, Sc, Y, La, Lu, Ti and V, excellent OLED performance can be achieved basically irrespective of the presence and/or extent of the conjugated system of delocalized electrons in the substantially organic electron transport compound comprising the polar group as defined in the present invention.

In the very final stage, it was additionally recognized that various combinations of polar matrices and doping metals can be advantageously used in combination with silver cathodes, particularly in case that the cathode is adjacent to the substantially organic layer comprising the polar matrix and the electropositive metal.

VI. EXAMPLES

Auxiliary materials

4, 4', 5, 5'-tetracyclohexyl-l, Γ, 2, 2^*, 3, 3"-hexamethyl-2, 1, 3, 3^,-tetrahydn>-lH, 1 Ή-biimidazolc, CAS 1253941-73-9, Fl ;

2,7-di(naphtal n-2-yl)spiro[fluorene-9,9'-xanthene], LUMO (CV vs. FcVFc) -2.63 V, WO2013/149958, F2;

biphenyl]-3,3'-diarnine, WO2014/060526, F3;

carbazol-3-yl)phenyl]-amine, CAS 1242056-42-3, F4;

ethyl-lH-benzo[d]imidazole, CAS 1254961-38-0, F5;

2,7-di([l , 1 '-biphenyl]-4-yl)spiro[fluorene-9,9'- xanthene], LUMO (CV vs. Fc⁺ Fc) -2.63 V, WO2013/149958, CAS 1466444-48-3, F6.

Auxiliary procedures

Cyclic voltammetry

The redo potentials given at particular compounds were measured in an argon deaerated, dry 0.1M THF solution of the tested substance, under argon atmosphere, with 0.1M tetrabutylammonium hexafluorophosphate supporting electrolyte, between platinum working electrodes and with an Ag AgCl pseudo-standard electrode, consisting of a silver wire covered by silver chloride and immersed directly in the measured solution, with the scan rate 100 mV/s. The first ran was done in the broadest range of the potential set on the working electrodes, and the range was then adjusted within subsequent runs appropriately. The final three runs were done with the addition of ferrocene (in 0.1M concentration) as the standard. The average of potentials corresponding to cathodic and anodic peak of the studied compound, after subtraction of the average of cathodic and anodic potentials observed for the standard Fc7Fc redox couple, afforded finally the values reported above. All studied phosphine oxide compounds as well as the reported comparative compounds showed well- defined reversible electrochemical behaviour.

Synthesis examples

The synthesis of phosphine oxide ETL matrix compounds is well described in many publications, besides the literature cited at particular compounds listed above and describing typical multistep procedures used for these compounds, the compound E6 was prepared, according to Bull. Chem. Soc. Jpn., 76, 1233—1244 (2003), quite specifically by an anionic rearrangement of the compound E2.

For yet unpublished compounds, the typical procedures were used, as exemplified below specifically for the compounds E5 and E8. All synthesis steps were carried out under argon atmosphere. Commercial materials were used without additional purification. Solvents were dried by appropriate means and deaerated by saturation with argon.

Synthesis example 1

[14':4'J" erphenyl]-3₅5-diylbis-dipheiiylphospriine oxide (E5)

Step 1 : 3 ,5-dibromo- 1 , 1 ^*:4', 1 "-terphenyl

All components (10.00 g (1.0 eq, 50.50 mmol) [1,1 '-biphenyl]-4-yl-boronic acid, 23.85 g

(1.5 eq, 75.75 mmol) 1,3,5-tribromobenzene, 1.17 g (2.0 mol%, 1.01 mmol) tetrakis(triphenyl phosphine)palladium(O), 10.70 g (2 eq, 101 mmol) sodium carbonate in 50 mL water, 100 mL ethanol and 310 mL toluene) were mixed together and stirred at reflux for 21 hours. The reaction mixture was cooled to room temperature and diluted with 200 mL toluene (three layers appear). The aqueous layer was extracted with 100 mL toluene, the combined organic layers were washed with 200 mL water, dried and evaporated to dryness. The crude material was purified via column chromatography (S1O₂, hexane/DCM 4:1 v/v) The combined fiactions were evaporated, suspended in hexane and filtered off to give 9.4 g of a white glittering solid (yield 48 %, HPLC purity 99.79 %).

Ste 2: [l, :4',l "-terphenyl]-3,5-diylbis-diphenylphosphine oxide

All components (5.00 g (1.0 eq, 12.9 mmol) 3 ,5-dibromo- 1 , Γ:4', 1 "-terphenyl from the previous step, 12.0 g (5.0 eq, 64.4 mmol) diphenyl phosphinc, 114 mg (5 mol%, 6.44 x 10^{' 4} mol) palladium(II) chloride, 3.79 g (3.0 eq, 38.6 mmol) potassium acetate and 100 mL N,N- dimethylformamide) were mixed together and stirred at reflux for 21 hours. Then the reaction mixture was cooled to room temperature; water was added (100 mL) and the mixture was stirred for 30 min, then filtered off. The solid was re-dissolved in DCM (100 mL), H₂0₂ (30 wt% aqueous solution) was added dropwise, and the solution was stirred overnight at room temperature. Then the organic layer was decanted, washed with water (100 mL) twice, dried over MgSC , and evaporated to dryness. The resulting oil was triturated in hot MeOH (100 mL) which induced the formation of a solid. After filtration hot, the resulting solid was rinsed with MeOH and dried, yielding 9.7 g of crude product. The crude material was re-dissolved in DCM and chromatographed on a short silica column, elution with ethyl acetate. After evaporation of the eluate to dryness, the resulting solid was triturated in hot MeOH (100 mL), followed by trituration in hot ethyl acetate (50 mL). After drying, the desired compound was obtained in 70 % yield (5.71 g). Finally, the product was purified using vacuum sublimation. The pure sublimed compound was amorphous, with no detectable melting peak on the DSC curve, glass transition onset at 86 °C, and started to decompose at 490 °C.

Synthesis example 2

(9,9-dmexyl-9H-fluorene-2,7-diyl)bis-diphenylphosphine oxide E8

2,7-Dibromo-9,9-dihexylfluorene (5.00 g, 1.0 eq, 10.2 mmol) was placed in a flask and deaerated with argon. Then anhydrous THF (70 mL) was added, and the mixture was cooled to -78 °C. 9.7 mL (2.5M solution in hexanes, 2.4 eq, 24.4 mmol) n-butyllithium were then added dropwise; the resulting solution was stirred for 1 h at -78 °C, and then progressively warmed to -50 °C. After slow addition of pure chlorodiphenylphosphine (4.0 mL, 2.2 eq, 22.4 mmol), the mixture was left to stir overnight till room temperature. MeOH (20 mL) was added to quench the reaction, and the solution was evaporated to dryness. The solid was re- dissolved in DCM (50 mL). H₂0₂ (30 wt% aqueous solution, 15 mL) was added dropwise, and the mixture left under stirring. After 24 fa, the organic phase was separated, washed subsequently with water and brine, dried over MgSC^, and evaporated to dryness. Purification by chromatography (silica, gradient elution from hexane / EtOAc 1 :1 v/v to neat EtOAc) provided the desired compound in 34 % yield (2.51 g). The material was then further purified by vacuum sublimation.

The pure sublimed compound was amorphous, with no detectable melting peak on the DSC curve, and decomposed at 485 °C.

Synthesis example 3

Diphenyl-(3-(spiro[fluorene-9,9'-xanthen]-2-yl)phenyl)phosphine oxide (El 4)

Step 1.

Synthesis of 2-bromospiro|fluorene-9,9'-xantheneJ

Chemical Formula: C13H B1O Chemical Formula: C₂₅H₁₅BrO

Molecular Weight: 259,10 Molecular Weight 411 ,30

2-Bromo-9-fluorenone (10.00 g, 1.0 eq, 38.6 mmol) and phenol (34.9 g, 9.6 eq., 0.37 mol) were placed in a two-necked flask and degassed with argon. Methanesulfbnic acid (10.0 mL, 4.0 eq, 0.15 mol) was added, and the resulting mixture was refluxed for 4 days at 135°C. After cooling to room temperature,

DCM (10 mL) and water (130 mL) were added. Upon stirring, the material precipitates. After filtration and abundant washing with MeOH, it was finally triturated in hot EtOH (60 mL) for lh, which after filtration afforded the desired compound (11.0 g, 69%), > 99% purity according to GCMS.

Step 2.

Synthesis of diphenyl(3^spiro[nuoren -9^9'-xanthen]-2--yl)phenyl)pho8phine oxide

Chemical Formula: C_isH₁₅BrO Chemical Formula: C,₈H₁₄BrOP Chemical Formula:

Molecular Weight: 411 ,30 Molecular Weight: 357,19 Molecular Weight: 808,68

Ni(dppp)C½ [1 ,3-Bis(diphenylprioephino)propane]nickel(!l} chloride C2₇H_¾P₂CI₂Ni 575,0 g/mol

2-Bromospiro[fluorene-9,9'-xanthene] (10.0 g, 1.2 eq, 24.3 mmol) was charged in a two-necked flask, degassed with argon, and dissolved in anhydrous THF (240 mL). To this solution were added Mg° (827 mg, 1.4 eq, 34.0 mmol), followed by iodomethane (414 mg, 0.12 eq., 2.92 mmol), and the resulting mixture was refluxed for 2 hours. Then, this Grignard solution was cannulated to an anhydrous solution of (3-bromophenyl)dipheny!phosphine oxide (7.23 g, 1.0 eq., 20.3 mmol) and [ 1 ,3-bis(diphenylphosphino)propane]nickel(II) chloride (263 mg, 2.0 mol%, 0.49 mmol) in THF (200 mL). The resulting mixture was refluxed overnight after which it was quenched by addition of water (5 mL). The organic solvent was removed under reduced pressure, and the compound extracted with CHCI₃ (200 mL) and water ( 100 mL); the organic phase was decanted, further washed with water (2 x 200 mL), dried over MgS0₄ and evaporated to dryness. The crude product was filtered over silica; upon elution with n-hexane / DCM 2:1, apolar impurities was removed, while the desired was isolated using pure DCM, After removal of the DCM, the product was triturated in ethylacetate (100 mL), filtered off, and dried under vacuum yielding the title compound (7.5 g, 61%). Finally, the product was purified by sublimation (78% yield).

Mater rjrqgerties:

DSC: Melting point: 264°C (peak), sublimed material

CV; LUMO vs. Fc (THF): -2.71 V (reversible)

Synthesis example 4

Diphenyl-(4-(spiro[fluorene-9,9'-xanthen]-2-yl)phenyl)phosphine oxide (E15)

Chemical Formula: C_2SH ·, ₅BrO Chemical Formula: C₁₈H ₁₄BiOP Chemical Formula:

Molecular Weight: 411 ,30 Molecular Weight: 357,19 Molecular Weight: 608,66

Ni(dppp)CI₂ [1 ,3-Bis(diphenylphosphino)prapaneJnickeKII) chloride ΟΣ Η^Ρ^ΝΙ 575,0 g/mol

2-Bromospiro[fluorene-9,9'-xantheneJ (10.0 g, 1.2 eq, 24.3 mmol) was charged in a two-necked flask, degassed with argon, and dissolved in anhydrous THF (240 mL), To this solution were added Mg° (827 mg, 1.4 eq, 34.0 mmol), followed by iodomethane (414 mg, 0.12 eq., 2.92 mmol), and the resulting mixture was refluxed for 2 hours. Then, this Grignard solution was cannulated to an anhydrous solution of (4-bromopheny!)diphenylphosphine oxide (7.23 g, 1.0 eq., 20.3 mmol) and [l ,3-bis(diphenylphosphino)propane]nickel(Il) chloride (263 mg, 2.0 mol%, 0.49 mmol) in THF (200 mL). The resulting mixture was refluxed overnight after which it was quenched by addition of water (5 mL). The organic solvent was removed under reduced pressure, and the compound extracted with DCM (2 L) and water (500 mL); the organic phase was decanted, dried over MgSC¾ and evaporated to dryness. The crude product was purified by chromatography over silica, elution with DCM / MeOH 99: 1. The fractions containing the product were merged and evaporated to dryness. The resulting solid was then triturated in ethylacetate (50 mL), filter off, and dried under vacuum yielding the title compound (4.0 g, 32%). Finally, the product was purified by sublimation (77% yield).

Material properties: DSC: Melting point: 255°C (peak), sublimed material

CV: LUMO vs. Fc (THF): -2.65 V (reversible)

Device examples

Comparative example 1 (blue OLED)

A first blue emitting device was made by depositing a 40 nm layer of HTM2 doped with PD2 (matrix to dopant weight ratio of 97:3 wt%) onto an ITO-glass substrate, followed by a 90 run undoped layer of HTML Subsequently, a blue fluorescent emitting layer of ABH113 (Sun Fine Chemicals) doped with NUBD370 (Sun Fine Chemicals) (97:3 wt%) was deposited with a thickness of 20 nm. A 36 nm layer of the tested compound was deposited on the emitting layer together with the desired amount of the metallic element (usually, with 5 wt% Mg) as ETL, Subsequently, an aluminium layer with a thickness of 100 nm was deposited as a cathode.

The observed voltages and quantum efficiencies at a current density 10 mA/cm² are reported in the Table 1.

Comparative example 2 (organic diode)

A similar device was produced as in Example 1, with the difference that the emitter was omitted, and each combination matrix-dopant was tested in two different thicknesses of the ETL (40 and 80 nm) and with two different dopant concentrations (5 and 25 wt%). The observed voltages at the current density 10 mA/cm² and, wherever measured, optical absorbances of the device, are reported in the Table 2.

Comparative example 3 (blue or white OLED)

A similar device was produced as in. Example 1 , with the difference that there were combined various compositions of semiconducting materials in the ETL with various emitter systems.

The results were evaluated similarly as in Example 1 and are summarized in Table 3

Comparative example 4 (blue OLED)

In device of Example 1 , Al cathode was replaced with the sputtered indium tin oxide (ITO) cathode in combination with the Mg or Ba doped ETL. The results showed that ETLs based on divalent metal doped phosphine oxide matrices having redox potentials vs Fc+ Fc in the range from -2.24 V to -2,81 V are applicable also in top emitting OLEDs with cathode made of transparent semiconductive oxide.

Comparative example 5 (transparent OLED)

In transparent devices having p-side (substrate with 1TO anode, HTL, EBL) as in Example 1, and sputtered 100 nm thick ITO cathode as in Example 4, polymeric emitting layer comprising blue emitting polymer (supplied by Cambridge Display Technology) was successfully tested. The results reported in Table 4 (together with the n-side composition of the device, which in all cases comprised a 20 nm thick HBL consisting of F2 and ETL1 consisting of E2 and D3 in weight ratio 7:3 and having a thickness given in the table) show that ETLs based on phosphine oxide compounds doped with divalent metals are applicable even with polymeric LELs having very high LUMO levels about -2.8 V (in terms of their redox potential vs. FCVFC reference). Without metal doped ETL, the devices had (at current density 10 mA/cm²) very high voltages, even when EILs made of pure metal were deposited under the ITO electrode.

Table 4

Comparative example 6 (metal deposition using linear vaporization source)

Evaporation behaviour of Mg in a linear evaporation source was tested. It was demonstrated that Mg can be deposited from linear sources with the rate as high as 1 nm/s without spitting, whereas point evaporation sources spit Mg particles at the same deposition rate significantly. Comparative example 7 (metal + metal salt electrical doping in the same ETL)

Mixed ETL comprising a matrix combined with LiQ + either Mg or W₂(hpp)4 a combined two-component doping system showed the superiority of the salt + metal combination.

Comparative example 8 (tandem white OLED)

On an ITO substrate, following layers were deposited by vacuum thermal evaporation:

10 nm thick HTL consisting of 92 wt% auxiliary material F4 doped with 8 wt% PD2, 135 nm thick layer of neat F4, 25 nm thick blue emitting layer ABHl 13 (Sun Fine Chemicals) doped with NUBD370 (Sun Fine Chemicals) (97:3 wt%), 20 nm thick layer ABH036 (Sun Fine Chemicals), 10 nm thick CGL consisting of 95 wt% inventive compound E12 doped with 5 wt% Mg, 10 nm thick HTL consisting of 90 wt% auxiliary material F4 doped with 10 wt% PD2, 30 nm thick layer of neat F4, 15 nm thick layer of neat F3, 30 nm thick proprietary phosphorescent yellow emitting layer, 35 ran thick ETL of auxiliary material F5, 1 nm thick LiF layer and aluminium cathode. The diode operated at 6.81 V had EQE 24.4 %.

Comparative example 9 (tandem white OLED)

The example 8 was repeated with Yb in the CGL instead of Mg. The diode operated at 6.80 V had EQE 23.9 %.

Comparative example 10 (tandem white OLED)

The example 9 was repeated with compound E6 instead of El 2 in the CGL. The diode operated at 6.71 V had EQE 23.7 %.

Comparative example 11 (charge injection into adjacent or admixed high-LUMO ETM in a blue OLED)

Example 1 was repeated with following modifications:

On the ITO substrate, 10 nm thick HTL consisting of 92 wt% auxiliary material F4 doped with 8 wt% PD2 followed by 130 nm thick layer of neat F4 were deposited by VIE. O top of the same emitting layer as in example 1, a 31 nm thick HBL of F6 and on top of it the doped ETL according to Table 5 were deposited subsequently, followed with the aluminium cathode. All deposition steps were done by VTE under pressure below 10^"2 Pa.

Table 5 l!g Yb Li

36nm 36nm 36nm 36iim

ETM;Mg(95;5) ETM:Yb(95:5) ETM;Ba(95:5) ETM:Li(99,5:0,5)

ETM

V at

V at 10mA/cm² V at lOttiA/cm² V at 10m A/cm² l OmA/crn²

E5 2.58 4.2 3.7

Εδ -2.62 6.1 3.5

E7 -2.81 > 5.2 3.7

E13 -2,78 3.9 3.2 3.4

P -2.71 3.3

A < -3.J 0 3 J 3.9 .1.3

A4 < -3.10 4.1 3.2

,'..-::F6 -2.63 3.3 3.3

<*__. n n. 3.9 3.7

Cl l -2.45 >10.0 5.9 5.3

The experiments convincingly showed that compounds E5, E6, E7, E13 and E14 as well as

A2 and A4 and their mixtures with ETMs which are fee of a polar group provided very good electron injection into adjacent F6 layer, despite its highly negative reduction potential. Due to highly negative reduction potential and low polarity of the F6 layer, the model device mimicks also the properties of devices comprising emitting layers made of light emitting polymers.

Comparative example 12 (charge injection into adjacent or admixed Mgh-LUMO ETM in a blue OLED)

Example 11 was repeated under replacement F6 in HBL with BIO. Composition of ETLs and results are shown in Table 6.

Table 6

The results showed that semiconducting materials based on phosphine oxide matrices doped with divalent metals allow an efficient electron injection also into a CBP layer having even more negative redox potential than the HBL matrix of the previous example.

Comparative example 13 (applicability of semiconducting materials based on phosphine oxide matrices doped with divalent metals in very thick ETLs)

Example 11 was repeated with ETLs having thickness 150 nm. The results are shown in Table 7.

Table 7

Comparison with Table 5 demonstrates that devices utilizing in ETLs the inventive devices comprising compounds with highly negative reduction portentials E5, E6, E7, El 3, E14, A2 and A4 show practically the same operational voltages as devices of Example 11 , despite the thickness of the ETL increased more than four times. The experiment shows that the OLED design according to present invention enables easily tuning the size of optical cavity in electronic devices comprising emitting layers with very negative redox potentials, like light emitting polymers.

Inventive example 14 (bottom emitting OLED with, reflecting silver cathode)

The device of Example 1 1 was reproduced with replacement of aluminium in the cathode with silver, compound CI as the electron, transport matrix and Mg as the electropositive element. With Ag and Al cathode, the device operated at comparable voltages 4.0 and 3.9 V, but the quantum efficiency with Ag was 5.3 %, whereas with Al 4.9 %. The comparative device comprising as ETL compound CI doped with Dl operated with Al cathode at the voltage 4.7 V with quantum efficiency 3.3 %, and with Ag cathode at 5.7 V and quantum efficiency 2.5 %.

The features disclosed in the foregoing description, in the claims and in the accompanying drawings may both separately and in any combination be material for realizing the invention in diverse forms thereof. Reference values of physico-chemical properties relevant for the present invention (first and second ionization potential, normal boiling point, standard redox potential) are summarized in Table 8.

Table 8

Element ^' 1° eV ² eV ² eV ² °C V

Li 5.391 75.640 81.031 3 -3.04

Na 5.139 47.286 52.425 890 -2.713 Mg 7.646 15.035 22.681 1 1 10 -2,372

Al 5.986 18.829 24.835 2470 -1.676

Ca 6.1 13 1 1.872 17.985 1487 -2.84

ME 7.434 15.640 23.074 21 0 -1.18

Zn 9.394 17.964 27.358 907 -0.793

Sr 5.695 11.030 16.725 1380 -2.89

Ba 5.212 10.004 15.216 1637 -2.92

Sm 5.644 1 1.07 16.714 1900 **

Eu 5.670 1 1.241 16.91 1 1440 -1.99

Yb 6.254 ΊζΪ7( 18.430 1430 -2.22

Bi 7.286 16.69 23.976 1560 0.317

Timing Zhang, Julian R. G. Evans, Shoufeng Yang: Corrected Values for Boiling Points and Enthalpies of Vaporization of Elements in Handbooks. From: Journal of Chemical & Engineering Data. 56. 201 1, p. 328-337; the values fit with values given in articles for individual elements in current German version of Wikipedia. http://en.wikipedia.org wiki/Ionization_eneTgies of the elements %28data_page%29

Used abbreviations

CGL charge generating layer

CV cyclic voltarnrnetry

DCM dichloromethane

DSC differential scanning calorimetry

EIL electron injecting layer

EQE external quantum efficiency of electroluminescence

ETL electron transporting layer

ETM electron transport matrix

EtOAc ethyl acetate Fc⁺/Fc ferrocen i um/ferrocene reference system h hour

HBL hole blocking layer

HIL hole injecting layer

HOMO highest occupied molecular orbital

HTL hole transporting layer

HTM hole transport matrix

ITO indium tin oxide

LUMO lowest unoccupied molecular orbital

LEL light emitting layer

LiQ lithium 8-hydroxyquinolinolate

MeOH methanol

mol% molar percent

mp melting point

OLED organic light emitting diode

QA quality assurance

RT room temperature

THF tetrahydrofuran

UV ultraviolet (light)

vol% volume percent

v/v volume/volume (ratio)

VTE vacuum thermal evaporation wt% weight (mass) percent

Claims

Claims

1. Electronic device comprising at least one light emitting layer between an anode and a substantially silver cathode, the device further comprising between the cathode and the anode at least one mixed layer comprising

(i) at least one substantially covalent electron transport matrix compound comprising at least one polar group selected from phosphine oxide group or diazole group, and

(ii) in substantially elemental form, an electropositive element selected from

substantially non-radioactive alkali metals, alkaline earth metals, rare earth metals, and transition metals of the fourth period of the Periodic table having proton numbers 22, 23, 24, 25, 26, 27, 28, 29.

2. Electronic device according to claim 1, wherein the reduction potential of the electron transport matrix compound, if measured by cyclic voltammetry under the same conditions, has the value which is more negative than the value obtained for tris(2- benzo [d]thiazol-2-yl)phenoxyaluminum, preferably more negative than for 9,9', 10,10'- tetraphenyl-2,2'-bianthracene or 2,9-di([l,l '-biphenyl] -4-yl)-4,7-diphenyl- 1,10- phenanthroline, more preferably more negative than for 2,4,7,9-tetraphenyl-l , 10- phenanthroline, even more preferably more negative than for 9, 10-di(naphthalen-2-yl)-2- phenylanthracene, even more preferably more negative than for 2,9-bis(2- methoxyphenyl)-4,7-diphenyl- 1 , 10-phenanthroline, even more preferably more negative than for 9,9'-spirobi [fluorene] -2,7-diylbis(diphenylphosphine oxide), even more preferably more negative than for 4,7-diphenyl-l , 10-phenanthroline, even more preferably more negative than for 1 ,3 ,5-tris( 1 -phenyl- 1 H-benzimidazol-2-yl)benzene, most preferably more negative than for pyrene and still preferably more negative than the value obtained for [1,1 ^,-binaphthalen]-2,2^,-diylbis(diphenylphosphine oxide).

3. Electronic device according to claim 1 or 2, wherein the silver content in the cathode is at least 99.5 weight %, more preferably at least 99.8 weight %, even more preferably at least 99.9 weight %, most preferably at least 99.95 weight %.

4. Electronic device according to any of claims 1-3, wherein the electropositive element is selected from Li, Na, K, Be, Mg, Ca, Sr, Ba, Sc, Y, La, Sm, Eu, Tb, Yb, Lu, Ti, V and Mn.

5. Electronic device according to any of claims 1-4, wherein the phosphine oxide group is part of a substantially covalent structure comprising at least three carbon atoms directly attached to the phosphorus atom of the phosphine oxide group and having overall count of covalently bound atoms, preferably selected from C, H, B, Si, N, P, O, S, F, CI, Br and I, in the range 16-250 atoms, more preferably in the range 32-220 atoms, even more preferably in the range 48-1 0 atoms, most preferably in the range 64-160 atoms,

6. Electronic device according to any of claims 1-5, wherein the phosphine oxide group is selected from phosphine oxide substituted with three monovalent hydrocarbyl groups or one divalent hydrocarbylene group forming with the phosphorus atom a ring arid one monovalent hydrocarbyl group, and the overall count of carbon atoms in the hydrocarbyl groups and the hydrocarbylene group is 8-80 carbon atoms, preferably 14-72 carbon atoms, more preferably 20-66 carbon atoms, even more preferably 26-60 carbon atoms, most preferably 32-54 carbon atoms.

7. Electronic device according to any of claims 1-6, wherein the diazole group is an

imidazole group.

8. Electronic device according to any of claims 1-7, wherein the electron transport matrix compound comprises a conjugated system of at least ten delocalized electrons.

9. Electronic device according to my of claims 1-8, wherein the electron transport matrix compound comprises at least one aromatic or heteroaromatic ring.

10. Electronic device according to claim 9, wherein the electron transport matrix compound comprises at least two aromatic or heteroaromatic rings which are either linked by a covalent bond or condensed.

11. Electronic device according to any of claims 1-10, wherein the molar ratio of the

electropositive element to the electron transport compound is lower than 0.5, preferably lower than 0.4, more preferably lower than 0.33, even more preferably lower than 0.25, even more preferably lower than 0.20, even more preferably lower than 0.17, most preferably lower than 0.15, still preferably lower than 0.13, still but less preferably lower than 0.10.

12. Electronic device according to any of claims 1-1 1, wherein the molar ratio of the

electropositive element to the electron transport compound is higher than 0.01, preferably higher than 0.02, more preferably higher than 0.03, even more preferably higher than 0.05, most preferably higher than 0.08.

13. Electronic device according to any of claims 1-12, which is an inverted device.

14. Electronic device according to any of claims 1-12, which is a top-emitting device with the cathode having thickness less than 30 nm, preferably less than 25 nm, more preferably less than 20 nm, most preferably less than 15 nm. Electronic device according to any of claims 1-14, wherein the silver cathode is corrugated and/or combined with a light-scattering layer.