Electroluminescent device

Abstract

A solid state light-emitting device comprising: a first electrode coupled to a first charge injecting layer; a second electrode coupled to a second charge injecting layer; an emissive layer comprising a perovskite material, wherein the emissive layer is provided between the first and second charge injecting layers; and wherein the bandgaps of the first and second charge injecting layers are larger than the bandgap of the emissive perovskite layer.

Claims

1. A solid state light-emitting device comprising: a first electrode coupled to a first charge injecting layer; a second electrode coupled to a second charge injecting layer; an emissive layer comprising an organometal halide perovskite material or a metal halide perovskite material, wherein the emissive layer is provided between the first and second charge injecting layers; and wherein the optical bandgaps of the first and second charge injecting layers are larger than the optical bandgap of the emissive perovskite layer; wherein said organometal halide perovskite material or metal halide perovskite material contains an AMX.sub.3 unit cell, where A is a monovalent organic cation or a monovalent metal cation, M is a divalent cation and X is a halide anion; and wherein at least one of the charge injecting layers form a type-I heterojunction with the emissive layer.

2. A solid state light-emitting device comprising: a first electrode coupled to a first charge injecting layer; a second electrode coupled to a second charge injecting layer; an emissive layer comprising an organometal halide perovskite material or a metal halide perovskite material, wherein the emissive layer is provided between the first and second charge injecting layers; and wherein the first charge injecting layer consists essentially of a hole injecting polyelectrolyte, and the second charge injecting layer has an optical bandgap that is larger than the optical bandgap of the emissive perovskite layer; wherein said organometal halide perovskite material or metal halide perovskite material contains an AMX.sub.3 unit cell, where A is a monovalent organic cation or a monovalent metal cation, M is a divalent cation and X is a halide anion; and wherein at least one of the charge injecting layers form a type-I heterojunction with the emissive layer.

3. A solid state light-emitting device as recited in claim 1 wherein at least one of the charge injecting layers has an optical bandgap of from 1.5 eV to 5 eV.

4. A solid state light-emitting device as recited in claim 1 wherein the emissive layer has a thickness of less than 100 nm.

5. A solid state light-emitting device as recited in claim 1 wherein at least one of the charge injecting layers is formed of a semiconducting material.

6. A solid state light-emitting device as recited in claim 1 wherein at least one of the charge injecting layers is formed of an organic semiconducting material.

7. A solid state light-emitting device as recited in claim 1 wherein a charge injecting layer consists essentially of a hole injecting organic semiconducting material and is selected from the group consisting of PEDOT:PSS, PANI (polyaniline), polypyrrole, substituted polypyrrole, doped poly(ethylene dioxythiophene) (PEDOT).

8. A solid state light-emitting device as recited in claim 1 wherein a charge injecting layer consists essentially of an electron injecting organic semiconducting material and is selected from the group consisting of poly(fluorene)s, including F8, TFB, F8BT or F8-TFB AB copolymer (95:5 F8:TFB).

9. A solid state light-emitting device as recited in claim 1 wherein the charge injecting layer is an electron injecting inorganic semiconducting material and is selected from the group consisting of titanium dioxide (TiO.sub.2), zinc oxide (ZnO), magnesium zinc oxide (MgZnO) and aluminium-doped ZnO (AZO).

10. A solid state light-emitting device as recited in claim 1 wherein the charge injecting layer is a hole injecting organic semiconducting material and is selected from the group consisting of polyfluorenes (preferably F8, TFB, PFB or F8-TFB) or Spiro-OMeTAD or polycarbazole (preferably poly(9-vinylcarbazole)) or 4,4′-Bis(N-carbazolyl)-1,1′-biphenyl.

11. A solid state light-emitting device as recited in claim 1 wherein at least one of the first and second electrodes is formed of a transparent conductive material.

12. A solid state light-emitting device as recited in claim 11 wherein the first electrode is an anode and the transparent conductive material is selected from: indium tin oxide (ITO), fluorine doped tin oxide (FTO), indium zinc oxide, graphene, carbon nanotubes, and a metal with a thickness of less than 20 nm.

13. A solid state light-emitting device as recited in claim 1 wherein there is an insulating layer formed between either or both of the charge injecting layers and the light-emitting layer.

14. A solid state light-emitting device as claimed in claim 13 wherein the insulating layer is formed of an oxide or nitride and has a thickness of less than 30 nm.

15. A solid state light-emitting device as recited in claim 13 wherein the insulating layer consists essentially of aluminium oxide, silicon dioxide, silicon nitride, zinc oxide modified with aluminium oxide, nickel oxide or magnesium oxide.

16. A solid state light-emitting device as recited in claim 13 wherein the insulating layer consists essentially of poly(ethyleneimine) (PEI), polyethylenimine-ethoxylated (PEIE), polystyrene (PS) or poly(methylmethacrylate) (PMMA).

17. A solid state light-emitting device as claimed in claim 13 wherein the insulating layer is deposited by atomic layer deposition, ALD.

18. A device according to claim 1, wherein a thin layer of <30 nm of a material selected from molybdenum trioxide or tungsten trioxide is deposited: between a first charge injecting layer and a first electrode, between a second charge injecting layer and a second electrode, between a first charge injecting layer and an emissive layer, between a second charge injecting layer and an emissive layer, between a first electrode and an emissive layer, or between a second electrode and an emissive layer.

19. A solid state light-emitting device as claimed in claim 1 wherein said organometal halide perovskite or metal halide perovskite has an AMX.sub.3 structure, where A is a monovalent organic cation or a monovalent metal cation, M is a divalent cation and X is a halide anion.

20. A solid state light-emitting device as claimed in claim 1 wherein divalent cation M is a divalent metal cation.

21. A solid state light-emitting device as claimed in claim 20 wherein the divalent metal cation is tin (Sn.sup.2+) or lead (Pb.sup.2+).

22. A solid state light-emitting device as recited in claim 1, wherein the monovalent organic cation is a primary, secondary or tertiary ammonium cation [HNR R.sup.2R.sup.3].sup.+, wherein each of R.sup.1, R.sup.2 and R.sup.3 may be the same or different and is selected from hydrogen, an unsubstituted or substituted C.sub.1-C.sub.20 alkyl group and an unsubstituted or substituted C.sub.5-C.sub.18 aryl group.

23. A solid state light-emitting device as recited in claim 1, wherein the monovalent organic cation is of the form [R.sup.1R.sup.2N—CH═NR.sup.3R.sup.4].sup.+: ##STR00003## wherein each of R.sup.1, R.sup.2, R.sup.3 and R.sup.4 may be the same or different and is selected from hydrogen, an unsubstituted or substituted C.sub.1-C.sub.20 alkyl group and an unsubstituted or substituted C.sub.5-C.sub.18 aryl group.

24. A solid state light-emitting device as recited in claim 1, wherein the monovalent organic cation is of the form (R.sup.1R.sup.2N)(R.sup.3R.sup.4N)c═N.sup.+R.sup.5R.sup.6: ##STR00004## wherein each of R.sup.1, R.sup.2, R.sup.3, R.sup.4, R.sup.5 and R.sup.6 may be the same or different and is selected from hydrogen, an unsubstituted or substituted C.sub.1-C.sub.20 alkyl group and an unsubstituted or substituted C.sub.5-C.sub.18 aryl group.

25. A solid state light-emitting device as recited in claim 1, wherein the monovalent metal cation is an alkali metal cation.

26. A solid state light-emitting device as recited in claim 1, wherein the monovalent metal cation is caesium (Cs.sup.+) or rubidium (Rb.sup.+).

27. A solid state light-emitting device as recited in claim 1, wherein X is a halide anion selected from chloride, bromide, iodide, and fluoride and, in the AMX.sub.3 unit cell each halide may be the same or different.

28. A solid state light-emitting device as recited in claim 1, wherein the organometal halide perovskite material or metal halide perovskite material contains an Al-i BMX.sub.3 unit cell, wherein: A and B are each a monovalent organic cation or a monovalent metal cation as claimed in any one of claims 22 to 26, where A and B are different; M is a divalent metal cation; X is a halide anion; and i is between 0 and 1.

29. A solid state light-emitting device as recited in claim 1, wherein the organometal halide perovskite material or metal halide perovskite material contains an AMX.sub.3-kY.sub.k unit cell, wherein: A is a monovalent cation; M is a divalent metal cation; X and Y are each a halide anion, where X and Y are different; and k is between 0 and 3.

30. A solid state light-emitting device as recited in claim 1, wherein the organometal halide perovskite material or metal halide perovskite material contains an AM.sub.1-jN.sub.jX.sub.3 unit cell, wherein: A is a monovalent cation; M and N are each a divalent metal cation; X is a halide anion; and j is between 0 and 1.

31. A solid state light-emitting device as recited in claim 1, wherein the organometal halide perovskite material or metal halide perovskite material contains an A.sub.1-i B.sub.iM.sub.1-jN.sub.jX.sub.3-kY.sub.k unit cell, wherein: A and B are each a monovalent cation, where A and B are different; M and N are each a divalent metal cation; X and Y are each a halide anion, where X and Y are different; and where i is between 0 and 1, j is between 0 and 1, and k is between 0 and 3.

32. A method of manufacturing a solid state light-emitting device, the method comprising: providing a first electrode on a substrate; depositing a first charge injecting layer over the first electrode; depositing an emissive perovskite layer over the first charge injecting layer; depositing a second charge injecting layer over the emissive perovskite layer; and depositing a second electrode over the second charge injecting layer; and wherein the bandgaps of the first and second charge injecting layers are larger than the bandgap of the emissive perovskite layer.

33. A method of manufacturing a solid state light-emitting device as claimed in claim 32 wherein the first electrode is an anode and is formed by the deposition of a transparent conductive material, preferably indium tin oxide (ITO) or fluorine doped tin oxide (FTO).

34. A method of manufacturing a solid state light-emitting device as claimed in claim 28 wherein the emissive perovskite layer is composed of a single, homogenous phase perovskite material.

35. A method of manufacturing a solid state light-emitting device as recited in claim 32 wherein the emissive layer has a thickness of less than 100 nm.

36. A method of manufacturing a solid state light-emitting device as recited in claim 32 wherein a thin insulating layer is deposited between either or both of the charge injecting layers and the light-emitting layer.

37. A method as claimed in claim 36 wherein the insulating layer is formed of an oxide or nitride and has a thickness of less than 30 nm.

38. A method as claimed in claim 36 wherein the insulating layer is selected from the group consisting of aluminium oxide, silicon dioxide, silicon nitride, zinc oxide modified with aluminium oxide, nickel oxide or magnesium oxide.

39. A method as claimed in claim 36 wherein the insulating layer is an insulating polymer selected from the group consisting of poly(ethyleneimine) (PEI), polyethylenimine-ethoxylated (PEIE), polystyrene (PS) and poly(methylmethacrylate) (PMMA).

40. A method as claimed in claim 36 wherein the insulating layer is deposited by atomic layer deposition, ALD.

41. A method of manufacturing a solid state light-emitting device as recited in claim 32 wherein at least one of the first charge injecting layer and the second charge injecting layer is formed of a semiconductor material.

42. A method of manufacturing a solid state light-emitting device as recited in claim 32 wherein a thin layer of <30 nm of a material selected molybdenum trioxide and tungsten trioxide is deposited between: the transparent conductive electrode and the perovskite layer, between a charge injecting layer and a conductive electrode, between the transparent conductive electrode and a charge injecting layer, between the perovskite layer and a charge injecting layer, or between the perovskite layer and a conductive electrode.

43. A method of manufacturing a solid state light-emitting device as recited in claim 32 wherein deposition of the layers is performed using one or more of the following deposition techniques: vacuum thermal evaporation, spin coating, direct-write printing, inkjet printing, lithographic patterning, and solution deposition.

44. An electronic device comprising an LED display, wherein the display comprises one or more_light-emitting devices as recited in claim 1.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) The invention is diagrammatically illustrated, by way of example, in the accompanying drawings, in which:

(2) FIG. 1a shows a device architecture comprising CH.sub.3NH.sub.3PbI.sub.3-xCl.sub.x perovskite, according to a first embodiment of the invention;

(3) FIG. 1b shows a single unit-cell of an ABX.sub.3 perovskite crystal;

(4) FIG. 1c shows an energy level diagram of different layers of materials in an infra-red perovskite LED (PeLED);

(5) FIG. 1d illustrates absorbance, electroluminescence and photoluminescence spectra of CH.sub.3NH.sub.3PbI.sub.3-xCl.sub.x perovskite;

(6) FIG. 1e shows photothermal deflection spectroscopy (PDS) of CH.sub.3NH.sub.3PbI.sub.3-xCl.sub.x and CH.sub.3NH.sub.3PbBr.sub.3 perovskite thin films on fused silica substrate, showing the clear absorption edges of the materials;

(7) FIG. 2a shows a device architecture comprising CH.sub.3NH.sub.3PbBr.sub.3 perovskite, according to a further embodiment of the invention;

(8) FIG. 2b depicts electroluminescence, photoluminescence and absorbance spectra of CH.sub.3NH.sub.3PbBr.sub.3 perovskite on a PEDOT:PSS covered substrate;

(9) FIG. 2c shows, in the top graph, the combined luminance and current density vs. voltage characteristics of a green PeLED and, in the bottom graph, the external quantum efficiency vs. voltage of green PeLED;

(10) FIG. 2d shows the external quantum efficiency vs. current density of a green PeLED;

(11) FIG. 3a shows a device architecture comprising a CH.sub.3NH.sub.3PbI.sub.3-xCl.sub.x perovskite layer, according to a further embodiment of the invention;

(12) FIG. 3b shows the electroluminescence of the device of FIG. 3a;

(13) FIG. 4 shows how brightness and quantum efficiency of the light-emitting devices decreases with increasing perovskite thickness;

(14) FIG. 5 shows the optical absorption of the perovskite with different compositions of bromide and iodide;

(15) FIG. 6 shows the photoluminescence spectra of mixed and pure halide perovskite samples;

(16) FIG. 7a depicts two graphs, the top graph showing combined radiance and current density against voltage characteristics of an infrared PeLED, the bottom graph showing external quantum efficiency against voltage of PeLED;

(17) FIG. 7b shows a graph of radiance against current density of PeLED under steady state and pulse measurement conditions;

(18) FIG. 7c shows a graph of photoluminescence quantum efficiency (PLQE) of TiO.sub.2/CH.sub.3NH.sub.3PbI.sub.3-xCl.sub.xF8 film as a function of absorbed photon flux;

(19) FIG. 8 shows plots of radiance, current density and external quantum efficiency vs. time (or pulse width) for the CH.sub.3NH.sub.3PbI.sub.3-xCl.sub.x PeLED at a driving voltage of 7 V;

(20) FIGS. 9a and 9b show changes in radiance, current density, and EQE when applying a fixed +5 V square voltage pulse with a width of 5 ms and a frequency of 20 Hz to a CH.sub.3NH.sub.3PbI.sub.3-xCl.sub.xPeLED, when respectively, the background voltage was varied from (a) 0 to +4 V and from (b) 0 to −6 V;

(21) FIG. 10 shows hysteresis effects in radiance, current density and external quantum efficiency during current-voltage measurements;

(22) FIG. 11 depicts a graph showing combined plots of radiance vs. voltage (solid lines) and EQE vs. voltage (dash lines) of CH.sub.3NH.sub.3PbI.sub.3-xCl.sub.x PeLED devices with different perovskite layer thicknesses;

(23) FIG. 12 shows an Atomic Force Microscopy (AFM) image of a CH.sub.3NH.sub.3PbI.sub.3-xCl.sub.x perovskite thin film;

(24) FIG. 13 shows a graph of combined photon flux per steradian, and current density vs. voltage characteristics of TiO.sub.2/F8 and TiO.sub.2/CH.sub.3NH.sub.3PbI.sub.3-xCl.sub.x/F8 devices;

(25) FIG. 14 shows an AFM image of the CH.sub.3NH.sub.3PbBr.sub.3 perovskite thin film;

(26) FIG. 15a shows the combined luminance and current density vs. voltage characteristics of ITO/PEDOT:PSS/CH.sub.3NH.sub.3PbBr.sub.2I/F8/Ca/Ag red PeLED;

(27) FIG. 15b shows the EQE vs. voltage of a red PeLED;

(28) FIG. 16a shows a variation of the device of FIG. 1a comprising an aluminium oxide layer;

(29) FIGS. 16b and 16c show, respectively, the current density and radiance versus applied bias voltage for devices with an ITO/TiO.sub.2/Al.sub.2O.sub.3/CH.sub.3NH.sub.3PbI.sub.3-xCl.sub.x/F8/MoO.sub.3/Ag structures;

(30) FIGS. 16d and 16e show, respectively, the external and calculated internal quantum efficiency versus applied bias voltage for devices comprising a metal oxide layer;

(31) FIGS. 16f and 16g show, respectively, the radiance and calculated internal quantum efficiency versus current density for a device comprising a metal oxide layer;

(32) FIG. 16h shows electroluminescence versus wavelength for a device comprising a metal oxide layer;

(33) FIG. 17a is a plot of luminance versus voltage characteristics for ITO/ZnO & TiO.sub.2/Al.sub.2O.sub.3/perovskite (CH.sub.3NH.sub.3PbBr.sub.3)/F8/MoO.sub.3/Ag devices;

(34) FIG. 17b is a plot of electroluminescence versus wavelength for a ITO/ZnO/Al.sub.2O.sub.3/perovskite (CH.sub.3NH.sub.3PbBr.sub.3)/F8/MoO.sub.3/Ag device;

(35) FIG. 18 is a plot of electroluminescence versus wavelength for a ITO/ZnO/PEI/CH.sub.3NH.sub.3PbBr.sub.1.2Cl.sub.1.8/CBP/MoO.sub.3/Au device and a ITO/ZnO/PEI/CH.sub.3NH.sub.3PbCl.sub.3/CBP/MoO.sub.3/Au device;

(36) FIG. 19a is a plot of electroluminescence versus wavelength for a ITO/PEDOT:PSS/CH.sub.3NH.sub.3Pb(Br.sub.0.4Cl.sub.0.6).sub.3/CBP/Ca/Ag device; and

(37) FIG. 19b is a plot of absorbance against energy (eV) for each of a series of devices comprising the architecture ITO/PEDOT:PSS/peroskite/CBP/Ca/Ag device wherein the thin perovskite films have the formula CH.sub.3NH.sub.3PbBr.sub.3-xCl.sub.x and the bromide content varies between 0% and 100%.

DETAILED DESCRIPTION OF THE INVENTION

(38) Broadly speaking, embodiments of the invention relate to perovskite-based light-emitting devices (LEDs), which may be formed using different architectures/device structures.

(39) Example Structure 1

(40) FIG. 1a shows device architecture 10 according to an embodiment of the invention. The PeLED is fabricated on a transparent substrate 12, which may be a glass substrate. Generally speaking, the PeLED illustrated here comprises a first electrode 14 coupled to a first charge injecting layer 16, a second electrode 22 coupled to a second charge injecting layer 20, and an active layer 18 formed of an organometal halide perovskite material. The active layer 18 is sandwiched between the first and second charge injecting layers 16, 20. One or both of the charge injecting layers 16, 20 may be formed of a semiconductor material, such as titanium dioxide (TiO.sub.2). The active perovskite layer may be formed of CH.sub.3NH.sub.3PbI.sub.3-xCl.sub.x perovskite. Advantageously, the double heterostructure architecture shown here enables the injected charges (holes and electrons) to be confined within the active (emissive) perovskite layer. This may be achieved by forming the PeLED using charge injecting layers 16, 20 from a material that has a large bandgap relative to the bandgap of the perovskite layer, separately chosen for electron and hole confinement. This confinement within the active layer may be improved by use of a thin active perovskite layer, e.g. having a thickness of less than 100 nm (or <60 nm, or even <20 nm). Large bandgap semiconductor materials enables spatial confinement of the electrons and holes in the active layer, and this is enhanced further by use of a thin perovskite layer, resulting in radiative recombination and light emission.

(41) Preferably, the first electrode 14 may be formed of a transparent, conductive material, such as indium tin oxide (ITO). One or both of the first and second charge injecting layers 16, 20 may be formed from a semiconductor, such as titanium dioxide (TiO.sub.2), a polymer material such as poly(9,9′-dioctylfluorene) (F8). One or both of the first and second charge injecting layers 16, 20 may actively inject charge into the active perovskite layer 18, or one or both of layers 16, 20 may simply act as a charge transport and charge blocking layer. The second electrode 22 may be formed of any conductive material such as, but not limited to MoO.sub.3/Ag, Ca/Ag, Au etc.

(42) For illustrative purposes, device 10 may be formed of a simple 3-layered structure of TiO.sub.2/CH.sub.3NH.sub.3PbI.sub.3-xCl.sub.x/F8, sandwiched between indium tin oxide (ITO) and MoO.sub.3/Ag as the first electrode 14 (cathode) and second electrode (anode) 22 respectively. Here, ITO is the cathode and thus, the first charge injecting layer 16 (titanium oxide) transports electrons and injects them into the active perovskite layer 18, while MoO.sub.3/Ag is the cathode and the second charge injecting layer 20 (F8) transports the holes and injects them into the active layer 18. In this example, the first and second charge injecting layers 16, 20 are formed from large bandgap semiconductors. The large bandgaps of the semiconductors (relative to the perovskite layer bandgap), enable confinement of the injected charges in the perovskite layer for better light emission. The first charge transport and injecting layer 16 is formed of a thin (25 nm) layer of titanium dioxide (TiO.sub.2) that is deposited using atomic layer deposition (ALD) onto an ITO electrode 14 which forms a coating on glass substrate 12. In this case, the TiO.sub.2 layer 16 serves as an efficient electron injector, as well as a hole blocking layer for the perovskite device.

(43) A thin layer of perovskite precursor was deposited on the TiO.sub.2 by spin coating. Annealing at 100° C. converts the precursor into a 15 nm CH.sub.3NH.sub.3PbI.sub.3-xCl.sub.x perovskite thin film. The perovskite layer 18 was designed to be thin (e.g. <100 nm) in order to spatially confine electrons and holes for radiative recombination. Given the small exciton binding energy in this class of materials, the confinement of injected charges within a thin “well” in the perovskite layer enhances electron-hole capture and improves radiative recombination. The perovskite film was capped with the second charge injecting layer 20, which was a 50 nm layer of F8 polymer.

(44) The F8 polymer layer 20 has a deep ionization potential and a shallow electron affinity, which advantageously results in the formation of a type-1 heterojunction with the perovskite layer 18. Thus, the F8 material confines holes within the perovskite well, and blocks electrons from exiting via the anode (i.e. the second electrode 22). A high work function MoO.sub.3/Ag anode was used to provide ohmic hole injection into the F8 polymer.

(45) Thus, in more detail, ITO coated glass substrates 12 were cleaned successively with acetone and isopropanol. TiO.sub.2 (25 nm) was grown onto the cleaned substrates at 225° C. with atomic layer deposition (ALD), using titanium tetrachloride (TiCl.sub.4) and water as precursors. The substrates were transferred into a nitrogen filled glovebox for further fabrication. The CH.sub.3NH.sub.3PbI.sub.3Cl.sub.x perovskite precursor solution was spin coated onto the TiO.sub.2 at 3000 rpm for 60 seconds, and annealed at 100° C. for 5 minutes to give a thin perovskite film with an average thickness of ˜15 nm, as determined by atomic force microscopy (AFM). A solution of F8 in chlorobenzene (10 mg/mL) was spin coated onto the perovskite layer at 3000 rpm for 60 seconds to give a 50 nm film. MoO.sub.3 (5 nm) and Ag (100 nm) were successively deposited by vacuum thermal evaporation.

(46) FIG. 1b shows a single unit-cell of an ABX.sub.3 perovskite crystal. In embodiments, A may be methylammonium, B may be lead, and X may be I, Br or Cl.

(47) FIG. 1c shows an energy level diagram of different layers of materials in an infra-red PeLED formed using the device structure illustrated in FIG. 1a. In particular, the diagram shows the conduction and valence band levels with respect to vacuum. This demonstrates the higher bandgaps for TiO.sub.2 and F8 than for the perovskite layer.

(48) Turning now to FIG. 1d, this shows a graph of the absorbance, electroluminescence and photoluminescence spectra of CH.sub.3NH.sub.3PbI.sub.3-xCl.sub.x perovskite. Absorption onset occurs at approximately 760 nm, consistent with previous reports such as that of M. M. Lee et al (ibid). The non-zero baseline in the absorption spectra can be attributed to light scattering and interference effects, a clear absorption edge was measured in photothermal deflection spectroscopy (PDS), which is a technique immune to optical scattering artifacts (see FIG. 1e). A strong near-infrared photoluminescence (PL), centered at 773 nm, was measured when the perovskite thin film was excited with a 532 nm green cw-laser. Using the integrating sphere method (described in M. Era, Applied Physics Letters 1994, 65, 676) a relatively high photoluminescence quantum efficiency (PLQE) of 26% was measured. The electroluminescence (EL) of the infrared PeLED is slightly blue shifted from the PL, peaking at 754 nm. The emission band is narrow with a full width at half maximum (FWHM) of 35 nm. No electroluminescence was observed from the F8 polymer, indicating that F8 serves only as a hole transporting and an electron blocking layer, and does not participate in light emission.

(49) FIG. 7a shows the current density vs. voltage characteristics and the corresponding radiance of the infrared PeLED of FIG. 1a. A clear turn-on of light emission was observed at 1.5 V, close to the photon emission energy. For the best performing device, a radiance of 6.8 W sr.sup.−1 m.sup.−2 was achieved at a driving voltage of 6.2 V, where the corresponding current density was 605 mA cm.sup.−2. This is comparable to some of the best colloidal quantum dot infrared light-emitting devices (see for example, N. Tessler et al, Science 2002, 295, 1506, and L. Sun et al, Nat Nano 2012, 7, 369). The highest external quantum efficiency (EQE) of 0.23% was achieved at 494 mA cm.sup.−2, calculated assuming a Lambertian emission profile. This gives an internal quantum efficiency (IQE) of 1.0%, calculated using the relation IQE=2n.sup.2 EQE (see N. C. Greenham et al, Advanced Materials 1994, 6, 491). We selected the refractive index of glass (n=1.5) to estimate IQE because we consider light to be emitted isotropically into the glass substrate, given the thinness (<<A) of our emissive perovskite layer. The EQE rises with increasing voltage and current density, indicating that a high density of charges is required for efficient radiative recombination.

(50) To investigate the reasons for the decline in radiance and efficiency beyond 600 mA cm.sup.−2, pulse voltage measurements were adopted where a bias as high as 14 V was applied to drive currents up to 1500 mA cm.sup.−2 (see FIG. 7b). Using 14 V square voltage pulses at a width of 1 ms and frequency of 100 Hz, a two-times higher radiance of 13.1 W sr.sup.−1 m.sup.−2 was achieved at a current density of 1467 mA cm.sup.−2. This indicates that the device degradation and efficiency drop-off are driven by heating at high current densities since a higher radiance could be achieved when the device is given time to cool in between the pulses.

(51) Interestingly, during the pulse experiments it was observed that the radiance and quantum efficiency of the PeLED increase across the duration of the voltage pulse, while the current density remains approximately constant. FIG. 8 shows plots of radiance, current density and external quantum efficiency vs. time (or pulse width) for the CH.sub.3NH.sub.3PbI.sub.3-xCl.sub.x PeLED at a driving voltage of 7 V. In the single pulse width experiment, 25 ms square voltage pulses with a period of 250 ms were applied and the radiance and current density were monitored over the duration of the 25 ms pulse. For the multiple pulse width experiment, square pulses with widths ranging from 50 μs to 25 ms were applied to the device, all with periods of 10× pulse width. Both experiments show that radiance and EQE increase with the pulse width while current density remains approximately constant.

(52) FIGS. 9a and 9b demonstrate that this increase in light-emitting efficiency is related to the history of electric field polarization across the device. In the experiments, a device which is pre-polarized with an external bias emits at a higher radiance and efficiency. Over the duration of the voltage pulse, the perovskite device becomes increasingly polarized, hence giving higher electroluminescence with longer pulse durations.

(53) Specifically, the data shown in FIGS. 9a and 9b was obtained from an experiment in which a fixed +5 V square voltage pulse with a width of 5 ms and a frequency of 20 Hz was applied to the CH.sub.3NH.sub.3PbI.sub.3-xCl.sub.x PeLED. The background voltage was varied from (a) 0 to +4 V and from (b) 0 to −6 V, and the resulting changes in radiance, current density and EQE were monitored over the duration of the pulse. The profiles of the applied voltage are presented in the top-most graph in each Figure.

(54) When the background voltage was increased from 0 to +4 V, an increase in radiance and EQE and a slight decrease in current density were observed over the 5 ms pulse. When the background voltage was decreased into the reverse bias from 0 to −6 V, the radiance and EQE were measured to be higher at the start of the pulse, and gradually decreased over the duration of the pulse. The pulse current density was slightly higher when the background voltage was at reverse bias.

(55) From these results, it is clear that the perovskite device emits more strongly when pre-polarized with an electrical bias. The application of both forward and reverse bias in the background triggers an increase in light emission during the fixed +5 V pulse. When the background voltage is 0 V, the radiance increases over the duration of the pulse as a result of increasing polarization over time. At +4 V background, the perovskite is already pre-polarized, hence yielding a higher emission from the beginning of the pulse. At a negative background voltage, the perovskite starts off being polarized and is hence more emissive, but the emission gradually decays as the positive pulse voltage removes the polarization and eventually causes a reversal in the polarization direction.

(56) It is believed that the polarization at such slow timescale is a result of ionic trap (or defect) migration across the active layer; a pre-polarization voltage may drive the ionic traps out of the bulk emissive layer, therefore suppressing non-radiative trap-mediated recombination and enhancing the radiative bimolecular recombination. This model is also consistent with the observed changes in current density due to device polarization. The ionic traps migrate in a direction that partially screens the applied background voltage. Hence, at a positive background bias, the screening causes the internal field during the +5 V pulse to be slightly diminished, giving a smaller current density. At a negative background bias, the field is enhanced during the +5 V pulse, giving a higher current density.

(57) This polarization effect is also reflected in the hysteretic current-voltage characteristics of the PeLED, as shown in FIG. 10. When the device is swept from −4 V (reverse bias) to +4 V (forward bias), the radiance and EQE are lower but the current density is higher, as compared to a voltage sweep from +4 V to −4 V. This is due to the polarization history of the device during the voltage sweep where a history of higher polarization (i.e. +4 V decreasing to −4 V) leads to higher radiance and EQE and a smaller current density, consistent with observations in the pulse experiments. It is postulated that a high polarization voltage drives the ionic traps out of the bulk emissive layer, therefore suppressing non-radiative trap-mediated recombination and enhancing the radiative bimolecular recombination.

(58) In the experiments, an increase in EL quantum efficiency was observed with both increasing voltage and increasing current density (see FIGS. 7a and 7c). In order to investigate the relation of radiative recombination with respect to charge (or excitation) density, photoluminescence quantum efficiency (PLQE) as a function of laser excitation flux was measured (see FIG. 7c). A thin film structure of TiO.sub.2/CH.sub.3NH.sub.3PbI.sub.3-xCl.sub.x/F8 that is identical to the earlier devices for measurement was fabricated. The PLQE rises with laser photon flux in a trend that is similar to the rise of device quantum efficiency with current density, confirming that high charge densities are essential for efficient radiative recombination. In FIG. 7c, IQE vs. current density was compared in the same plot as PLQE vs. absorbed photon flux and a good agreement was found between the quantum efficiencies obtained via laser excitation and electrical injection. The extrapolation of the IQE suggests that higher electroluminescence quantum efficiency may be achievable at higher current densities. It is notable that the quantum efficiency of the pulsed device also exhibits the same linear increase with current density, although efficiencies are lower compared to the steady state measurements due to the relatively short 1 ms pulse width used (see FIG. 8). The lower device EL quantum efficiencies compared to PLQE suggest non-radiative losses by electrical injection, which may be reduced by the optimization of device fabrication process if these losses are predominantly due to high leakage currents.

(59) The applicants have previously demonstrated that radiative recombination in CH.sub.3NH.sub.3PbI.sub.3-xCl.sub.x perovskites is bimolecular in nature (F. Deschler et al, The Journal of Physical Chemistry Letters 2014). The need for high excitation densities for efficient radiative recombination suggests the presence of a competing non-radiative pathway. Since bimolecular recombination kinetics follow an n.sup.2 relationship (where n=excitation density), a radiative bimolecular pathway can dominate at higher charge densities. With the fluxes produced in the devices of the present invention, charge densities are relatively low and the competing non-radiative channels dominate, giving rise to a modest EL quantum efficiency.

(60) It is interesting to note that a high PLQE of 10% can be achieved in a TiO.sub.2/CH.sub.3NH.sub.3PbI.sub.3-xCl.sub.x/F8 film structure. This is compared to a PLQE of 26% for a pristine perovskite film of equivalent thickness (15 nm) at an equivalent laser excitation flux. This appears to confirm the formation of a charge confining well structure in the devices of the present invention, where excitations are minimally quenched at the two heterojunctions, despite an emissive layer thickness of only 15 nm. To further demonstrate the effectiveness of our thin well structure, devices using thicker perovskite films were fabricated (see FIG. 11). The radiance and quantum efficiency of the devices decrease markedly at larger perovskite thicknesses, demonstrating the need for spatial confinement of charges to ensure a high rate of electron-hole capture and radiative recombination. In addition, reabsorption losses are likely to be lower in the thinner films. It should be noted that these perovskite films are annealed for a longer time (30 min) to ensure the complete formation of perovskites for the thicker layers.

(61) These findings so far suggest that higher charge densities, higher polarization and thinner emission layers are possible routes towards enhancing the electroluminescence efficiency. FIG. 12 shows an Atomic Force Microscopy (AFM) image of the CH.sub.3NH.sub.3PbI.sub.3-xCl.sub.x perovskite thin film, showing incomplete film coverage over a TiO.sub.2 coated substrate. The AFM image shows that the surface coverage of the CH.sub.3NH.sub.3PbI.sub.3-xCl.sub.x perovskite is incomplete (˜30% voids) in such thin layers, causing possible contact between the TiO.sub.2 and F8 layers.

(62) In order to investigate the effects of electrical shunts at TiO.sub.2/F8 interfaces, a TiO.sub.2/F8 device was fabricated without the thin perovskite interlayer. FIG. 13 shows a graph of combined photon flux per steradian (dash lines) and current density (solid lines) vs. voltage characteristics of TiO.sub.2/F8 and TiO.sub.2/CH.sub.3NH.sub.3PbI.sub.3-xCl.sub.x/F8 devices. As shown, TiO.sub.2/F8 device gives higher current density compared to the perovskite device. Electroluminescence in TiO.sub.2/F8 device is negligible. The current density of the TiO.sub.2/F8 device was found to be several times higher than the perovskite device at forward bias, indicating that there are possible current losses at the TiO.sub.2/F8 interface that do not contribute to electroluminescence. Optimization of thin film formation and morphology towards complete perovskite coverage is therefore expected to enhance device radiance and quantum efficiency. It is further noted that the EL photon flux from the F8-only device is 3 orders of magnitude lower than the perovskite device, confirming that the F8 layer does not participate in light emission but serves only as a hole transport and electron blocking layer.

(63) To investigate whether the efficiency of the PeLED of FIG. 1a could be improved, a device comprising a metal oxide interlayer was fabricated. As shown in FIG. 16a, device 10 may further comprise a thin insulating layer of metal oxide 52 provided between the first charge injecting layer 16 and the active perovskite layer 18. Without wishing to be bound by theory, it is believed that the thin insulating layer 52 (in this case Al.sub.2O.sub.3) functions to improve the quality of perovskite film formation and to modify the work function of the charge injecting layer for favourable electron injection. The insulating layer may also suppress luminescence quenching of the perovskite near the charge injection layer. The thin insulating layer 52 may be formed of aluminium oxide, silicon dioxide, zinc oxide, nickel oxide or silicon nitride. Furthermore, a layer of a hole-injecting agent 54 may be provided between the second charge injecting layer 20 and the second electrode 22 (and/or between the active perovskite layer 18 and the first charge injecting layer 16). The hole-injecting agent 54 is typically a thin layer (e.g. <30 nm thick) of a material selected from molybdenum trioxide and tungsten trioxide, and may act to increase the hole injection efficiency into the active perovskite layer 16. Thus, FIG. 16a shows a device with an ITO/TiO.sub.2/Al.sub.2O.sub.3/CH.sub.3NH.sub.3PbI.sub.3-xCl.sub.x/F8/MoO.sub.3/Ag structure.

(64) To demonstrate the effect of a thin insulating layer of metal oxide on electroluminescence, devices were fabricated using different thickness of the first charge injecting layer 16 (which in this case comprises titanium oxide), and a device was fabricated with an additional metal oxide layer, i.e. thin insulating layer 52, deposited over the first charge injecting layer 16. In this case, the thin insulating layer 52 comprises aluminum oxide (Al.sub.2O.sub.3). FIGS. 16b and 16c show, respectively, the current density and radiance versus applied bias voltage for these devices. Each layer may be deposited using the fabrication techniques described in detail below. As shown in FIG. 16c, the brightness of the device comprising the additional Al.sub.2O.sub.3 interlayer 52 is higher than those devices comprising the titanium dioxide charge injecting layers of different thicknesses.

(65) FIGS. 16d and 16e show, respectively, the external and calculated internal quantum efficiency versus applied bias voltage for a device comprising a thin insulating layer of metal oxide compared to devices with varying thicknesses of titanium dioxide (and no additional thin insulating layer). As shown, the EQE (0.76%) of the device comprising the additional Al.sub.2O.sub.3 interlayer 52 rises with increasing voltage, while the internal quantum efficiency (IQE) is 3.43%. Compared with the EQE and IQE of a device without the additional thin insulating layer (see FIG. 7a and corresponding description above), this indicates that the additional thin insulating layer is required for efficient radiative recombination.

(66) FIGS. 16f and 16g show, respectively, the radiance and calculated internal quantum efficiency versus current density for a device comprising a metal oxide thin insulating layer compared to devices with varying thicknesses of titanium dioxide (and no additional thin insulating layer). The device comprising the Al.sub.2O.sub.3 thin insulating layer exhibits a much higher radiance at a current density of 363 mA cm.sup.−2 than the devices comprising no additional Al.sub.2O.sub.3 thin insulating layer (which also show a decline in radiance beyond around 600 mA cm.sup.−2). Device degradation and efficiency drop-off may be driven by heating at high current densities. However, it is clear from FIG. 16h, which shows electroluminescence versus wavelength for these devices, that a markedly higher electroluminescence may be achievable by devices including the thin insulating layer.

(67) In a further example of a device to show the effect of incorporating a thin film of an insulating layer on electroluminescence, a device with the same architecture as FIG. 16 was prepared. A first charge injecting layer 16 (a 25 nm electron injecting zinc oxide layer) was deposited on a glass/ITO substrate 14. A 1 nm thin insulating layer 52 of aluminium oxide was deposited over the first charge injecting layer 16. A CH.sub.3NH.sub.3PbBr.sub.3 perovskite layer 18 was deposited on the thin insulating layer 52. A hole injecting layer 20 comprising F8 was deposited on the perovskite layer 18 followed by a thin layer of molybdenum trioxide 54 of less than 20 nm. Finally, a silver electrode 22 was deposited on the molybdenum trioxide layer 54. Two devices were prepared in this manner, one in which the perovskite layer was annealed for 15 minutes and one where it was annealed for 40 minutes before deposition of the hole injecting layer 20 thereon.

(68) Two further equivalent devices were prepared which were identical except that in place of the 1 nm aluminium oxide on the zinc oxide electron injecting layer 16, the two devices contained a 1 nm aluminium oxide on a titanium oxide electron injecting layer 16.

(69) FIG. 17a is a plot of luminance versus voltage characteristics for the ITO/ZnO & TiO.sub.2/Al.sub.2O.sub.3/perovskite/F8/MoO.sub.3/Ag devices. It can be seen that as for the device described above, the inclusion of the thin insulating layer of aluminium oxide appears to function to improve the quality of perovskite film formation and to modify the work function of the charge injecting layer for favourable electron injection. The insulating layer may also suppress luminescence quenching of the perovskite near the charge injection layer.

(70) As shown in FIG. 17b, the device ITO/ZnO/Al.sub.2O.sub.3/perovskite/F8/MoO.sub.3/Ag device in which the perovskite layer was annealed for 40 minutes, achieves electroluminescence at around 530 nm. The electroluminescence peak is a sharp, strong peak.

(71) Example Structure 2

(72) In order to demonstrate the application of organometal halide perovskite as visible light emitters, the larger bandgap CH.sub.3NH.sub.3PbBr.sub.3 perovskite material was used as a green emitter in a PeLED device of the present invention. Since a larger bandgap makes it more difficult to achieve electron injection from TiO.sub.2 into the perovskite conduction band, an inverted device structure (relative to that shown in FIG. 1a) was used.

(73) FIG. 2a shows an example of an inverted device architecture 24. Here, the PeLED is fabricated on a transparent substrate 26, which may be a glass substrate. Generally speaking, the PeLED illustrated here comprises a first electrode 28 coupled to a first charge injecting layer 30, a second electrode 36 coupled to a second charge injecting layer 34, and an active layer 32 formed of an organometal halide perovskite material. The active layer 32 is sandwiched between the first and second charge injecting layers 30, 34. One or both of the charge injecting layers 30, 34 may be formed of a semiconductor material.

(74) Preferably, the first electrode 28 may be formed of a transparent, conductive material, such as indium tin oxide (ITO). One or both of the first and second charge injecting layers 30, 34 may actively inject charge into the active perovskite layer 32, or one or both of layers 30, 34 may simply act as a charge transport and charge blocking layer.

(75) The second electrode 36 may be formed of any conductive material such as, but not limited to MoO.sub.3/Ag, Ca/Ag, Au etc.

(76) For illustrative purposes, the visible light emitting PeLED device 24 is fabricated using an ITO/PEDOT:PSS/CH.sub.3NH.sub.3PbBr.sub.3/F8/Ca/Ag architecture. Thus, here, the first charge injecting layer 30 is formed from a conductive polymer material, PEDOT:PSS, and the second charge injecting layer 34 is formed from F8. In this structure, the injected charges are confined within the active layer 32 by providing the charge injecting layers 30, 34 from a material that has a large bandgap relative to the bandgap of the perovskite layer. Additionally, confinement of the injected charges (holes and electrons) within the active (emissive) perovskite layer may be improved further by forming the PeLED with a thin active perovskite layer, e.g. having a thickness of less than 100 nm (or <60 nm, or even <20 nm).

(77) Here, ITO is the anode (electrode 28) and is coupled to the first charge injecting layer 30, which may be formed from PEDOT:PSS and is used to inject holes into the active perovskite layer 32. The second electrode 36 is the cathode and may be formed from silver/calcium. The second electrode 36 is coupled to the second charge injecting layer 34, which may be formed from a poly(9,9′-dioctylfluorene) (F8) polymer semiconductor material. The F8 material serves as a spacer layer for electron transport and to prevent emission quenching near the interface with the second electrode 36 (rather than actively injecting charge into the perovskite layer). Thus, the width of the spacer layer may be chosen such that electrons are injected into the active perovskite layer 32, and in the example described here, the F8 layer has a thickness of 50 nm. In order to spatially confine injected charges for efficient radiative recombination, the active perovskite layer 32 (formed from CH.sub.3NH.sub.3PbBr.sub.3) was deposited as a film of thickness 20 nm.

(78) FIG. 14 shows the film morphology of the perovskite thin film. The AFM image of the CH.sub.3NH.sub.3PbBr.sub.3 perovskite thin film shows incomplete film coverage over a PEDOT:PSS coated substrate, and specifically shows an extended perovskite film with crystal clusters which are surrounded by voids.

(79) To fabricate the structure shown in FIG. 2a, ITO coated glass electrodes 28 were cleaned successively with acetone and isopropanol, followed by 10 minutes oxygen plasma treatment. PEDOT:PSS (Clevios P VP AI 4083) was spin coated onto the substrate at 6000 rpm for 30 seconds, and annealed at 140° C. for 30 minutes in a nitrogen atmosphere. The CH.sub.3NH.sub.3PbBr.sub.3 perovskite precursor solution was spin coated onto PEDOT:PSS at 3000 rpm for 30 seconds and annealed at 100° C. for 15 minutes to give a perovskite film thickness of −20 nm. A solution of F8 in chlorobenzene (10 mg/mL) was spin coated onto the perovskite layer at 3000 rpm for 30 seconds to give a 50 nm film. Ca (20 nm) and Ag (100 nm) were successively deposited by vacuum thermal evaporation. The device thus obtained was tested without encapsulation.

(80) As shown in FIG. 2b, bright green electroluminescence was achieved at 517 nm. The CH.sub.3NH.sub.3PbBr.sub.3 perovskite PL is slightly red-shifted from the EL, peaking at 524 nm with a PLQE of 7% when excited with a 405 nm blue cw-laser. The device turns on at 3.3 V and reaches a maximum luminance of 364 cd m.sup.−2 at a current density of 123 mA cm.sup.−2, as shown in FIG. 2c. This gives an efficiency of 0.3 cd A.sup.−1, or an EQE of 0.1% and IQE of 0.4% assuming a Lambertian emission profile. Similar to the infrared device, the EL quantum efficiency increases with injection current density (see FIG. 2d), demonstrating a need for high charge densities to achieve efficient radiative recombination.

(81) Red electroluminescence at 630 nm was also demonstrated using a CH.sub.3NH.sub.3PbBr.sub.2I mixed halide perovskite in the same device architecture, showing the versatility and wide bandgap tunability of these organometal halide perovskite materials. FIG. 15a shows the combined luminance and current density vs. voltage characteristics of ITO/PEDOT:PSS/CH.sub.3NH.sub.3PbBr.sub.2I/F8/Ca/Ag red PeLED. A maximum luminance of 16.2 cd m.sup.−2 was achieved at a current density of 55 mA cm.sup.−2. FIG. 15b shows the EQE vs. voltage of red PeLED. Peak EQE of 0.018% or 0.03 cd A.sup.−1 was achieved at 5.7 V.

(82) Example Structure 3

(83) FIG. 3a shows device architecture 38 according to an embodiment of the invention. The PeLED is fabricated on a transparent substrate 40, which may be a glass substrate. Generally speaking, the PeLED illustrated here comprises a first electrode 42 coupled to a charge injecting layer 44, a second electrode 50, and an active (emissive) layer 48 formed of an organometal halide perovskite material. The active layer 48 is provided between the charge injecting layer 44 and the second electrode 50. In this example device structure, the second electrode 50 directly injects charges into the active perovskite layer 48, and thus can be considered a charge injecting layer. The second electrode 50 may be formed from a conductive material such as Au. The charge injecting layer 44 may be formed from of a semiconductor material, such as titanium dioxide (TiO.sub.2) which injects electrons from the first electrode 44 (the anode) into the active perovskite layer. The active perovskite layer may be formed of CH.sub.3NH.sub.3PbI.sub.3-xCl.sub.x material.

(84) Advantageously, the architecture shown here enables the injected charges (holes and electrons) to be confined within the active (emissive) perovskite layer, by forming the PeLED using charge injecting layers 44, 50 from materials that have a large bandgap relative to the bandgap of the perovskite layer. Additionally, confinement of the injected charges (holes and electrons) within the active (emissive) perovskite layer may be improved further by forming the PeLED with a thin active perovskite layer, e.g. having a thickness of less than 100 nm (or <60 nm, or even <20 nm).

(85) Here, ITO is the anode (electrode 42) and is coupled to the first charge injecting layer 44, which may be formed from titanium dioxide or a similar conductive material. The anode 42 generates electrons which pass into the charge injecting layer 44 which injects the electrons into the active perovskite layer 48. The second electrode 50 (i.e. the second charge injecting layer) directly injects holes into the active perovskite layer 48. The second electrode 50 may be formed from Au.

(86) In order to spatially confine injected charges within the active perovskite layer 48 for efficient radiative recombination, a further conductive layer or charge transport layer 46 may be provided between the first charge injecting layer 44 and the perovskite layer 48. The charge transport layer 46 may be formed of a semiconducting material such as poly[2-methoxy-5-(2′-ethylhexyloxy)-1,4-phenylenevinylene (MEH-PPV). Furthermore, a layer of a wetting agent (not shown in FIG. 3a) may be included either between the perovskite layer and the titanium oxide layer or, if present, between the perovskite layer and the charge transport layer 46. This is typically a thin layer (e.g. <30 nm thick) of a material selected from molybdenum trioxide and tungsten trioxide. This can act to increase the charge injection efficiency into the active perovskite layer. As shown in FIG. 3b, the device 38 achieves infrared electroluminescence at around 780 nm.

(87) The device architecture 38 may comprise an ITO/TiO.sub.2/MEH-PPV/MoO.sub.3/perovskite/Au architecture. The example device comprises an n-type semiconductor 44 coated on to a transparent conducting substrate 42, followed by a semiconducting material 46 such as poly[2-methoxy-5-(2′-ethylhexyloxy)-1,4-phenylenevinylene (MEH-PPV). A thin layer of molybdenum trioxide (MoO.sub.3) was deposited onto the semiconducting material MEH-PPV as a surface wetting layer on which, a perovskite (CH.sub.3NH.sub.3PbI.sub.3-xCl.sub.x) layer 48 was deposited. A gold (Au) electrode 50 was deposited on top of the perovskite layer for electrical contact.

(88) In more detail, ITO coated glass substrates were cleaned successively with acetone and isopropanol. TiO.sub.2 (25 nm) was grown onto the cleaned substrates at 225° C. with atomic layer deposition (ALD), using titanium tetrachloride (TiCl.sub.4) and water as precursors. The substrates were transferred into a nitrogen filled glovebox for further fabrication. A solution of MEH-PPV (poly[2-methoxy-5-(2′-ethylhexyloxy)-1,4-phenylenevinylene) in chlorobenzene was spin-coated onto the TiO.sub.2 layer. MoO.sub.3 (3 nm) was deposited onto MEH-PPV. The CH.sub.3NH.sub.3PbI.sub.3-xCl.sub.x perovskite precursor solution was spin coated onto the MoO.sub.3 at 3000 rpm for 60 seconds, and annealed at 100° C. for 45 minutes to give a thin perovskite film. Au (100 nm) was deposited by vacuum thermal evaporation.

(89) Example Structure 4

(90) Using a similar procedure to Example 1, a device with the following architecture was prepared: ITO/ZnO/PEI/perovskite/CBP/MoO.sub.3/Au. The example device comprises a zinc oxode (ZnO) electron injecting layer 16 (20 nm) coated on to a transparent ITO-coated glass conducting substrate, followed by a thin layer of the insulating polymer polyethylenimine (PEI) (<5 nm) on which, a perovskite (CH.sub.3NH.sub.3PbBr.sub.3-xCl.sub.x) layer 18 (80 nm) was deposited. A layer of the hole injecting material 4,4′-bis(N-carbazolyl)-1,1′-biphenyl (CBP) (40 nm) was deposited onto the perovskite film followed by molybdenum trioxide (15 nm) and finally a gold (Au) (150 nm) electrode was deposited for electrical contact.

(91) FIG. 18 shows electroluminescence spectra for the devices prepared in this manner. As can be seen, the device prepared using CH.sub.3NH.sub.3PbBr.sub.1.2Cl.sub.1.8 had peak luminescence at about 480 nm in the blue region, thus enabling blue LEDs to be prepared.

(92) Example Structure 5

(93) Using a similar procedure to Example 2, a device with the following architecture was prepared: ITO/PEDOT:PSS/perovskite/CBP/Ca/Ag. The example device comprises a hole injecting layer like-PEDOT:PSS (30 nm) coated on to a transparent ITO-coated glass conducting substrate, followed by a perovskite (CH.sub.3NH.sub.3PbBr.sub.3-xCl.sub.x) layer (80 nm). A layer of electron injection material like 4,4′-bis(N-carbazolyl)-1,1′-biphenyl (CBP) (40 nm) was deposited onto the perovskite film followed by calcium (25 nm) and silver (Ag) (150 nm) electrode for electrical contact.

(94) The electroluminescence spectrum for a LED with the above device structure comprising the perovskite CH.sub.3NH.sub.3Pb(Br.sub.0.4Cl.sub.0.6).sub.3 is shown in FIG. 19a, with a peak at approximately 490 nm in the blue region of the spectrum.

(95) A series of devices were prepared of the formed above comprising a thin emissive perovskite layer having the formula CH.sub.3NH.sub.3PbBr.sub.3-xCl.sub.x in which the bromide content in the films was varied between 0% and 100%. Using photothermal deflection spectroscopy, it was possible to prepare a plot of absorbance against energy (eV) for each of the thin perovskite films. As can be seen from FIG. 19b, the materials show clear absorption edges with a shift towards blue as the percentage of bromide in the film increased.

(96) Materials and Fabrication Techniques

(97) When the perovskite materials described above are used in the manufacture of organic optoelectronic devices, they are preferably dissolved to form solutions. The solutions may be used in solution processing techniques to form the active/emissive layer of such devices. The electrodes of the PeLEDs may be deposited by thermal evaporation. The emissive layer, the hole injecting layer and/or interlayer(s) may be deposited by solution processing, e.g. spin coating. Preferred devices of the invention are also encapsulated to avoid ingress of moisture and oxygen. Conventional encapsulation techniques may be used.

(98) The PeLED substrate may be formed from any material conventionally used in the art such as glass or plastic. Optionally the substrate is pre-treated to improve adhesion thereto. Preferably the substrate is transparent. Preferably the substrate also has good barrier properties to prevent ingress of moisture or oxygen into the device.

(99) The anode may comprise any material with a workfunction suitable for injection of holes into the light emitting layer. Preferably the anode is transparent. Representative examples of materials for use as a transparent anode include indium tin oxide (ITO) and indium zinc oxide (IZO). If the anode is not required to be transparent (e.g. if the cathode is transparent or it is a bottom emitting device) then opaque conducting materials such as opaque metals may be used as the anode.

(100) The anode may comprise a single layer or may comprise more than one layer. For example, the anode may comprise a first anode layer and an auxiliary conductive layer between the anode and the hole injection layer such as a layer of organic conductive material between the anode and the hole injection layer.

(101) The anode may be deposited on the substrate by thermal evaporation. The anode may be between 10 nm to 200 nm thick.

(102) The hole injection layer may be deposited by a solution-based processing method. Any conventional solution-based processing method may be used. Representative examples of solution-based processing methods include spin coating, dip coating, slot die coating, doctor blade coating and ink-jet printing. In embodiments, spin coating is the preferred deposition method. The parameters used for spin coating the hole injection layer such as spin coating speed, acceleration and time are selected on the basis of the target thickness for the layer.

(103) After deposition, the hole injection layer is preferably annealed by heating, e.g. at 150 to 200° C. for 5 to 30 minutes in air. The thickness of the hole injection layer may be 15 to 100 nm, or between 30 to 50 nm.

(104) As mentioned above, the devices may incorporate a thin insulating layer formed between either or both of the electrodes and the light-emitting layer. The insulating layer is preferably formed of an oxide or nitride, and more preferably the insulating layer is selected from the group consisting of aluminium oxide, silicon dioxide, silicon nitride, zinc oxide, nickel oxide or magnesium oxide. The interlayer can play an important role in improving the device efficiency and the lifetime of LEDs (because exciton quenching at the interface between the charge injecting layer(s) and the emissive layer is prevented or minimised). Preferably, the insulating layer is deposited by atomic layer deposition, ALD.

(105) The light-emitting (emissive) layer may be prepared by depositing a solution as hereinbefore defined on the charge injecting layer or, when present, the interlayer. Any conventional solution-based processing method may be used. Representative examples of solution-based processing methods include spin coating, dip coating, slot die coating, doctor blade coating and ink-jet printing. In embodiments, the depositing is by spin coating. The parameters used for spin coating the light emitting layer such as spin coating speed, acceleration and time are selected on the basis of the target thickness for the light emitting layer. After depositing, the light emitting layer is preferably dried, e.g. at 100-150° C. in a glove box. The thickness of the light emitting layer may be between 50 to 350 nm or between 75 to 150 nm.

(106) The cathode may comprise any material having a workfunction allowing injection of electrons into the light-emitting layer. In order to provide efficient injection of electrons into the device, the cathode preferably has a workfunction of less than 3.5 eV, more preferably less than 3.2 eV, most preferably less than 3 eV. Work functions of metals can be found in, for example, Michaelson, J. Appl. Phys. 48(11), 4729, 1977. The cathode may consist of a single material such as a layer of gold. Alternatively, it may comprise a plurality of metals, for example a bilayer or trilayer of a low workfunction material and a high workfunction material such as calcium and aluminium as disclosed in WO 98/10621; elemental barium as disclosed in WO 98/57381, Appl. Phys. Lett. 2002, 8 1(4), 634 and WO 02/84759; or a thin layer of metal compound, in particular an oxide or fluoride of an alkali or alkali earth metal, to assist electron injection, for example lithium fluoride as disclosed in WO 00/48258; barium fluoride as disclosed in Appl. Phys. Lett. 2001, 79(5), 2001; and barium oxide.

(107) The cathode may be opaque or transparent. Transparent cathodes are particularly advantageous for active matrix devices because emission through a transparent anode in such devices is at least partially blocked by drive circuitry located underneath the emissive pixels. A transparent cathode comprises a layer of a hole injecting material that is sufficiently thin to be transparent. Typically, the lateral conductivity of this layer will be low as a result of its thinness. In this case, a layer of hole injecting material is used in combination with a thicker layer of transparent conducting material such as indium tin oxide.

(108) The cathode may be deposited by thermal evaporation. The cathode may be 100 to 400 nm thick, or between 200 to 350 nm thick.

(109) Suitable encapsulants include a sheet of glass, films having suitable barrier properties such as silicon dioxide, silicon monoxide, silicon nitride or alternating stacks of polymer and dielectric as disclosed in, for example, WO 01/81649 or an airtight container as disclosed in, for example, WO 01/19142. In the case of a transparent cathode device, a transparent encapsulating layer such as silicon monoxide or silicon dioxide may be deposited to micron levels of thickness, although in one preferred embodiment the thickness of such a layer is in the range of 20-300 nm. A material for absorption of any atmospheric moisture and/or oxygen that may permeate through the substrate or encapsulant may optionally be disposed between the substrate and the encapsulant.

(110) CH.sub.3NH.sub.3PbI.sub.3-xCl.sub.x Perovskite Synthesis.

(111) Methylammonium iodide (CH.sub.3NH.sub.3I) was prepared by adding 33 wt % methylamine solution in ethanol (24 mL) and 57 wt % hydroiodic acid in water (10 mL) to 100 mL of absolute ethanol. The reaction mixture was stirred at room temperature in a nitrogen atmosphere. The solvent was removed by rotary evaporation until white crystals started to appear. The product was collected using Büchner funnel filtration and was dried overnight under vacuum at 80° C. The mixed halide perovskite precursor solution was prepared by mixing CH.sub.3NH.sub.3I and PbCl.sub.2 in a 3:1 molar ratio in anhydrous N,N-dimethylformamide to give a concentration of 5 wt %.

(112) CH.sub.3NH.sub.3PbBr.sub.3 Perovskite Synthesis.

(113) Methylammonium bromide (CH.sub.3NH.sub.3Br) was prepared by adding 33 wt % methylamine solution in ethanol (24 mL) and 48 wt % hydrobromic acid in water (8.5 mL) to 100 mL of absolute ethanol. The reaction mixture was stirred at room temperature. The solvent was removed by rotary evaporation. The obtained white crystals were washed with anhydrous diethyl ether and recrystallized in ethanol. The perovskite precursor solution was prepared by mixing CH.sub.3NH.sub.3Br and PbBr.sub.2 in a 3:1 molar ratio in anhydrous N,N-dimethylformamide to give a concentration of 5 wt %. For the CH.sub.3NH.sub.3PbBr.sub.2I mixed-halide precursor solution, the individual CH.sub.3NH.sub.3PbBr.sub.3 and CH.sub.3NH.sub.3PbI.sub.3 precursor solutions were mixed in a molar ratio of 2:1 to give an overall concentration of 5 wt %.

(114) CH.sub.3NH.sub.3PbBr.sub.3-xCl.sub.x Perovskite Synthesis.

(115) CH.sub.3NH.sub.3PbBr.sub.3 precursor solution was synthesised by mixing CH.sub.3NH.sub.3Br and Pb(CH.sub.3COO).sub.2 in 5:1 molar stoichiometric ratio in N,N-Dimethylformamide (DMF) to get a 0.5 M solution. Similarly, the CH.sub.3NH.sub.3PbCl.sub.3 precursor solution was synthesised by mixing CH.sub.3NH.sub.3Cl and Pb(CH.sub.3COO).sub.2 in 5:1 molar stoichiometric ratio in a mixed solvent comprising of dimethyl sulfoxide (DMSO) and DMF in the ratio of 40:60 v/v to get a 0.5M solution. To make the desired CH.sub.3NH.sub.3PbBr.sub.3-xCl.sub.x [0≤x≤1] precursor solutions the above mentioned CH.sub.3NH.sub.3PbBr.sub.3 and CH.sub.3NH.sub.3PbCl.sub.3 precursor solutions were mixed together in the required stoichiometric ratios.

(116) In embodiments, the perovskite layer is thin (<100 nm) in order to achieve efficient electroluminescence. To demonstrate this, devices were fabricated using thicker perovskite films. As shown in FIG. 4, the brightness and quantum efficiency of the devices decrease markedly at larger perovskite thicknesses, demonstrating the preference for spatial confinement of charges to ensure a high rate of electron-hole capture and radiative recombination. In addition, reabsorption losses are likely to be lower in the thinner films.

(117) Preferably, the perovskite is prepared such that it consists of a single homogenous phase, which may achieve efficient emission. Photothermal deflection spectroscopy (PDS) technique was used to determine the absorption spectra of various methyl ammonium mixed halide perovskite films, and to determine the homogeneity of the perovskite layer. FIG. 5 shows the optical absorption of the perovskite with different compositions of bromide and iodide (measured using PDS). The absorption spectra shows a clean band-gap with no detectable absorption from the neat iodide (CH.sub.3NH.sub.3PbI.sub.3) phase up to the detection limit of 5 orders of magnitude below the band edge absorption, demonstrating a clean band-edge and hence a homogenous perovskite phase (i.e. no areas with more concentrated amounts of bromide or iodide).

(118) FIG. 6 shows the photoluminescence spectra of the mixed and pure halide perovskite samples. The spectra show sharp and well-defined emission peaks, even in the mixed halide perovskite samples, indicating the presence of homogenous composition throughout the perovskite layer. If more than one phase/type of the perovskite was present within the layer, there would be an appearance of two or more emission peaks.

(119) Infrared PeLED Fabrication.

(120) ITO coated glass substrates were cleaned successively with acetone and isopropanol. TiO.sub.2 (25 nm) was grown onto the cleaned substrates at 225° C. with atomic layer deposition (ALD), using titanium tetrachloride (TiCl.sub.4) and water as precursors. The substrates were transferred into a nitrogen filled glovebox for further fabrication. The CH.sub.3NH.sub.3PbI.sub.3Cl.sub.x perovskite precursor solution was spin coated onto the TiO.sub.2 at 3000 rpm for 60 seconds, and annealed at 100° C. for 5 minutes to give a thin perovskite film with an average thickness of −15 nm, as determined by atomic force microscopy (AFM). A solution of F8 in chlorobenzene (10 mg/mL) was spin coated onto the perovskite layer at 3000 rpm for 60 seconds to give a 50 nm film. MoO.sub.3 (5 nm) and Ag (100 nm) were successively deposited by vacuum thermal evaporation. Devices were tested in air without encapsulation.

(121) Green PeLED Fabrication.

(122) ITO coated glass substrates were cleaned successively with acetone and isopropanol, followed by 10 minutes oxygen plasma treatment. PEDOT:PSS (Clevios P VP AI 4083) was spin coated onto the substrate at 6000 rpm for 30 seconds, and annealed at 140° C. for 30 minutes in a nitrogen atmosphere. The CH.sub.3NH.sub.3PbBr.sub.3 perovskite precursor solution was spin coated onto PEDOT:PSS at 3000 rpm for 30 seconds and annealed at 100° C. for 15 minutes to give a perovskite film thickness of −20 nm. A solution of F8 in chlorobenzene (10 mg/mL) was spin coated onto the perovskite layer at 3000 rpm for 30 seconds to give a 50 nm film. Ca (20 nm) and Ag (100 nm) were successively deposited by vacuum thermal evaporation. Devices were tested in air without encapsulation.

(123) PeLED Characterization.

(124) Current vs. voltage characteristics were measured using a Keithley 2400 Source Measure Unit (SMU). Photon flux was measured simultaneously using a calibrated silicon photodiode centered over the light-emitting pixel. Radiance in W sr.sup.−1 m.sup.−2 and luminance in cd m.sup.−2 were calculated based on the emission function of the PeLED, and on the known spectral response of the silicon photodiode. External quantum efficiency was calculated, assuming a Lambertian emission profile. Electroluminescence spectra were measured using a Labsphere CDS-610 spectrometer.

(125) PLQE Measurement.

(126) The PLQE of the thin film samples was measured using an integrating sphere method, as reported in the literature (de Mello, J. C.; Wittmann, H. F.; Friend, R. H. Advanced Materials 1997, 9, 230). A continuous wave 532 nm green diode laser with an excitation power of 10-100 mW and a focused beam spot of ˜0.3 mm.sup.2 was used to photo-excite the samples. Emission was measured using an Andor iDus DU490A InGaAs detector. During measurements, the samples were kept at room temperature under a constant stream of nitrogen. Samples were prepared in the structure TiO.sub.2/CH.sub.3NH.sub.3PbI.sub.3-xCl.sub.x/F8 on top of a spectrosil fused silica substrate, similar to the PeLED device structure.

(127) Pulse LED Measurement.

(128) A square voltage pulse with a pulse width of 1 ms and frequency of 100 Hz was applied to the device using a HP8116A Pulse/Function generator. A dual-channel Agilent Technologies DSO6032A oscilloscope, set to 50Ω impedance, was connected in series to the device and the function generator to determine the current through the device. The electroluminescence was monitored using a calibrated Si photodiode, connected to the 2.sup.nd channel of the oscilloscope (1 MΩ) via a Thorlabs PDA200C photodiode amplifier.

SUMMARY

(129) The demonstration of bright visible and infrared electroluminescence from organometal halide based perovskite using the devices of the present invention shows great promise in the development of this class of materials for large area optoelectronic or electrically-pumped lasing applications. Furthermore, this work exemplifies the fact that an efficient solar cell material is generally also a good light emitter, as given by the Shockley-Queisser detailed balance limit calculations (W. Shockley and H. J. Queisser, Journal of Applied Physics 1961, 32, 510, and O. D. Miller et al, Photovoltaics, IEEE Journal of 2012, 2, 303).

(130) It is interesting that bright electroluminescence could be achieved using a simple thin emitter realized by solution processing. Given the versatility and low-cost, simple processability at room temperature of these organometal halide perovskite materials, they may quickly find their way into the displays and lighting industry.

(131) No doubt many other effective alternatives within the scope of the presently claimed invention will occur to the skilled person. It will be understood that the invention is not limited to the described embodiments and encompasses modifications apparent to those skilled in the art lying within the spirit and scope of the claims appended hereto.