Nanowire light emitting diode with directional emission and display comprising the same

Abstract

A full color display includes a plurality of pixels and has a white point, an emission direction, and an emission solid angle around the emission direction, the emission solid angle being characterized by a half-cone angle Θ. Each pixel includes: a sub-pixel including a red LED having a first geometry, the red LED emitting red light into the emission angle range such that a fraction of power emitted within the emission solid angle is at least 1.2*(1-cos(Θ) 2 ); a sub-pixel including a green LED having a second geometry, the green LED emitting green light into the emission angle range such that a fraction of power emitted within the emission solid angle is at least 1.2*(1-cos(Θ) 2 ); and a sub-pixel including a blue LED, the blue LED emitting blue light into the emission angle range such that a fraction of power emitted within the emission solid angle is at least 1.2*(1-cos(Θ) 2 ). The LEDs are configured such that white light emitted by the display in any direction within the emission solid angle has a color difference Du'v' from the white point of the display of less than 0.01.

Description

Technical Field

[0001] This disclosure relates to light-emitting diodes (LEDs), and more particularly to nanowire (NW) LEDs with directional emission. Background Technology

[0002] Nanowires (NWs) are elongated nanostructures with a short dimension of approximately a few nanometers to hundreds of nanometers (e.g., 1,000 nm or less) and a significantly longer length dimension extending along the axis (e.g., an aspect ratio of 5 or greater). Nanowire light-emitting diodes (NW LEDs) typically consist of many wire-shaped LEDs, typically having a diameter of approximately 10–1,000 nm and a length of approximately 100–10,000 nm. Summary of the Invention

[0003] The far-field pattern (FFP) of an LED is the intensity of light emitted at various angles, and is denoted as I(θ) in this paper. In conventional LEDs, the FFP is approximately Lambertian (i.e., I(θ) ~ cos(θ)), where light is emitted over a wide angular range. However, some applications (including displays) can benefit from directional emission, where the FFP directs more light into specific angles (e.g., a narrow angular range near the normal direction). It may be desirable to influence the FFP at the LED level (without using additional optics). Furthermore, if directionality is achieved, it may come at the cost of color uniformity across angles (since red, green, and blue pixels may have different directions of orientation), and avoiding this is desirable.

[0004] This application discloses an LED with an optical structure that promotes directional FFP, and a directional LED with good color uniformity.

[0005] The various aspects of this invention are summarized below.

[0006] Typically, in a first aspect, the invention is characterized by a full-color display comprising a plurality of pixels having a white dot, an emission direction, and an emission solid angle about the emission direction, the emission solid angle being characterized by a half-cone angle θ. Each pixel includes a red sub-pixel comprising a red LED (e.g., one or more red LEDs) having a first geometry, the red LEDs being configured to emit red light into an emission angular range such that the power fraction of the red sub-pixel emitted within the emission solid angle is at least 1.2*(1-cos(θ)). 2 The green subpixel includes a green LED (e.g., one or more green LEDs) having a second geometry, the green LEDs being configured to emit green light into an emission angular range such that the fraction of power of the green subpixel emitted within the emission solid angle is at least 1.2 * (1 - cos(θ)). 2); and a blue sub-pixel including a blue LED (e.g., one or more blue LEDs), the blue LEDs being configured to emit blue light into an emission angular range such that the power fraction of the blue sub-pixel emitted within the emission solid angle is at least 1.2*(1-cos(θ)). 2 The red, green, and blue LEDs are configured such that the white light emitted by the display in any direction within the emission solid angle has a chromaticity difference Du′v′ of less than 0.01 with the white point of the display.

[0007] Embodiments of a full-color display may include one or more of the following features and / or other features. For example, each of the blue LED, green LED, and red LED includes a light-emitting region and a reflective region located at an optical distance of less than 1 μm from the active region, such that optical interference within each LED causes light to be emitted in a preferred direction in the far field, wherein the respective optical distances between the light-emitting and reflective regions of the red LED, green LED, and blue LED are configured such that the respective preferred directions of light emitted from the red LED, green LED, and blue LED are within + / -10° of each other (e.g., within + / -5° of each other, within + / -3° of each other).

[0008] Each blue LED may include a blue emitting region that emits light at a blue wavelength, and a blue reflective region separated from the blue emitting region by a first optical distance. A green LED may include a green emitting region that emits light at a green wavelength, and a green reflective region separated from the green emitting region by a second optical distance. A red LED may include a red emitting region that emits light at a red wavelength, and a red reflective region separated from the red emitting region by a third optical distance. The ratios between the first optical distance and the blue wavelength, the second optical distance and the green wavelength, and the third optical distance and the red wavelength are all within 10% of each other (within 5%, within 3%).

[0009] Each LED can have a characteristic lateral dimension of 50 μm or less. In some embodiments, at least one of the red LED, green LED, and blue LED is a multifaceted LED, comprising a top surface and a plurality of sidewall surfaces, wherein, during operation of the multifaceted LED, light is emitted from each surface, and the cumulative surface area of ​​the sidewall surfaces is at least 20% of the surface area of ​​the top surface. At least 20% of the light emitted by the multifaceted LED can be emitted through the sidewall surfaces. Light emitted from the top surface can be preferentially emitted at a first angle relative to the top surface, and light emitted from one of the sidewalls can be preferentially emitted at a second angle relative to the top surface, wherein the first angle and the second angle are within + / -10° of each other.

[0010] In some embodiments, θ is in the range of 5° to 45°.

[0011] The red, green, and blue LEDs can each be represented by their respective shapes, so that the white light emitted by the display in any direction within the emission solid angle has a chromaticity difference of less than 0.01 Du′v′ with the white point of the display.

[0012] Red, green, and blue LEDs are characterized by their physical dimensions and emission wavelengths, respectively, with the ratio between the physical dimension and emission wavelength of each LED within + / - 10% of each other. The ratios for each LED can be approximately equal. The physical dimensions can be selected from the group consisting of: the LED's radius, height, characteristic vertical dimension, characteristic horizontal dimension, and the optical distance between the emitting area and the reflector.

[0013] In some embodiments, the white point has a chromaticity substantially equal to D65.

[0014] Each LED can include a metal-based reflector, and can be essentially based on the equation To configure, among which, θ represents the phase shift at the interface between the metal-based mirror and the adjacent layers of the LED, d is the optical thickness between the interface and the light-emitting region of the LED, θ is the angle ranging from 0° to 30°, and m is an integer ranging from 1 to 3. The optical thickness d between the interface and the light-emitting region is within 10 nm, satisfying the equation.

[0015] Typically, in another aspect, the present invention is characterized by an axially extending light-emitting element comprising: an n-side portion having one or more n-doped material layers; a p-side portion having one or more p-doped material layers; one or more light-emitting layers disposed between the n-side portion and the p-side portion, the one or more light-emitting layers being configured to emit light at a wavelength λ during operation of the light-emitting element; and a conductive layer in contact with the p-side portion, the conductive layer having a reflectivity of 80% or greater at λ. The light-emitting element has physical dimensions such that it supports at least two guiding modes for the propagation of light at λ along the axis, and the p-side portion has an axial optical thickness selected to preferentially enhance the emission of light at λ in a first mode compared to the emission of light at λ in a second mode.

[0016] Embodiments of the light-emitting element can include one or more of the following features and / or other aspects. For example, the light-emitting element can be substantially based on the equation... To configure, among which, d is the phase shift at the interface between the conductive layer and the p-side portion, d is the optical thickness of the p-side portion, θ is the angle in the range of 0° to 30°, and m is an integer in the range of 1 to 3.

[0017] The conductive layer can include silver, and the light-emitting device can be further configured such that d / λ is in the range of 0.55-0.70, where d is the optical thickness of the p-side portion.

[0018] The optical thickness of the p-side portion along the axis can be selected such that, for light at λ in the first mode, it results in constructive interference between light emitted from one or more light-emitting layers propagating toward the n-side portion and light emitted from one or more light-emitting layers propagating toward the p-side portion reflected from the conductive layer.

[0019] The light-emitting element is a nanowire light-emitting diode (NW LED).

[0020] One or more n-doped material layers can include an n-GaN layer.

[0021] One or more p-doped material layers can include p-GaN layers.

[0022] One or more n-doped material layers can include an n-TCO layer.

[0023] One or more p-doped material layers can include a p-TCO layer.

[0024] One or more light-emitting layers can include one or more quantum well (QW) layers. For example, one or more light-emitting layers can include three to eight QW layers.

[0025] The light-emitting element can emit light from the end of the light-emitting element that is opposite to the conductive layer.

[0026] The conductive layer can include a metal. The metal can be selected from the group consisting of gold, silver, and aluminum.

[0027] The light-emitting element can include a second conductive layer that contacts the n-doped portion.

[0028] λ can be a visible wavelength.

[0029] The axial dimension of the light-emitting element can be in the range of 0.5μm to 5μm, and the maximum lateral dimension orthogonal to the axis is in the range of 0.05μm to 0.5μm.

[0030] The light-emitting element can support fewer than 5 guiding modes.

[0031] In another aspect, the present invention is characterized by a light-emitting device comprising: a plurality of light-emitting elements, each light-emitting element being configured to emit light at a common peak wavelength λ1 in a common direction and a common angular range, the light-emitting elements being spaced apart from each other at a distance ranging from 0.05 μm to 1 μm.

[0032] The light-emitting device can include a plurality of second light-emitting elements, each configured to emit light at a common peak wavelength λ2 in a common direction and a common angular range, wherein λ2 is different from λ1. The light-emitting device can include a plurality of third light-emitting elements, each configured to emit light at a common peak wavelength λ3 in a common direction and a common angular range, wherein λ3 is different from λ1 and λ2. The light-emitting device can include a plurality of pixels, each pixel having at least three sub-pixels, a first sub-pixel having a first plurality of light-emitting elements, a second sub-pixel having a second plurality of light-emitting elements, and a third sub-pixel having a third plurality of light-emitting elements.

[0033] Typically, in another aspect, the present invention is characterized by a full-color display comprising a plurality of pixels having a white dot, each pixel comprising: a red sub-pixel having a plurality of nanowire light-emitting diodes (NW LEDs) configured to emit red light into an emission angular range, wherein 45% or more of the intensity of the red light is emitted into the angular range along a first direction and at a solid angle of 30° opposite to the emission point; a green sub-pixel having a plurality of NW LEDs configured to emit green light into the emission angular range, wherein 45% or more of the intensity of the green light is emitted into the angular range; and a blue sub-pixel having a plurality of NW LEDs configured to emit blue light into the emission angular range, wherein 45% or more of the intensity of the blue light is emitted into the angular range. Within the angular range, the white light emitted by the display has a chromaticity difference Du′v′ of less than 0.01 with the white dot of the display. Embodiments of the full-color display may include one or more features of other aspects.

[0034] Typically, in another aspect, the present invention is characterized by a full-color display comprising a plurality of pixels having a white dot, an emission direction, and an emission solid angle around the emission direction, the emission solid angle being characterized by a half-cone angle θ, each pixel comprising: a red sub-pixel having a plurality of red nanowire light-emitting diodes (NW LEDs), the plurality of red NW LEDs being configured to emit red light into the emission solid angle range such that the power fraction of the red sub-pixel emitted within the emission solid angle is at least 1.2 * (1 - cos(θ)). 2A green sub-pixel having multiple green NW LEDs configured to emit green light into an emission angular range such that the power fraction of the green sub-pixel emitted within the emission solid angle is at least 1.2*(1-cos(θ)). 2 ); and a blue sub-pixel having multiple blue NW LEDs configured to emit blue light into an emission angular range such that the power fraction of the blue sub-pixel emitted within the emission solid angle is at least 1.2*(1-cos(θ)). 2 The NW LED is configured such that the white light emitted by the display in any direction within the emission solid angle has a chromaticity difference Du′v′ of less than 0.01 with the white point of the display.

[0035] Embodiments of a full-color display may include one or more of the following features and / or other characteristics. For example, θ is in the range of 5° to 45°.

[0036] The red, green, and blue LEDs can each be represented by their respective shapes, so that the white light emitted by the display in any direction within the emission solid angle has a chromaticity difference of less than 0.01 Du′v′ with the white point of the display.

[0037] Red LEDs, green LEDs, and blue LEDs can be characterized by their physical dimensions and emission wavelengths, respectively, and the ratio between the physical dimension and emission wavelength of each LED is within + / - 10% of each other. In some embodiments, the ratios of each LED are approximately equal. The physical dimensions can be selected from the group consisting of: the radius, height, characteristic vertical dimension, characteristic horizontal dimension of the LED, and the optical distance between the emitting area and the reflector.

[0038] White dots can have a chromaticity that is essentially equal to D65.

[0039] Each LED can include a metal-based reflector, and can be essentially based on the equation To configure, among which, θ represents the phase shift at the interface between the metal-based mirror and the adjacent layers of the LED, d is the optical thickness between the interface and the light-emitting region of the LED, θ is the angle ranging from 0° to 30°, and m is an integer ranging from 1 to 3. The optical thickness d between the interface and the light-emitting region can be within 10 nm, satisfying the equation.

[0040] Other aspects and advantages will be apparent from the description, drawings and claims. Attached Figure Description

[0041] Figure 1A This is a schematic diagram illustrating the emission of NW LEDs from an array of nanowire (NW) LEDs.

[0042] Figure 1B It is a graph comparing the relative emission intensity from an exemplary directional transmitter and a Lambert transmitter at emission angles from 0° to 90°.

[0043] Figure 1C It is a graph comparing the fractions of launches from an exemplary directional launcher and a Lambert launcher at launch angles from 0° to 90°.

[0044] Figure 1D It is a graph comparing the light extraction from an exemplary directional emitter and a Lambertian emitter over the range of emission angles from 0° to 90°.

[0045] Figure 2A This is a schematic cross-sectional view of an exemplary NW LED.

[0046] Figure 2B This is a schematic cross-sectional view of another exemplary NW LED.

[0047] Figure 2C This is a schematic cross-sectional view of yet another exemplary NW LED.

[0048] Figure 3A This is a schematic cross-sectional view of an exemplary NW LED, showing the light path of light emitted by the LED.

[0049] Figure 3B This is a graph showing the emission intensity as a function of the angles of two exemplary NW structures in which the interference effect occurs.

[0050] Figure 3C It is a graph showing the total extraction efficiency into the air as a function of the thickness of the separated active region and the reflector.

[0051] Figure 4A This is a schematic cross-sectional view of an exemplary NW LED illustrating the guide light and sidewall light trails.

[0052] Figure 4B It is a diagram. Figure 4A A schematic cross-sectional view of the guiding pattern within an exemplary NW LED is shown.

[0053] Figure 5A This is a graph showing an exemplary emission distribution from an NW LED.

[0054] Figure 5B This is a graph showing another exemplary emission distribution from an NW LED.

[0055] Figure 5CThis is a graph showing another exemplary emission distribution from an NW LED.

[0056] Figure 5D This is a graph showing another exemplary emission distribution from an NW LED.

[0057] Figure 6A This is a schematic cross-sectional view illustrating the emission of a first guiding mode in an exemplary NW LED.

[0058] Figure 6B It is shown Figure 6A A schematic cross-sectional view of the emission of a second guiding mode in an exemplary NW LED is shown.

[0059] Figure 7A This is a graph illustrating Snell's law when GaN is extracted into the air.

[0060] Figure 7B It is a graph showing the emission intensity at various angle ranges.

[0061] Figure 8 This is a flowchart illustrating an exemplary method for designing NW LEDs.

[0062] Figure 9 This is a schematic cross-sectional view of the NW in an exemplary array of NW LEDs.

[0063] Figure 10A This is a graph showing exemplary spectral outputs of red, green, and blue LEDs that provide a D65 white point when properly mixed.

[0064] Figure 10B This is a graph showing the far-field emission as a function of angle for exemplary red NW LEDs, green NW LEDs, and blue NW LEDs.

[0065] Figure 10C This is a graph showing the far-field emission as a function of angle for another exemplary red NW LED, green NW LED, and blue NW LED.

[0066] Figure 10D It shows that it is aimed at Figure 10B The graph shows the trajectory of the white dots of the NW LED as a function of angle.

[0067] Figure 10E It is a comparison targeting the corresponding Figure 10B and Figure 10C The system's curve as a function of angle, Du′v′.

[0068] Figure 11AThis is a schematic cross-sectional view of an exemplary red LED.

[0069] Figure 11B This is a schematic cross-sectional view of an exemplary green LED.

[0070] Figure 11C This is a schematic cross-sectional view of an exemplary blue LED.

[0071] Figure 12A This is a graph comparing the normalized far-field emission as a function of angle for example red, green, and blue LEDs.

[0072] Figure 12B It is a graph comparing the curves of Du′v′ as a function of angle for two different systems.

[0073] Figure 12C This is a graph comparing the normalized far-field emission as a function of angle for another example red LED, green LED, and blue LED.

[0074] Figure 12D It is a comparison Figure 12B Compare the curves of Du′v′ as a function of angle for two different systems within a smaller angular range.

[0075] Figure 13 This is a schematic cross-sectional view of an exemplary device that includes three different LEDs.

[0076] Figure 14A This is a schematic cross-sectional view of an exemplary multifaceted LED.

[0077] Figure 14B This is a schematic cross-sectional view of another exemplary multifaceted LED.

[0078] Figure 15 This is a graph showing the FFP as a function of the polar angle, illustrating the contrast definition between the two lobes within the angular emission cone of an exemplary LED.

[0079] Figure 16 This is a schematic diagram of a display composed of LEDs. Detailed Implementation

[0080] Nanowire (NW) LEDs offer the opportunity for directional emission. The far-field pattern (FFP) of light emission from an NW LED can be affected by various effects, including individual and collective effects. Individual effects include the influence of the optical design of a single NW on its emission characteristics. Collective effects include the interaction of light emitted from an NW with other surrounding NWs, including diffraction effects from multiple NWs. The embodiments disclosed herein enable the realization of directional FFP by utilizing individual and / or collective effects.

[0081] Directional FFP can be quantified by the fraction of total emissions within a specific angular range. Figure 1A An array 100 of NW LEDs 110 arranged in a plane is schematically shown, and the emission I(θ) of a given NW LED is measured as a function θ perpendicular to the plane.

[0082] Figure 1B This is a graph comparing the emission of two examples from FFP. The vertical axis represents the intensity of the FFP, normalized to the peak emission at θ = 0°. The Lambert FFP I(θ) ~ cos(θ) is common for conventional LEDs. An exemplary directional NW LED can achieve FFP I(θ) ~ cos(θ). 2 This allows for better directionality than the Lambert FFP near the normal incidence.

[0083] Figure 1C It shows that it is aimed at Figure 1B The graphs show the fraction of total emission between normal incidence and a given angle for exemplary directional FFPs and Lambert FFPs. For the Lambert FFP, 25% of the emission is within 30°, while for the directional FFP, 35% of the emission is within 30°.

[0084] Directional FFP can also be quantified by the absolute extraction efficiency within a specific angular range (where extraction efficiency is the ratio of light emitted from the outside of the LED to light emitted from inside the LED). Figure 1D This is a graph showing the extraction efficiency versus angle range for two previous FFPs, assuming the LED has a total extraction efficiency of 60% in both cases. For the Lambert FFP, the extraction efficiency within 30° is 15%. For the directional FFP, the extraction efficiency within 30° is 22%.

[0085] Therefore, in Figure 1A-1D In the examples, the directional FFP achieved a 40% relative directionality improvement compared to a Lambertian emitter in the 0-30° range. Some embodiments are characterized by this directionality enhancement for Lambertian emitters. For some embodiments, Table 1 shows values ​​for the enhancement and emission fraction within the pyramid for various angle ranges. Some embodiments achieve at least such enhancement and such emission fraction within the pyramid.

[0086] 0-20° 20％ 14.5％ 0-20° 50％ 18％ 0-20° 100％ 24％ 0-30° 20％ 30％ 0-30° 40％ 35％ 0-30° 60％ 40％ 0-45° 10％ 55％ 0-45° 20％ 60％ 0-45° 30％ 65％

[0087] Individual effects on directionality

[0088] In some embodiments, the optical design of the NW LED is configured to influence its directivity. The optical design refers to the geometry of the NW LED, including its height, lateral dimensions / diameter, tapered shape, cross-sectional shape (e.g., circular, hexagonal, etc.). It further includes the thickness and refractive index of the various layers comprising the NW LED.

[0089] Exemplary NW LEDs include active regions having one or more light-emitting layers (e.g., quantum wells or dual heterostructures). In the following discussion, group III nitride LEDs will be used, but other material systems can be considered. An exemplary group III nitride NW epitaxial stack may be as follows (starting from the growth substrate): an n-GaN layer; an InGaN underlayer (e.g., a host or superlattice, possibly n-doped); a GaN underspacer layer (e.g., undoped and / or n-doped); an active region, such as a QW / barrier stack (e.g., undoped); a GaN underspacer layer (e.g., undoped and / or p-doped); an AlGaN electron blocking layer (e.g., p-doped); and a p-GaN layer.

[0090] Typically, the growth substrate can be a sapphire, silicon, SiC, freestanding GaN, GaN, or InGaN buffer layer grown on another material. The growth substrate can be patterned with a mask to facilitate the growth of the new wave (NW) within the mask openings. After growth, the NW can be transferred to a substrate assembly (e.g., where the n-side or p-side of the NW faces the substrate), and the substrate can be removed. n-type and p-type contacts are formed to the NW. Typically, several NWs are connected in parallel to form an LED. The LED can form a sub-pixel of a display system.

[0091] Some embodiments influence the FFP by means of optical interference caused by the p-contacts. In some embodiments, light emitted from the active region is reflected by the p-contacts, thereby causing interference. The thickness and refractive index of various layers can be configured to control this interference, as will be discussed below.

[0092] In some embodiments, the p-contact may include a highly reflective metal layer, including Ag and / or Al. The p-contact may further include a transparent conductive layer (e.g., a transparent conductive oxide (TCO)), for example, between p-GaN and the metal.

[0093] Figure 2A An exemplary structure of an NW LED 200 is shown. Specifically, the NW LED 200 is a single line of an NW LED. The NW 200 has an epitaxial layer including a p-GaN layer 220, an active region 210, an n-GaN layer 250, a p-contact composed of a metal layer 240 and a TCO layer 230, and an n-contact composed of another TCO layer 255. The contacts may further extend laterally and contact other NWs in the array of NWs of the NW LED.

[0094] Other exemplary partial structures for NW are Figure 2B and 2C The diagram shows different layers for the p-contact structure compared to NW 200. For example, in some cases, the p-contact includes at least one Ag layer 280 with a thickness of at least 50 nm (e.g., 100 nm or more, 200 nm or more) to provide sufficient reflectivity. Layer 280 can include metals other than Ag, such as metals alloyed with Ag. Additional metals can be present in small traces (e.g., through doping or diffusion), existing as thin layers in the Ag layer (either in the bulk of the layer or at its surface). Figure 2B The structure shown also includes an interface layer 270 present at the Ag / semiconductor interface. Interface layer 270 can be a thin layer of another metal (e.g., Ti, Al, Ni, Pt) and can promote low contact resistance. Interface layer 270 can have a thickness of 10 nm or less (e.g., 3 nm or less, 10 Å or less, 3 Å or less, such as 1 Å). It can be an alloy layer of Ag and another layer. Interface layer 270 can be deposited or diffused to the Ag / semiconductor interface.

[0095] In these various embodiments, sufficient reflectivity of the light emitted by the NW is desired. In some embodiments, for light emitted from the active region, the single-bounce reflectivity under normal incidence of the contact stack is 80% or greater (e.g., 85% or greater, 90% or greater, 93% or greater, 95% or greater, 97% or greater, such as 99% or greater). The interface layer 270 can be configured to ensure this reflectivity, for example, by ensuring that the interface layer is thin enough to limit its absorption. Figure 2B The stack shown also includes a metal stack 285 between the Ag layer 280 and the substrate 260. This stack can provide various functions (e.g., encapsulation, diffusion barrier, adhesion, thermal conduction, mechanical compliance) and can include Ag, Al, Au, Ni, Ti, Pt, Sn, Cu and other metals.

[0096] Figure 2C An alternative metal stack is shown, wherein a TCO layer 258 (such as ITO, ZNO, IZO, CuO or other TCO) is further present between the interface layer 270 and the semiconductor layer 220.

[0097] In embodiments with a metal layer, the p-contact acts as a reflector, causing interference between the light reflected by the reflector and the light emitted by the active region 210. Figure 2AThe interference is illustrated in the following manner: light emitted downward from the active region 210 (in the p direction) is reflected upward by the metal layer 240 and interferes with light emitted upward from the active region 210 (in the n direction). Such a geometry can be called a half-cavity geometry because the interference is caused by only one mirror (compared to two mirrors in a conventional cavity).

[0098] Not wanting to be bound by theory, in the following text, before discussing the embodiments in NW LEDs, we first discuss the half-cavity interference of LEDs in the case of planar geometry (i.e., for LEDs with infinite lateral range).

[0099] An important parameter of the interference is the optical thickness between the active region and the mirror. In this disclosure, d1 represents the optical thickness between the top of the active region and the p-mirror, and dc represents the optical thickness between the center of the active region and the p-mirror. Figure 2A The values ​​d1 and dc for NW 200 are shown. For clarity, the optical thickness is the product of the physical thickness and the refractive index at the wavelength of interest (e.g., the peak emission wavelength of the active region). For example, for an epitaxial stack with layers as given in the table below (with refractive indices at the wavelengths of interest):

[0100] p-metals Ag 100 0.05+2.5*i ITO interval ITO 50 1.8 p-GaN GaN 100 2.5 Active area GaN / InGaN 40 2.6

[0101] The optical thicknesses d1 and dc are as follows:

[0102] d1 = 100 x 2.5 + 50 x 1.8 = 340 nm

[0103] dc=20x2.6+100x2.5+50x1.8=392nm.

[0104] d1 can be a useful quantity characterizing the optical interference of light emitted from the top (p-side) of the active region; dc can be a useful quantity characterizing the optical interference of light emitted from the middle of the active region. In some embodiments, most of the light is emitted near the top of the active region (e.g., via the p-side QW); in some embodiments, the light is emitted across the active region, and dc roughly characterizes the center of the emission. In some embodiments, the active region is thin (e.g., less than 10 nm), and d1 and dc are practically equal. Light propagating from the active region 210 and reflected by the mirror (metal layer 240) experiences a phase shift proportional to d1 / dc. The metal of the mirror further causes a phase shift (e.g., if the mirror is Ag-based, it can cause a phase shift of approximately -π / 2). The total phase shift of the emitted light is then the sum of the phase shift caused by propagation in the transparent layer and the reflection at the mirror. Depending on the value of this total phase shift, constructive or destructive interference may occur in a given direction.

[0105] Figure 3A The illustration shows the cumulative phase shift propagating within layer 330 of NW 300 due to reflection from mirror 340. In this example, NW layers 320 and 330 have a constant refractive index n, and the light-emitting layer 310 is located at a distance t from mirror 340 (in this case, the optical thickness is d1 = n*t).

[0106] The electric field emitted in direction θ due to the interference between the two waves is Where E0 is the intensity of the emission field in the host medium. The reflectivity of the mirror is k = 2π / λ, where λ is the emitted wavelength, t is the thickness between the emitting layer 310 and the mirror 340, and θ is the emission angle. When the total phase shift is: When constructive interference occurs in a given direction, where m is an integer. In this case, the material is homogeneous, and the optical distance d1 is simply equal to n*t.

[0107] The emission intensity I(θ) is the square of the electric field E. If the mirror is perfect (i.e., |r| = 1), the emission intensity can reach a maximum enhancement of 4 (constructive interference) and a minimum value of 0 (destructive interference).

[0108] Figure 3B This is a graph illustrating the I(θ) produced by this interference for a GaN LED (n = 2.5) with an Ag mirror. I(θ) is normalized by I0 (the value in the bulk material). Therefore, a value of I(θ) = 2 means that the emission intensity is enhanced by a factor of two through the interference effect.

[0109] Two exemplary values ​​of t were considered for this curve. Case 1: When t = 112 nm, constructive interference occurs in the vertical direction, with a value close to 4. Case 2: When t = 120 nm, constructive interference occurs at θ = 20°.

[0110] Figure 3C The net extraction efficiency resulting from this interference effect is shown. Figure 3C It is for and Figure 3A The graph shows the total extraction to air of the same LED in -B as a function of thickness t. The extraction oscillates with t, with a period of λ / 2n (approximately 90 nm). In this example, by perfectly tuning t, an extraction efficiency higher than 20% into the air can be achieved. If a specific directionality is sought, the optimal value of t can be varied; the extraction efficiency can be higher than 10% (e.g., 15% or more, 20% or more). Other designs can promote even higher extraction efficiencies (e.g., 30% or more, 40% or more, 50% or more).

[0111] In more complex structures with non-uniform refractive indices, the term t*n should be replaced by the optical thickness d1 (or dc), and the condition for constructive interference under normal incidence is:

[0112] In summary, by configuring the optical design of the half-cavity, interference can be enhanced or suppressed in some directions. The configuration of the optical design may include: selecting the thickness and refractive index of the layer; selecting a metal mirror with a desired reflectivity and / or phase shift; configuring the optical thickness to induce interference in one direction; configuring contacts with a metal reflector and a TCO layer; selecting the radius or shape of the NW; and other means known in the art. In some embodiments, configuring the optical design includes selecting an angle and configuring the optical length of the light reflected by the mirror to obtain constructive or destructive interference at said angle.

[0113] Some embodiments of the present invention utilize this half-cavity interference in NW LEDs. The light emission in this case is more complex than that of simple planar geometry, and will now be discussed.

[0114] Figure 4A Two light emission paths in an NW LED 400 are shown. In this example, the NW 400 includes a substrate 440, a p-contact layer 432, a semiconductor layer 430, a light-emitting region 410, another semiconductor layer 420, and a non-contact layer 422. In the other paths, the light-emitting region 410 emits light into the semiconductor layer 420 along two different paths. Light emitted near the horizontal direction can escape from the sidewalls and will be referred to as sidewall light. This is for light generation within the sidewall light cone, such as... Figure 4A The dotted line in the diagram illustrates the angle of the light cone (counted horizontally). This angle is asin(n_ext / n_in), where n_ext and n_in are the refractive indices of the external and LED media, respectively (this expression also holds for light emitted from the center of the active region, but provides the correct order of magnitude for other emission locations). Therefore, if the NW 400 is GaN-based (refractive index ~2.5) and the external medium has a refractive index of 1 (or 1.5), the angle of the light cone is approximately 24° (or 37°).

[0115] On the other hand, light emitted to the outside of the light cone is reflected by the NW sidewalls due to total internal reflection, forming guide light.

[0116] Figure 4BThe diagram illustrates the guiding light bouncing several times within layer 420 and forming guiding modes that propagate along the axis of NW 400. Various modes with different symmetries are possible (e.g., TE, TM, HE, and HM symmetries in the case of a circular NW). Each of these modes can be characterized by a propagation angle θ along the NW axis. These guiding modes are analogous to those in an optical fiber. Typically, the number of modes increases with the NW radius (or more generally, its lateral dimension).

[0117] Figure 5A An example of the resulting emission pattern is shown (i.e., the intensity contrast angle of light emitted from the active region). Within the sidewall cone, light can be emitted as a broad continuous spectrum. Outside the cone, light is emitted in a series of discrete directions corresponding to the guiding modes (modes 1 and 2 are shown here, with corresponding propagation angles of approximately 20° and 45°).

[0118] Due to interference effects, some embodiments can achieve directionality by selectively exciting certain guiding modes and / or sidewall light. Selective excitation means a change (increase or decrease) in the amount of power emitted in one direction.

[0119] In some embodiments, sidewall light is undesirable because it is emitted near the horizontal direction. The optical design of the NW can be configured to induce destructive interference near the horizontal direction to suppress this emission. In some embodiments, the optical design is configured to produce destructive interference at an angle θ of 90° or 85° or 75° or 70° and / or at one or more angles in the range of 70°–85°.

[0120] Figure 5B Another example of an emission pattern is shown in an embodiment in which sidewall light emission is suppressed. In some embodiments, the amount of light emitted from the sidewall of the NW LED is 30% or less (e.g., 25% or less, 20% or less, 15% or less, 10% or less, 5% or less).

[0121] In some embodiments, selective excitation of some guiding modes is sought against others. For example, the NW can have a radius that supports two guiding modes propagating at angles θ1 and θ2. The optical design of the NW is configured to induce constructive interference at angle θ1 and / or destructive interference at angle θ2, such that mode 1 is preferentially excited. Figure 5C An exemplary emission diagram of this embodiment is shown, wherein emission is suppressed in mode 2.

[0122] Figure 5D An exemplary emission diagram is shown for an embodiment of mode 2 and sidewall light suppression.

[0123] Depend on Figure 3BThe interference patterns shown can be adapted to achieve this effect because they enhance emission at small angles and suppress emission at intermediate angles close to 50°. This selective excitation of the guiding modes can lead to directional FFPs because some modes emit in specific directions outside the NW within the FFP.

[0124] In some embodiments, an emission direction is selected; and the optical design of the NW is configured to achieve constructive or destructive interference in the emission direction. Embodiments include methods for configuring LEDs as taught herein.

[0125] In some embodiments, the NW supports at least two guiding modes. The NW includes a mirror, and optical interference between the emitted light and the mirror suppresses the emission into a guiding mode that propagates within a predetermined angular range. In some embodiments, the emission is substantially suppressed within a range of 40-50° (and / or 30-40°, 50-60°, 60-70°). In some embodiments, 10% or less of the total emission (e.g., 5% or less, 2% or less) is emitted within the suppressed angular range.

[0126] In some embodiments, the NW supports at least two guiding modes. The NW includes a reflector, and the optical interference of the emitted light with the reflector causes 30% or more (e.g., 40% or more, 50% or more, 60% or more, 70% or more) of the total emission to be guided into one mode; and / or 20% or less (e.g., 10% or less, 5% or less, 2% or less) of the emission to be emitted into other modes.

[0127] In some embodiments, the NW supports at least two guiding modes. The NW includes a mirror, and optical interference between the emitted light and the mirror results in selective excitation of a portion of the modes. This leads to a directional FFP, the directionality of which can be described by the values ​​in Table 1.

[0128] The change in emission to a mode can be quantified in absolute terms. In some embodiments, emission enhancement is sought, and a mode or subset of modes is selectively excited and receives 30% or more (e.g., 40% or more, 50% or more, 60% or more, 70% or more) of the total light emitted by the active region. In some embodiments, emission suppression is sought, and a mode or subset of modes is selectively excited and receives 20% or less (e.g., 10% or less, 5% or less, 2% or less) of the total light emitted by the active region. The change in emission to a mode can be quantified by its relative change due to interference. In some embodiments, a mode receives a first emission in the absence of interference (i.e., in an infinite NW without a mirror) and a second emission in the presence of interference (i.e., in the presence of a mirror), and the change in emission is quantified as the ratio of the second emission to the first emission. In some embodiments, emission enhancement is sought, and this ratio can be 1.5 or greater (e.g., 2 or greater, 2.5 or greater, 3 or greater, 3.5 or greater, 4 or greater). Ratios higher than 2 can be non-negligible because they can be caused by wave interference (as opposed to incoherent reflections alone). In some embodiments, emission suppression is sought, and the ratio can be 1 or less (e.g., 0.8 or less, 0.5 or less, 0.3 or less, 0.2 or less, 0.1 or less, 0.05 or less, 0.01 or less).

[0129] In such embodiments, the interference conditions may not be strictly given by the phase-matching equations described above: these involve planar systems, and the specific optical designs that cause interference of modes in the NW will be slightly different. However, given the teachings of this paper, such optical designs can be determined for specific geometries (e.g., through computer simulations, including numerical modeling known in the art).

[0130] In some embodiments, modes with a propagation angle close to the NW axis are preferred because such modes can have a more directional FFP. In particular, (after transforming the angle to the external medium via Snell's law) the FFP of some modes is approximately maximized at the angle corresponding to their propagation angle. This is especially true when the propagation angle of the mode places it inside the escape cone on the top surface of the NW.

[0131] This is Figure 6A and 6B As shown in the figure, Figure 6A and 6B The illustration shows emission from two guidance modes derived from nanowire 600. Figure 6AIn the model, the guiding mode propagates along the axis of NW 60 at a low internal angle θin. This angle is smaller than the critical angle for extraction into an external medium (e.g., from GaN to air). When incident on the top surface of the NW, the mode escapes the surface with good efficiency because its dominant wave vector can propagate in the external medium, and the mode's FFP is primarily oriented in the external direction θout given by Snell's law (n_in sinθin = n_out sinθout). Therefore, good directivity can be obtained using this mode.

[0132] Figure 7A The diagram illustrates Snell's law in the case of GaN extracted into air. For example, a mode with a propagation angle of 10° (or 5°, 7°) in GaN NW can have a maximum emission FFP of approximately 25° (or 12°, 20°) in air, which may be suitable for some applications.

[0133] on the contrary, Figure 6B The guiding mode in an NW 600 propagating at a high angle θin greater than the critical angle is shown. Direct extraction of the main wave vector is not possible; instead, the mode is diffracted in all directions. Therefore, this mode may be detrimental to directional emission.

[0134] Therefore, some embodiments can facilitate transmission to low-angle modes. In some embodiments, modes with a propagation angle of 20° or less (e.g., 15° or less, 10° or less, 5° or less) receive enhanced transmission. Modes with a propagation angle of 25° or greater (e.g., 30° or greater, in the range of 40-50°, 50-60°, or 40-60°) receive suppressed transmission. Figure 7B Various angle ranges are illustrated. In some embodiments, emission to a guided mode within the top light cone is preferred, and emission to a guided mode outside the light cone and into the continuum of the sidewall light cone is not preferred.

[0135] In some embodiments, the top surface of the NW has a non-planar shape that modifies the emission direction of one or more modes. The top surface may be concave, convex, or textured.

[0136] Besides interference effects, the shape of the NW itself can be an important parameter for optical design. The radius / lateral dimension of the NW can be selected to carry a desired number of guiding modes with desired symmetries (e.g., 1, 2, 3, 4, 5, less than 10) at the peak emission wavelength of the active region. Some mode symmetries have preferred optical coupling with certain electric dipole orientations. For example, in the case of c-plane group III nitride emitters (and some other semiconductors), the emitting dipoles are in-plane. Such dipoles can preferentially couple to modes with HE and TE symmetries (compared to modes with TM symmetries). Therefore, the radius of the NW can be selected to produce modes with desired symmetries and desired propagation angles or desired far-field radiation patterns. The interference effect is then configured to preferentially emit to these modes and suppress emission to modes propagating at other angles.

[0137] Figure 8 An example of method 800 for achieving this is shown. In an initial step 810, the radius / lateral shape of the NW is selected, and in step 820 (e.g., by modeling an NW that can be assumed to be infinitely long), the supporting modes of the NW are determined. In step 830, some modes are selected as desired or undesirable; this selection can be achieved by defining the figure of merit of the FFP (e.g., directional emission in a 30° cone) and determining which modes contribute to that figure of merit (e.g., by calculating their far-field patterns). Based on this, in step 840, the remainder of the optical design of the NW (layer thickness, mirror material, and location) is configured to generate interference to selectively excite the modes as specified.

[0138] In some embodiments, the lateral dimension is selected such that the NW carries three modes. The first mode (e.g., HE mode) has a low propagation angle and receives a large amount of light emission; this mode has a directional FFP. The second mode (e.g., TE mode) has a large propagation angle; emission entering this mode is suppressed by interference effects. The third mode has symmetry (e.g., TM mode), which results in low light emission entering this mode because it cannot couple well to light emission from the active region. A sufficient fraction of the total emission is guided in the first mode, thereby producing a directional FFP.

[0139] In some embodiments, the NW has an active region containing only one (or two, or three) emitting QWs. In some embodiments, substantially all light is emitted from a region with a thickness of 20 nm or less (e.g., 15 nm or less, 10 nm or less, 5 nm or less). Such a thin active region can promote interference effects. It should be understood that other layers (e.g., a substantially non-emitting "dummy" InGaN quantum well layer) may be present, for example, to improve material quality.

[0140] Some embodiments include devices characterized by NWs emitting at different wavelengths, and the NW optics are different for each emission wavelength, such that interference occurs in or near the same direction regardless of the emission wavelength. For example, a display could include three sets of NWs emitting red, blue, and green radiation respectively. Constructive interference under normal incidence is sought. Each set of NWs includes an active region, p-GaN, and Ag-based mirrors (wherein... The thickness of the p-GaN is selected for each emission wavelength to ensure that the interference conditions are met. This results in normal-incident constructive interference for all NWs. The following example illustrates how this can be achieved:

[0141]

[0142] In some embodiments, it may not be possible to independently control the thickness of the p-GaN layer for each NW group. In this case, the same effect can be achieved by forming a transparent conductive layer between the p-GaN and the p-metal, selecting the thickness of the transparent conductive layer for each NW group to achieve the desired interference. The following illustrates how this can be achieved when the p-GaN thickness is the same for all NW groups:

[0143]

[0144] While the refractive index, thickness, and wavelength given herein are merely examples, those skilled in the art will more generally understand how to configure optical designs of NWs of different wavelengths to achieve simultaneous interference in the same or near direction.

[0145] While the above discussion focuses on the half-cavity effect on the p-side of an LED, other cavity effects can be utilized to manipulate extraction and influence the FFP. For example, in devices where light is emitted from the p-side (e.g., if the p-side contacts are transparent or translucent), a half-cavity effect can be used on the n-side of the device. This can be achieved with an n-contact having a metal-based mirror, as discussed above in the case of p-contacts. Alternatively, an epitaxial reflector (also known as a distributed Bragg reflector (DBR)) can be integrated into an epitaxial stack. This DBR can be positioned at an appropriate distance from the emitting region to induce interference. More generally, one or more reflectors (including metal mirrors, DBRs, metal / DBR hybrids) can be placed on one or both sides of the active region to induce interference and influence the FFP. For group III nitride LEDs, the DBR can be made of GaN, AlGaN, AlInGaN, or AlInN layers.

[0146] In some embodiments, the lateral regions surrounding the luminescent region have a relatively low refractive index. This reduces sidewall light emission. For example, the lateral regions between the N and W sides comprise air or a transparent material with a low refractive index (e.g., 1.4 or less, 1.3 or less, 1.2 or less). Low refractive index materials can be obtained by various methods, including using porous media (e.g., nanoporous dielectrics comprising oxide or nitride dielectrics); or using materials in which nanoscale low refractive index inclusions have been dispersed (e.g., hollow dielectric spheres with an air core dispersed in a dielectric or sol-gel or other transparent medium).

[0147] In some embodiments, the metal overlaps with the NW sidewalls, for example because the metal (n-contact or p-contact) is deposited on a sample having the NW and a slightly over-etched filler / planarization material (making the top surface of the NW slightly protrude from the planarization material). This can lead to optical loss if the emitted light couples to the absorbing plasma. Therefore, in embodiments, the LED is configured to minimize absorption, including absorption due to plasma coupling. For example, the metal and geometry can be selected to reduce surface plasmon coupling. This can be achieved using a metal with a suitable refractive index. Depending on the emission wavelength, a given metal (e.g., Ag, Al, Ti, Ni, Au, other metals) can have lower plasmon absorption because the plasma frequency is far from the emission frequency. This can be further achieved by replacing sharp corners (e.g., at the top of the NW) with angled, curved, or circular surfaces (including semi-polar planes). In some embodiments, the metal overlaps with the NW sidewalls and is configured to induce extraction of surface plasmons.

[0148] In some embodiments, overlap between the metal and the NW sidewalls is avoided. This can be achieved, for example, by a planarization layer having a surface that is precisely or substantially flush with the top surface of the NW (e.g., within + / - 1 nm, 3 nm, 5 nm, 10 nm, or 20 nm). This can be achieved by inserting a TCO layer between the top surface of the NW and the metal. The surface of the TCO (in contact with the metal) can be smoother than the top surface of the NW, thereby reducing or eliminating surface topologies that lead to metal absorption. In some embodiments, a planarization is formed around the NW, and the planarization is slightly over-etched, resulting in a surface that protrudes from the NW, characterized by a height. A TCO is formed to contact the NW, and the TCO has a top surface characterized by a second height of 80% or less (e.g., 50% or less, 20% or less, 10% or less, 5% or less, 1% or less) of a first height. A metal portion is formed over the TCO.

[0149] Directional collective effect

[0150] In some embodiments, a plurality of NWs are configured in a predetermined spatial arrangement (such as a two-dimensional array of NWs). The array may periodically have a square, rectangular, or triangular lattice. In some embodiments, the array can be quasi-periodic. For example, it may have short-scale periodicity with long-scale variations (e.g., it may form a metasurface with spatially varying array spacing).

[0151] Collective optical effects in such NW arrays are known in the art. For example, the array can be used as a diffraction grating (or photonic crystal) to diffract light propagating in a plane to out-of-plane directions. Embodiments are able to combine such collective effects with individual effects to achieve improved directivity. Thus, the techniques disclosed above are used to configure the optical design of individual NWs to preferentially emit light in some directions. The optical design of the array can be further configured to diffract lateral light out of plane. Therefore, embodiments include methods for configuring both the design of individual NWs and the design of NW arrays to achieve a desired FFP.

[0152] refer to Figure 9 The extended array 900 of NWs 910, 920, and 930 supports various optical modes. Each NW 910, 920, and 930 supports a guiding mode 911 in the vertical direction (i.e., along the axis of the NW). Furthermore, the sidewall light emitted by each NW can be guided in-plane (because the array of NWs forms an effective waveguide in the lateral direction), resulting in a lateral guiding mode 912. These guiding modes can reduce the overall light extraction efficiency and should be extracted.

[0153] The embodiments can utilize interference effects to selectively excite certain vertical modes, as taught herein. Furthermore, the optical design of the NW array 900 can be configured to diffract laterally guided modes in a desired direction. For example, the spacing of the NWs can be selected to satisfy the Bragg condition for diffraction at the desired emission angle.

[0154] Color / chromaticity effect

[0155] Some embodiments are characterized by a non-white light (NW) serving as the emitter of a display. The display includes multiple pixels, each having at least three sub-pixels that emit essentially blue, green, and red radiation, respectively. In such embodiments, it may be desirable to avoid substantial differences in the color feed (FFP) between sub-patterns, as these differences can lead to variations in color angle (COA). For example, if the display emitter can be configured to emit white light (e.g., at the D65 color point), its perceived color can vary depending on the angle of light emission because the different sub-pixels have different FFPs.

[0156] Therefore, in some embodiments, each subpixel is configured to provide the desired FFP, and three subpixels are configured such that their FFPs are similar. This can be used to reduce COA variation. COA variation can be quantified by variations in chromaticity in a color space (such as a (xy) or (uv) or (u'v') color map). In particular, the (u'v') color map may be considered due to its improved color uniformity, and the corresponding chromaticity difference Du'v' can be used to quantify the color difference. Given a desired white chromaticity of a display (such as D65), each emission direction has a chromaticity difference of 5E-3 or less (e.g., 3E-3 or less, 1E-3 or less) from D65.

[0157] Figures 10A-10E These are a series of curves illustrating the COA effect. Figure 10A This is a graph showing the intensity as a function of wavelength for a set of RGB LEDs configured to emit white light with a D65 chromaticity when properly mixed. The peak wavelength / FWHM (in nm) of the LEDs are 460 / 25, 530 / 30, and 650 / 45, respectively. This selection of primary colors enables coverage of the DCI-P3 color gamut. Figure 10B A graph of FFP as a function of angle is shown for the first orientation system, where the B, G, and R primary colors have significantly different FFPs. Figure 10C This is a graph showing the FFP as a function of angle for the second orientation system, where the primary colors B, G, and R have fairly similar FFPs. Figure 10D It shows that it is aimed at Figure 10B The exemplary system is plotted as a curve of chromaticity versus angle in the (u'v') diagram. Here, each point corresponds to a 5° step in θ. Under normal incidence, the system emits at D65 (marked by a circular symbol), as designed. However, as the angle increases, the chromaticity shifts significantly from D65. Figure 10E This is a graph showing the chromatic distance Du'v' from D65 as a function of the emission angle. For the first system, Du'v' acquires a large value: at θ = 30°, Du'v' = 5e-3. Conversely, for the second system ( Figure 10C The Du'v' value remains low: at θ = 30°, Du'v' = 1.5e-3.

[0158] In some embodiments, the optical design of each subpixel is configured to achieve a near-constant COA. For example, the number of guiding modes in the blue, green, and red subpixels is substantially equal. Interference effects caused by mirrors in each subpixel cause enhancement or suppression at the same angle. In some embodiments, one or more dimensions of the subpixel are proportional to its emission wavelength (i.e., for each subpixel, the thickness or distance divided by the wavelength is equal). For example, for each NW type, the radius divided by the emission wavelength is substantially equal. An example is as follows (radius / emission wavelength): 225 / 450nm, 260 / 520nm, 310 / 620nm. Because Maxwell's equations are scaled with wavelength, this scaling enables similar optical behavior from the subpixels and promotes the same FFP. In some embodiments, the scaling takes into account the dispersion of the refractive index, making the optical path equal in the subpixel. The optical path may be associated with the NW radius or with the interference of the emitted light by the mirrors.

[0159] In some embodiments, the COA variation is small in a first angular range but large in a second angular range. This can be acceptable if the display system relies primarily on emission within the first angular range. For example, the optical system collects light emitted between 0° (normal incidence) and the cutoff angle, and substantially does not collect light emitted between the cutoff angle and 90° (horizontal incidence). In such embodiments, the COA can be lower than desired in the first angular range. For example, within an emission cone at + / -20° (or + / -10°, + / -30°, + / -45°, + / -90°) from normal incidence, the emitted chromaticity is less than 5E-3 (e.g., 3E-3 or less, 1E-3 or less).

[0160] Therefore, in some embodiments, the display is configured to emit white chromaticity. The display has emission over an angular range, the angular average chromaticity of which is essentially white chromaticity. It has local emission containing local chromaticity in each direction within the angular range, and the absolute Du'v' difference between each local chromaticity and the average chromaticity is less than 1E-3.

[0161] In some embodiments, instead of a simple chromaticity diagram, a more complex color space (such as CIELAB, CIECAM, or CAMO2-UCS) can be used. In this case, a color difference metric and a standard for COA are selected. For example, the color space could be CIELAB, and the color difference could be the Euclidean distance within that space. The white point is the target color of the display (e.g., D65).

[0162] Some implementations combine basic directionality with small variations in COA.

[0163] other

[0164] While various embodiments refer to GaN-based LEDs, the teachings apply to other materials. NW LEDs may include materials made of other group III nitride compounds (including AlGaN, InGaN, AlInGaN) or other semiconductors (including AlGaAs, AlInGaP, AlInGaAsP, ZnO).

[0165] The thickness values ​​disclosed herein are subject to some tolerance. For example, these values ​​can be understood as having a tolerance of + / -5nm (e.g., + / -10nm, + / -15nm, + / -20nm).

[0166] Although the discussion in this disclosure focuses on NW LEDs, the teachings can be further applied to other types of LEDs. For example, although not having an NW shape, microLEDs can have well-defined mode structures. As taught herein, the optical design of an LED can be configured to induce optical interference with a mirror to cause selective excitation (enhancement or suppression) of some optical modes. Furthermore, although not having an NW geometry, a group of RGB microLEDs can exhibit preferred directionality. As taught herein, the optical design of LEDs of various colors can be configured such that the FFP of each color is substantially similar, thereby avoiding the COA effect.

[0167] As an example, LEDs can have planar geometry; for instance, they can be microLEDs with lateral dimensions ranging from 1 to 20 μm and vertical dimensions ranging from 1 to 10 μm. In a specific example, an LED is a parallelepiped with dimensions of approximately 4 × 4 × 1 μm. In such LEDs, the lateral dimensions can be large enough that many modes excited by the emitting medium exist regardless of the LED's wavelength, making it potentially redundant to scale the lateral dimensions of R / G / B LEDs to their wavelengths. Nevertheless, the teachings disclosed above can be applied to ensure that R / G / B LEDs have similar FFPs to reduce COA artifacts. For example, LEDs can still benefit from interference effects between the emitting layer and the mirror; in this case, scaling the optical thickness (d1) between the emitting layer and the mirror can produce similar COA for each color. That is, for R / G / B LEDs, the ratio of optical thickness to wavelength can be substantially equal.

[0168] Figures 11A to 11C An example of such an embodiment is illustrated. Figure 11A-11CRed, green, and blue LEDs (1100, 1110, and 1120), each with a planar shape, are shown. LEDs 1100, 1110, and 1120 each have a reflector 1130 (e.g., a p-mirror). LED 1100 has an emission layer 1101 (e.g., a red QW layer) separated from the reflector 1130 by a semiconductor layer 1102. LED 1110 has an emission layer 1111 (e.g., a green QW layer) separated from the reflector 1130 by a semiconductor layer 1112. LED 1120 has an emission layer 1121 (e.g., a blue QW layer) separated from the reflector 1130 by a semiconductor layer 1122. The corresponding optical thicknesses d1R, d1G, and d1B between the emitter and the reflector (i.e., the optical thicknesses of layers 1102, 1112, and 1122, respectively) are configured to be substantially proportional to the corresponding emission wavelengths λR, λG, and λB, such that each LED has a similar FFP. For example, each FFP has a preferred emission angle θR, θG, θB (where the FFP has a peak value at the corresponding angle), and the angles are substantially equal. In some embodiments, these angles are equal within + / -5° (or + / -1°, + / -3°, + / -7°, + / -10°).

[0169] Figures 12A to 12D The method is further illustrated by modeling the FFP and chromaticity of a planar LED with a half-cavity effect. The LEDs have corresponding emission wavelengths of 450 nm, 520 nm, and 620 nm. They comprise a quantum well at a distance t from a silver mirror, which is placed in a GaN dielectric (n = 2.5) and emitted into air. In a simplified case, all three LEDs share the same value t = 115 nm (this value is optimized for extraction of the blue LED). This yields... Figure 12A The FFP shown in the graph indicates that the green and red LEDs have suppressed emission at large angles because they are detuned (the value of t is too low for extraction at these longer wavelengths). Instead, some embodiments correct for this by having near-constant t / λ values. Precisely constant values ​​are achieved with t_b = 115nm, t_g = 133nm, t_r = 158nm. Conversely, we use t_b = 110nm, t_g = 133nm, t_r = 163nm, i.e., the blue and red LEDs are detuned from the target values ​​by -5nm and +5nm respectively, to illustrate that small deviations from the nominal values ​​are acceptable. The resulting FFP... Figure 12B The curves in the graph show that the differences between the three FFPs are far less than... Figure 12A Significantly significant.

[0170] Next, the LED was tuned to emit D65 chromaticity under normal incidence, and Figure 12CThe graph shows the obtained COA metric, Du'v'. For Figure 12A For the system, COA is significant, while for Figure 12B In the examples provided, COA was significantly reduced. Figure 12D It shows Figure 12C An enlarged version. Figure 12D This embodiment maintains Du'v' below 3E-3 in a 30-degree cone and below 6E-3 in a 45-degree cone. This illustrates how appropriate scaling of the LED size can achieve a low COA. In this example, dispersion of the refractive index with wavelength is neglected, but it can certainly be taken into account (instead of making the physical distance t proportional to the wavelength, the optical distance d1 is proportional to the wavelength).

[0171] This teaching can be extended to other geometric characteristics of LEDs. For example, in some embodiments, LEDs can have sloping or non-vertical sidewalls, and the slope / shape / angle of the sidewalls is configured such that the FFP is substantially equal for R / G / B LEDs. Figure 13 An array 1300 with R / G / B LEDs having non-vertical sidewalls is illustrated. Specifically, blue LED 1310, green LED 1320, and red LED 1330 are all supported by a common substrate 1301. Each LED includes a reflector (reflectors 1312, 1322, and 1332) and a light-emitting region (regions 1311, 1321, and 1331, respectively). The reflectors and active regions are separated by corresponding semiconductor stacks (stacks 1313, 1323, and 1333, respectively). The value of d1 (the optical thickness of stacks 1313, 1323, and 1333) and the geometry of the LEDs are configured to ensure similar FFP between the LEDs to reduce COA. For example, the value of d1 can be selected first according to the resonant conditions of each wavelength, and the rest of the geometry (such as sidewall angles, LED heights, and surface states of various surfaces) can be further configured to make the FFP substantially similar. In some embodiments, some surfaces are roughened. For example, the top surface of an LED is roughened (e.g., by chemical, photochemical, photoelectrochemical, or dry etching). This roughness can interact with the cavity effect to provide a more diffuse FFP with high extraction and less susceptibility to COA artifacts.

[0172] Optical thickness, as the product of physical thickness and refractive index, is expressed in units of length. Embodiments may include LEDs with emitting and reflecting regions (e.g., mirrors). For the reflecting region to cause interference in emitted light with only a few lobes, the optical distance should be sufficiently small. For example, it should be 20 times or less the wavelength of light (e.g., 10 times or less, 5 times or less, 3 times or less, 2 times or less). Alternatively, for visible wavelengths, it can be 10 μm or less (e.g., 5 μm or less, 2 μm or less, 0.5 μm or less).

[0173] For clarity, in the case of LEDs with relatively thick active regions, unless otherwise defined, the optical distance is measured from the physical center of the active region. If the actual center of emission is known (i.e., if it is known that emission originates from the p-side of the active region), those skilled in the art will know to use this value instead of the physical center of the active region.

[0174] This teaching applies to a wide variety of microLEDs. MicroLEDs can be distinguished from conventional LEDs by their size (typically, a lateral dimension of 50 μm or less). In some embodiments, the microLED chip has a characteristic lateral dimension of 50 μm or less (e.g., 25 μm or less, 20 μm or less, 15 μm or less, 10 μm or less, 7 μm or less, 5 μm or less, 3 μm or less).

[0175] Furthermore, microLED chips can be distinguished from conventional LED chips by their light-emitting characteristics, including the ratio of light emitted from the top surface to light emitted from the sides. In some embodiments, the LED has a substantially planar top surface that is parallel to the plane of the light-emitting area, and sidewalls intersecting the top surface (these can be planar or curved, vertical or inclined). All surfaces can be between a semiconductor (including III-V semiconductors, III-nitride compounds, AlInGaAsP compounds) and an external medium (including air or a transparent medium, such as silicone or epoxy resin or sealant). Some microLEDs are characterized by having a relatively large ratio of light emitted from the sidewalls to the total emitted light (for clarity, this ratio can be defined as (light emitted from the sidewalls) / (light emitted from the sidewalls + top)). This ratio can be 20% or greater (e.g., 30% or greater, 40% or greater, 50% or greater, e.g., 55%). This contrasts with conventional thin-film LEDs, where this ratio can be less than 10% (and sometimes less than 5%). This contrasts with other large-size LEDs (such as those that retain a sapphire substrate), where light is emitted from the substrate rather than the semiconductor sidewalls themselves. In some embodiments, the growth substrate of the LED is removed.

[0176] More broadly, some embodiments are characterized by a combination of small lateral dimensions and having at least four faces, each emitting at least 5% (or 10%) of the total emitted light. These can be referred to as multifaceted microLEDs and encompass microLEDs with sidewall emission. These LEDs are also characterized by their physical dimensions, as the surface area of ​​the faces is generally approximately proportional to the light emission from the faces. In some embodiments, each of the faces has a surface area of ​​at least 5% (or 10%) of the total surface area for light emission. In some embodiments, the LED has a top face and a plurality of sidewalls, and the cumulative surface area of ​​the sidewalls is at least 20% (or 50%, 100%) of the surface area of ​​the top face.

[0177] In multifaceted micro-LEDs, light emitted from each facet has a facet-like light field (FFP). These facet-like FFPs combine to form the overall FFP for the micro-LED. This is in... Figure 14A As shown in the figure, Figure 14A An LED 1400 with sloping sidewalls is shown. An FFP (Surface-to-Face) from the top surface 1410 and the sloping sidewalls 1420 is shown. In this embodiment, the surface FFPs are jointly configured to obtain the desired FFP.

[0178] In some embodiments, the LED has a light-emitting region and a reflective region separated by optical thickness, and the ratio of optical thickness to emission wavelength is selected to facilitate emission in a given direction (i.e., at a selected angle in the FFP). The FFPs from each face can cooperate to jointly generate directional emission. For example, a first face has a first preferred emission angle, and a second face has a second preferred emission angle, and these angles are substantially equal (e.g., within + / -20 degrees, less than + / -10 degrees, less than + / -5 degrees). This can be achieved by configuring a combination of optical thickness (such that the cavity effect produces an angular peak within the LED) and configuring the geometry of the sidewalls.

[0179] exist Figure 14B An example is shown, illustrating a multifaceted LED 1450. Optical distances are configured to produce preferential emission in a first and second internal direction within the LED chip, as taught herein. The geometry of the faces (here, the tilt angle of the sloping sidewalls) is configured such that light propagating in the first and second internal directions is emitted to the outside of the LED through both faces in a first and second external direction, the external directions being substantially similar.

[0180] In some embodiments, one or more surfaces (including the top surface and sidewalls) are not optically smooth, but textured or roughened. The RMS roughness can be 100 nm or greater (e.g., 500 nm or greater, 1 μm or greater).

[0181] These teachings can be further applied to multiple LEDs (e.g., emitting at different wavelengths, such as R / G / B LEDs), which are configured to have substantially similar FFPs or the same preferred emission directions.

[0182] In some embodiments, the FFP is characterized by having several angular lobes (e.g., several local maxima at different emission angles). At least two lobes may exist (e.g., three or more lobes, five or more lobes). The contrast of the FFP lobes (defined as the ratio of the maximum intensity at a lobe to the minimum intensity between two lobes) may be 1.1 or greater (e.g., 1.2 or greater, 1.5 or greater, 2 or greater, 5 or greater). This high contrast characterizes an LED with a preferred emission direction. For simplicity, the FFP can be averaged over the azimuth angle before calculating the contrast. The contrast can be calculated within the desired angular emission cone. Figure 15 The illustration shows the definition of contrast between two lobes within an angular emission cone.

[0183] As previously mentioned, the disclosed LEDs (including NW LEDs and micro LEDs) can be incorporated into displays to provide beneficial COA characteristics. Reference Figure 16 The exemplary display 1600 includes an active region 1610 surrounded by a bezel 1620. The active region consists of a pixel array 1630, with each pixel consisting of three sub-pixels (e.g., R, G, B sub-pixels) 1640. Each sub-pixel 1640 consists of one or more LEDs (e.g., multiple NW LEDs), each LED being designed to emit light at the same wavelength. The display 1600 also includes an electronic controller 1650 and a power supply 1660, which are configured to provide drive signals to the pixels. The size and resolution of the display 1600 can be varied as needed. Smaller displays, such as those used in wearable devices (e.g., smartwatches), medium-sized displays (e.g., for smartphones), and larger displays (e.g., desktop monitors, TVs).

[0184] The embodiments may further include a method of configuring a display, the method comprising:

[0185] The emission characteristics of the display are defined by the emission cone, which is characterized by its direction and half-angle.

[0186] Select a display white point and chromaticity tolerance with white chromaticity (e.g., 1E-3);

[0187] The LEDs are configured with non-Lambertian emission patterns FFPr, FFPg, and FFPb such that when the display emits a white dot, all directions within the emission cone have a chromaticity whose Du'v' distance to the white chromaticity is less than the chromaticity tolerance.

[0188] The embodiments further include a device obtained by the method. In this document, a non-Lambertian FFP is an FFP that deviates significantly (e.g., at least 10%) from the direction of a Lambertian FFP within the emission cone.

[0189] The embodiments may include methods of operating LEDs as taught herein, and display systems (including AR and VR displays) that include LEDs as taught herein.

[0190] Examples may include purple LEDs, blue LEDs, green LEDs, yellow LEDs, and red LEDs. For clarity, the corresponding ranges for peak wavelengths are: 400-430nm, 430-480nm, 500-550nm, 550-600nm, and 600-680nm.

[0191] In addition to its effect on directionality, embodiments can be configured to achieve a desired value for the Purcell factor FP (which characterizes the relative change in spontaneous emission rate caused by the photonic environment compared to emission in an infinitely homogeneous semiconductor). As is known in the art, the value of FP can be controlled by the value of d1 and cavity interference effects. In some embodiments, FP is 0.8 or greater (e.g., 0.9 or greater, 1 or greater, 1.1 or greater, such as 1.2).

[0192] Other embodiments are described in the claims.

Claims

1. A full-color display comprising a plurality of pixels, the display having a white dot, an emission direction, and an emission solid angle surrounding the emission direction, the emission solid angle being a semi-cone angle. θ The representation, each pixel includes: This includes red sub-pixels comprising red LEDs having a first geometry, the red LEDs being configured to emit red light into an emission angular range such that the power fraction of the red sub-pixels emitted within the emission solid angle is at least [value missing]. ; The green sub-pixels include green LEDs having a second geometry, the green LEDs being configured to emit green light into an emission angular range such that the power fraction of the green sub-pixels emitted within the emission solid angle is at least [value missing]. ; as well as The blue sub-pixel includes a blue LED configured to emit blue light into an emission angular range such that the power fraction of the blue sub-pixel emitted within the emission solid angle is at least , The red LED, the green LED, and the blue LED are configured such that the white light emitted by the display in any direction within the emission solid angle has a chromaticity difference of less than 0.01 with the white point of the display. .

2. The display according to claim 1, wherein, Each of the blue LED, the green LED, and the red LED includes a light-emitting region and a reflective region located at an optical distance of less than 1 µm from the active region, such that optical interference within each LED causes light to be emitted in a preferred direction in the far field, wherein the respective optical distances between the light-emitting regions and the reflective regions of the red LED, the green LED, and the blue LED are configured such that the respective preferred directions of light emitted from the red LED, the green LED, and the blue LED are within + / -10° of each other.

3. The display according to claim 2, wherein, The respective preferred directions of the light emitted from the red LED, the green LED, and the blue LED are within + / -5° of each other.

4. The display according to claim 2, wherein, The respective preferred directions of the light emitted from the red LED, the green LED, and the blue LED are within + / -3° of each other.

5. The display according to claim 1, wherein: The blue LED includes a blue emitting region that emits light at a blue wavelength, and a blue reflective region that is separated from the blue emitting region by a first optical distance; The green LED includes a green emitting region that emits light at a green wavelength, and a green reflective region that is separated from the green emitting region by a second optical distance; as well as The red LED includes a red emitting region that emits light at a red wavelength, and a red reflective region located at a third optical distance from the red emitting region. The ratios between the first optical distance and the blue wavelength, the second optical distance and the green wavelength, and the third optical distance and the red wavelength are all within 10% of each other.

6. The display according to claim 1, wherein, Each LED has a characteristic lateral dimension of 50µm or less.

7. The display according to claim 6, wherein, At least one of the red LED, the green LED, and the blue LED is a multi-faceted LED, the multi-faceted LED including a top surface and a plurality of sidewalls, wherein light is emitted from each surface during operation of the multi-faceted LED, and the cumulative surface area of ​​the sidewalls is at least 20% of the surface area of ​​the top surface.

8. The display according to claim 7, wherein, At least 20% of the light emitted by the multifaceted LED is emitted through the sidewall surface.

9. The display according to claim 7, wherein, Light emitted from the top surface is preferentially emitted at a first angle relative to the top surface, and light emitted from one of the sidewalls is preferentially emitted at a second angle relative to the top surface, wherein the first angle and the second angle are within + / -10° of each other.

10. The display according to claim 1, wherein, θ Within the range of 5° to 45°.

11. The display according to claim 1, wherein, The red LED, the green LED, and the blue LED are each characterized by a corresponding shape, such that the white light emitted by the display in any direction within the emission solid angle has a chromaticity difference of less than 0.01 with the white point of the display. .

12. The display according to claim 1, wherein, The red LED, the green LED, and the blue LED are characterized by their physical size and emission wavelength, respectively, and the ratio between the physical size and the emission wavelength of each LED is within + / - 10% of each other.

13. The display according to claim 12, wherein, The ratio of each LED is equal.

14. The display according to claim 13, wherein, The physical dimensions are selected from the group consisting of: the radius, height, characteristic vertical dimension, characteristic horizontal dimension of the LED, and the optical distance between the light-emitting area and the reflector.

15. The display according to claim 1, wherein, The white dot has a chromaticity equal to D65.

16. The display according to claim 1, wherein, Each LED includes a metal-based reflector, and according to the equation To configure, among which, It characterizes the phase shift at the interface between adjacent layers of the metal-based reflector and the LED. d It is the optical thickness between the interface and the light-emitting area of ​​the LED, where θ is an angle ranging from 0° to 30°. m It is an integer in the range from 1 to 3.

17. The display according to claim 16, wherein, The optical thickness between the interface and the light-emitting area d Within 10 nm, satisfying the aforementioned equation.

18. A full-color display comprising a plurality of pixels, the display having a white dot, an emission direction, and an emission solid angle about the emission direction, the emission solid angle being a semi-cone angle. θ The representation, each pixel includes: Red sub-pixels comprising a plurality of red NW LEDs configured to emit red light into an emission angular range such that the power fraction of the red sub-pixels emitted within the emission solid angle is at least [value missing]. ; A green sub-pixel comprising a plurality of green NW LEDs configured to emit green light into an emission angular range such that the power fraction of the green sub-pixel emitted within the emission solid angle is at least [value missing]. ;as well as A blue sub-pixel comprising a plurality of blue NW LEDs configured to emit blue light into an emission angular range such that the power fraction of the blue sub-pixel emitted within the emission solid angle is at least , The NW LED is configured such that the white light emitted by the display in any direction within the emission solid angle has a chromaticity difference of less than 0.01 with the white point of the display. .

19. The display according to claim 18, wherein, θ Within the range of 5° to 45°.

20. The display according to claim 18, wherein, The plurality of red NW LEDs, the plurality of green NW LEDs, and the plurality of blue NW LEDs are each characterized by a corresponding shape, such that the white light emitted by the display in any direction within the emission solid angle has a chromaticity difference of less than 0.01 with the white point of the display. .

21. The display according to claim 18, wherein, The plurality of red NW LEDs, the plurality of green NW LEDs, and the plurality of blue NW LEDs are characterized by physical size and emission wavelength, respectively, and the ratio between the physical size and the emission wavelength of each LED is within + / - 10% of each other.

22. The display according to claim 21, wherein, The ratio of each LED is equal.

23. The display according to claim 21, wherein, The physical dimensions are selected from the group consisting of: the radius, height, characteristic vertical dimension, characteristic horizontal dimension of the LED, and the optical distance between the light-emitting area and the reflector.

24. The display according to claim 18, wherein, The white dot has a chromaticity equal to D65.

25. The display according to claim 18, wherein, Each LED includes a metal-based reflector, and according to the equation To configure, among which, It characterizes the phase shift at the interface between adjacent layers of the metal-based reflector and the LED. d It is the optical thickness between the interface and the light-emitting area of ​​the LED, where θ is an angle ranging from 0° to 30°. m It is an integer in the range from 1 to 3.

26. The display according to claim 25, wherein, The optical thickness d between the interface and the light-emitting area is within 10 nm, satisfying the equation.

Images