Light emitting devices and pixels containing the devices, displays and related methods of fabrication

Abstract

The invention provides an optoelectronic light emitting device comprising: a light emitting diode configured to emit first radiation; and a conversion layer extending above the light emitting diode and configured to convert at least part of the first radiation into second radiation by photoluminescence; the conversion layer being laterally delimited by a lateral surface and comprising at least one planar quantum well configured to emit the second radiation; a diffraction grating configured to extract at least part of the second radiation from the conversion layer; the diffraction grating being etched on an upper surface of the conversion layer; the device further comprising a lateral reflector having a reflective face laterally extending with respect to the conversion layer, facing at least part of the lateral surface of the conversion layer, the reflective face of the reflector and the lateral surface of the conversion layer being separated from each other by a material having an optical index n2, and wherein the distance separating the reflective face of the reflector from the lateral surface of the conversion layer is comprised between λ / (n2x40) and λ / (n2x2), where λ is the average wavelength of the second radiation.

Description

Technical Field

[0001] This invention relates to optoelectronic light-emitting devices, pixels comprising a plurality of optoelectronic light-emitting devices, displays comprising a matrix of such pixels, and methods for manufacturing the same. Background Technology

[0002] Generally, a key parameter for display devices is having the largest possible color gamut. In color synthesis, whether additive or subtractive, the color gamut, or range of colors, is a subset of the set of colors that a particular type of material or device can reproduce. Furthermore, lighting and display devices have continuously evolved by increasing the luminous flux and sharpness (i.e., resolution) of light-emitting surfaces. For example, this is reflected in the transformation from cathode ray tubes to liquid crystal displays (LCDs) and now OLED (organic light-emitting diode) screens. For lighting, this includes the shift from incandescent bulbs to LEDs (light-emitting diodes).

[0003] Using LED matrices for display or lighting purposes (e.g., headlights in the automotive industry) requires applying LED technology to semiconductors that emit pixelated matrices, such as those made of gallium nitride (GaN). Semiconductor technology can indeed create matrices with pixels ranging from one micrometer to tens of micrometers. However, GaN-based LEDs primarily emit blue light, with wavelengths around 460 nanometers.

[0004] To obtain other colors, it is necessary to perform light conversion operations within the pixel range. Since blue light has the highest energy in the visible spectrum, it can be converted into lower-energy green or red light through photoluminescence. Photoluminescence is a process in which a substance is excited by absorbing photons and then de-excited by re-emitting lower-energy photons. There are two forms of photoluminescence: fluorescence and phosphorescence. Fluorescence is fast photoluminescence, while phosphorescence is slow photoluminescence; fluorescence is suitable for display and lighting applications.

[0005] Fluorescence can be obtained through fluorophores, for example, in powder form bound to a polymer matrix.

[0006] These fluorophores must be submicroelectronic in size to make it possible to confine them within pixels of a few micrometers. Currently available materials are 3D quantum dots of semiconductors, such as cadmium selenide (CdSe), indium phosphide (InP), or silver indium sulfide (AgInS2), or nanophosphors of oxides, such as YAG:Ce (yttrium aluminum garnet) sulfide or fluoride.

[0007] Furthermore, a limiting factor for light conversion in micropixels is the additional thickness required. For technical reasons, particularly related to the aspect ratio of the structure, this thickness must be kept on the order of the pixel size; for example, for a 3µm pixel, the conversion thickness must be kept on the order of a maximum of 3µm; or for a 10µm pixel, on the order of a maximum of 10µm. Regardless of pixel size, Table 1 below shows the conversion thickness required to absorb 460nm blue light for different phosphor volume loadings in 3D quantum dots of indium phosphide (InP) or cadmium selenide (CdSe) nanophosphors or LuAG:Ce (cerium-doped lutetium aluminum garnet) nanophosphors and composite materials. Table 2 below shows the percentage of absorption achieved for different phosphors and loading rates in Table 1.

[0008] Table 1

[0009] Volumetric loading rate (%) InP CdSe LuAG:Ce 10 8 µm 5 µm 150 µm 20 3 µm 2.5 µm 90 µm 30 1.2 µm 1.2 µm 50 µm

[0010] Table 2

[0011] Volumetric loading rate (%) InP CdSe LuAG:Ce 10 96 94 94 20 98 96 95 30 99 97 95

[0012] Generally speaking, nanophosphors such as LuAG:Ce require a conversion thickness of tens of micrometers to fully or almost fully absorb blue light, which is incompatible with pixel sizes below 10µm.

[0013] Therefore, 3D quantum dots appear to be more promising than nanophosphors because the thickness required for effective conversion is much smaller than that of nanophosphors.

[0014] The article "Monolithic red / green / blue micro-LEDs with HBR and DBR structures" by Guan-Syun Chen et al. (IEEE Photonics Technology Letters, Vol. 30, No. 3, February 1, 2018) further describes a pixel comprising three gallium nitride (GaN) light-emitting diodes: the first without a conversion layer for emitting blue light, the second with a conversion layer for emitting red light, and the third with a conversion layer for emitting green light. The phosphors used are 3D quantum dots. These three diodes are separated from each other at the level of their active layers and, where appropriate, at the level of their conversion layers by light-absorbing layers to limit crosstalk, thereby improving color purity and contrast. A lower Bragg reflector, disposed beneath the substrate, is configured to reflect all three colors. An upper Bragg reflector is disposed on the conversion layers of the second and third diodes to block blue light.

[0015] 3D quantum dots, such as those used in this article, exhibit significant instability when affected by light flux or heat, as is the case in displays or lighting devices. Packaging solutions exist, but these solutions can only achieve weak brightness and low resolution.

[0016] In this context, a robust optoelectronic device is needed that can emit light radiation of a different color than blue in a stable manner and with good efficiency, even if the device is only micrometer-sized. A pixel composed of several such optoelectronic devices is also needed, emitting radiation with different saturated colors to achieve color display. Summary of the Invention

[0017] To at least partially solve the above problems, a photoelectric light-emitting device is proposed, comprising:

[0018] - A light-emitting diode, configured to emit a first radiation, and

[0019] - A conversion layer extending above a light-emitting diode and configured to convert at least a portion of the first radiation into second radiation by photoluminescence, the conversion layer being laterally defined by a side surface, wherein:

[0020] The conversion layer includes at least one planar quantum well configured to emit a second radiation.

[0021] A diffraction grating, configured to extract at least a portion of the second radiation from the conversion layer, is etched onto the upper surface of the conversion layer.

[0022] - A lateral reflector having a reflective surface that extends laterally relative to the conversion layer and faces at least a portion of the side surface of the conversion layer;

[0023] - The reflective surface of the reflector and the side surface of the conversion layer are separated from each other by a material having an optical index n2, and the distance separating the reflective surface of the reflector from the side surface of the conversion layer is between λ / (n2×40) and λ / (n2×2), where λ is the average wavelength of the second radiation.

[0024] By converting the first radiation through photoluminescence in a planar quantum well, the aforementioned absorption thickness problem encountered by nanophosphors can be reduced.

[0025] In fact, a conversion layer thickness typically between 0.1 and 1 micrometer is usually proven sufficient to absorb the majority of the first radiation in the conversion layer (made of semiconductor), thereby enabling the emission of the second radiation to be achieved in the quantum well via radiative recombination. For example, in the case where the first radiation has a spectrum predominantly in the blue, such as in the wavelength range between 430 nm and 490 nm, and the second radiation has a spectrum predominantly in the red, such as in the wavelength range between 600 nm and 700 nm, a thickness between 0.2 and 0.5 micrometers is sufficient for a planar quantum well conversion layer of AlInGaP / InGaP (aluminum indium gallium phosphide / indium gallium phosphide) type to absorb the majority of the first radiation.

[0026] Furthermore, planar quantum well structures do not suffer from the instability issues associated with 3D quantum dots. Such planar quantum well structures do indeed exhibit excellent flux resistance and temperature tolerance because they are typically epitaxially generated at high temperatures.

[0027] On the other hand, without special arrangements, the light power emitted by the planar quantum well conversion layer may prove to be relatively weak, as a significant portion of the second radiation is retained because the material forming the layer has a high optical index value (typically 3 or higher), and it is trapped in the layer by total internal reflection.

[0028] The aforementioned diffraction grating forms a one-dimensional or two-dimensional photonic crystal, allowing a larger portion of the second radiation to escape from the conversion layer and then propagate outside the device. In other words, this grating improves the extraction efficiency of the second radiation. It also makes it possible to apply a propagation direction and / or a given intensity profile and / or a given polarization to the radiation ultimately exiting the optoelectronic device to suit the target application.

[0029] The lateral reflector, in conjunction with the diffraction grating, further enhances the efficiency of extracting the second radiation from the conversion layer. In fact, a portion of the second radiation, propagating in the conversion layer in the form of a guided propagation mode parallel to its upper surface, terminates at a level that laterally defines the side surface of the layer. However, the reflector then sends this portion of the second radiation back into the conversion layer, where it can propagate again parallel to the upper surface of the conversion layer. Therefore, the reflector makes it possible to increase the length of this propagation mode's interaction with the diffraction grating, thereby improving the extraction efficiency of the second radiation.

[0030] From an optical perspective, the lateral reflector thus makes it possible to somehow reduce the diffraction grating and make its effective length greater than its actual geometric length. Due to this reflector, the extraction efficiency of the grating is the same as or nearly the same as that of an "effective" diffraction grating, which will have a much larger lateral extension than a diffraction grating etched on the conversion layer (typically at least twice as large, considering the reduced reflectivity of a lateral reflector, but in the case of an ideal lateral reflector with near-100% reflectivity, this "effective" diffraction may tend to infinity).

[0031] This proves particularly significant when the light-emitting device has a lateral dimension on the micrometer scale, such as around 10 micrometers. In fact, the diffraction grating etched on the upper surface of the conversion layer consists of only a limited number of periodically repeating patterns, for example, around ten patterns in the case of a 1D grating, leading to limited extraction efficiency. Furthermore, when the device has a lateral dimension on the micrometer scale, the thickness of the different layers of the device also decreases (the device itself also has a thickness on the micrometer scale), resulting in the fact that it is generally impossible or undesirable to etch the grating at significant depths. This reduced etching depth also limits the extraction efficiency.

[0032] Furthermore, digital simulation results indicate that for light-emitting devices without lateral reflectors, a diffraction grating 50 to 80 micrometers long, or even longer, is required to extract the main portion of the second radiation.

[0033] Furthermore, as an example, for a 1D grating with only 6 micrometers side length, consisting of seven repeating grating patterns configured to extract radiation with an average wavelength of 640 nanometers (wavelength in vacuum), numerical simulations show that without a side reflector, the far-field extraction efficiency is only about 15%. However, with a side reflector present, extending around the grating, the same grating (6 micrometers side length) allows for a far-field extraction efficiency of 80%.

[0034] Furthermore, maintaining a gap between the conversion layer and the reflective surface is advantageous in that when the reflective surface is metallic (coupling with it would result in power loss), plasma coupling with the metal of the reflector can be avoided.

[0035] Furthermore, the fact that the distance between the conversion layer and the reflecting surface is less than λ / (n²×2) avoids, from an optical perspective, significant discontinuities between the grating itself and the image on the reflecting surface. In other words, this allows for the acquisition of an effective grating with a near-regular grating period, without discontinuities between the actual grating and its different or multiple copies through the reflecting surface. This is significant because the extraction efficiency of an extended and regular "effective" grating is better than that of an extended effective grating with discontinuities between each copy of the initial grating.

[0036] The reflective surface of the reflector can extend around the conversion layer, along the main portion of the conversion layer's perimeter, or even along the entire perimeter of the conversion layer. Then, the reflective surface extends not only on a portion of the side surface of the conversion layer, but also around or almost around that side surface. The perimeter of the conversion layer defines a line that laterally surrounds the conversion layer, making it an edge. The fact that the reflective surface of the reflector extends around the conversion layer along the main portion of its perimeter indicates that the reflective surface of the reflector extends relative to one or more portions of the side surface of the conversion layer, which or these portions extend around the conversion layer for a total length greater than half the perimeter (this length may be a cumulative length) (that is, extending over the total length representing the main portion of the perimeter). As an example, if the conversion layer has a square cross-section with side length c, its perimeter is equal to 4c, and the reflective surface of the reflector then extends relative to one or more portions of the side surface of that layer, which occupy a length greater than 2c around the layer.

[0037] This arrangement allows for a further increase in the “effective” size of the diffraction grating by allowing the second radiation to propagate back and forth several times in the conversion layer, since the reflector then laterally confines the second radiation within the conversion layer.

[0038] The reflector may further include at least a first reflector and a second reflector, each of which extends laterally relative to the conversion layer and faces a portion of the side surface of the conversion layer, the second reflector being located opposite the first reflector relative to the conversion layer, and the first and second reflectors being substantially parallel to each other.

[0039] This arrangement allows the second radiation to propagate back and forth multiple times in the conversion layer through continuous reflection on the reflective surface in a simple manner. Therefore, this arrangement makes it possible to obtain an effective grating length that is much greater than its "actual" geometric length in a simple manner (not only twice as long, but even 10 times, or even 20 times greater due to such a large number of round trips).

[0040] In addition to the features described above, the described optoelectronic devices may include one or more of the following complementary and non-limiting features, which may be considered individually or in all technically possible combinations:

[0041] - The reflector is formed of metal at least on its surface;

[0042] - The upper surface of the conversion layer extends to an area smaller than 1000 square micrometers, or even smaller than 100 square micrometers or even smaller than 10 square micrometers;

[0043] - The light-emitting diode includes an active layer therefrom which it emits first radiation, and the device further includes a lateral confinement system, the lateral confinement system including the lateral reflector extending axially from the active layer to the conversion layer and configured to laterally confine the first radiation;

[0044] -The device further includes a first multilayer reflective filter that extends between the light-emitting diode and the planar quantum well and at least partially reflects the second radiation;

[0045] The device further includes a second multilayer reflective filter that extends over the planar quantum well and at least partially reflects the first radiation.

[0046] Another aspect of the present invention relates to a pixel comprising:

[0047] - As described above, optoelectronic devices, and

[0048] -Additional photoelectric light-emitting devices, including:

[0049] - An additional light-emitting diode, configured to emit the first radiation.

[0050] An additional conversion layer, extending above the additional light-emitting diode and configured to convert at least a portion of the first radiation into a third radiation by photoluminescence, the additional conversion layer being laterally defined by a side surface, the additional conversion layer including at least one additional planar quantum well configured to emit the third radiation, and an additional diffraction grating configured to extract at least a portion of the third radiation from the additional conversion layer, the additional diffraction grating being etched onto the upper surface of the additional conversion layer, and

[0051] - An additional lateral reflector having a reflective surface, at least a portion of which extends laterally to the additional conversion layer and faces the side surface of the additional conversion layer.

[0052] The present invention also relates to a pixel as described above, which further includes another photoelectric light-emitting device, said other photoelectric light-emitting device comprising another light-emitting diode, said other light-emitting diode being configured to emit the first radiation and being:

[0053] - Either there is no conversion layer.

[0054] -Or an additional conversion layer is provided, which is configured to convert at least a portion of the first radiation into a fourth radiation that is different from the second and third radiations.

[0055] The present invention also relates to a display comprising the pixel matrix described above.

[0056] This invention also relates to a method for manufacturing a photoelectric light-emitting device, comprising the following steps:

[0057] - Manufacture light-emitting diodes, which are configured to emit a first radiation, and

[0058] - A conversion layer is fabricated that extends over a light-emitting diode and is configured to convert at least a portion of the first radiation into second radiation by photoluminescence, the conversion layer being laterally defined by a side surface;

[0059] -The step of manufacturing the conversion layer includes the step of manufacturing a planar quantum well configured to emit the second radiation.

[0060] The method further includes the following steps:

[0061] - A diffraction grating is etched on the upper surface of the conversion layer, the diffraction grating being configured to extract at least a portion of the second radiation from the conversion layer, and

[0062] - Manufacture a side reflector having a reflective surface that extends laterally relative to the conversion layer and faces at least a portion of the side surface of the conversion layer. The reflective surface of the reflector and the side surface of the conversion layer are separated from each other by a material having an optical index n2, and the distance separating the reflective surface of the reflector from the side surface of the conversion layer is between λ / (n2×40) and λ / (n2×2), where λ is the average wavelength of the second radiation.

[0063] In this method, the light-emitting diode can be fabricated on a first substrate. The method further includes encapsulating the light-emitting diode and planarizing the contact surface above the light-emitting diode. The step of fabricating the conversion layer includes the following steps:

[0064] - On the second substrate, a stack including the planar quantum well is fabricated.

[0065] - Place the stacked layers on the contact surface, make contact with the contact surface, and fix the stacked layers on the contact surface, and

[0066] - Remove the second substrate.

[0067] The above method can also be performed in such a way that, when the device is presented, the manufactured device includes one or more of the above optional features.

[0068] The present invention also relates to a method for manufacturing pixels, comprising the following steps:

[0069] - Manufacture a first, second, and third light-emitting diode, each configured to emit a first radiation.

[0070] - Fabricating a conversion layer extending above the third light-emitting diode and configured to convert at least a portion of the first radiation into third radiation by photoluminescence, the conversion layer being laterally defined by a side surface, the fabrication step comprising:

[0071] - The step of fabricating at least one planar quantum well configured to emit the third radiation, and

[0072] - The step of etching a diffraction grating on the upper surface of the conversion layer, wherein the diffraction grating is configured to extract at least a portion of the third radiation from the conversion layer.

[0073] - Fabricating a conversion layer extending above the second light-emitting diode and configured to convert at least a portion of the first radiation into second radiation by photoluminescence, the conversion layer being laterally defined by a side surface, the fabrication step comprising:

[0074] - The step of fabricating at least one planar quantum well configured to emit the second radiation, and

[0075] - The step of etching a diffraction grating on the upper surface of the conversion layer, wherein the diffraction grating is configured to extract at least a portion of the second radiation from the conversion layer.

[0076] - For each conversion layer, a side reflector is fabricated having a reflective surface that extends laterally relative to the conversion layer and faces at least a portion of the side surface of the conversion layer. The reflective surface of the reflector and the side surface of the conversion layer are separated from each other by a material having an optical index n2, and the distance separating the reflective surface of the reflector from the side surface of the conversion layer is between λ / (n2×40) and λ / (n2×2), where λ is the average wavelength of the second radiation or the third radiation.

[0077] A better understanding of the invention and its various applications will be gained by reading the following description and examining the accompanying drawings. Attached Figure Description

[0078] The accompanying drawings are for illustrative purposes and do not limit the invention in any way.

[0079] Figure 1 The pixels that implement the teachings of this invention are schematically represented by cross-sectional and side views.

[0080] Figure 2 schematically represented by a top view Figure 1 One of the light-emitting devices of a pixel.

[0081] Figure 3 schematic representation Figure 2 The diffraction grating of the light-emitting device.

[0082] Figure 4 schematically representing by Figure 2 The cross-section of the light beam generated by the light-emitting device.

[0083] Figure 5 Schematic representation of something similar to Figure 2 The device shown is a light beam with a cross section produced by a light-emitting device having another diffraction grating.

[0084] Figure 6 schematically represented by a side view Figure 2 A cross-section of another embodiment of the light-emitting device.

[0085] Figure 7 Schematic representation of what is used in manufacturing Figure 1 The steps of the method for the device shown.

[0086] Figure 8 A more detailed schematic representation Figure 7 Some steps of the method shown.

[0087] Figure 9 Schematic and partial representation using cross-sectional and side views. Figure 1 The pixels are in the intermediate stage of their manufacturing process.

[0088] Figure 10 Schematic and partial representation using cross-sectional and side views. Figure 1 The pixels are in another intermediate stage of their manufacturing process.

[0089] Figure 11 Schematic and partial representation using cross-sectional and side views. Figure 1 The pixels are in another later stage of its manufacturing process.

[0090] Figure 12 Schematic representation using cross-sectional and side views Figure 1 The pixels are located just after the step of creating the conversion layer.

[0091] Figure 13 schematic representation Figure 1 The pixels, in the middle stage of its manufacturing, after the transfer of stacks intended to produce another conversion layer.

[0092] Figure 14 Schematic representation using cross-sectional and side views Figure 1 The pixels are located just after the step of creating another conversion layer. Detailed Implementation

[0093] Within the scope of this document, the terms “light-emitting diode” and “LED” are used indiscriminately.

[0094] This invention particularly relates to a photoelectric light-emitting device, comprising ( Figure 1 ):

[0095] - An electrically pumped light-emitting diode 10R, 10G is configured to emit a first light radiation whose spectrum extends primarily within a given wavelength band, for example, between 400 nm and 490 nm, and has a color, for example, corresponding to blue.

[0096] - Second and third conversion layers 20R, 20G extend above light-emitting diodes 10R, 10G and are configured to convert at least a portion of the first radiation by photoluminescence to obtain another light radiation, for example, red or green.

[0097] These optoelectronic devices 10R and 10G have particularly significant applications in manufacturing "color" pixels 100, such as... Figure 1 As shown, it can emit overall light radiation, and its color can be adjusted electronically.

[0098] To create such a pixel 100, for example, it can be combined as follows:

[0099] - A first photoelectric light-emitting device 1B includes an electrically pumped LED 10B configured to emit the aforementioned first light radiation (which in this particular instance has a color corresponding to blue). The first device 1B does not have a conversion layer.

[0100] - As described above, the second photoelectric light-emitting device 1R has a second conversion layer 20R configured to emit second light radiation via photoluminescence, the second light radiation having, for example, a color corresponding to red and a spectrum mainly located between 590 and 700 nanometers.

[0101] -The third photoelectric light-emitting device 1G as described above, wherein the third conversion layer 20G of the device 1G is configured to emit third light radiation by photoluminescence, the third light radiation having, for example, a color corresponding to green and a spectrum mainly located between 490 and 590 nanometers.

[0102] This "colored" pixel, based on three "blue" light-emitting diodes 10B, 10R, and 10G (two of which have second and third conversion layers 20R and 20G), is particularly significant because the fabrication of "blue" light-emitting diodes is well-mastered today, and they are highly efficient. Furthermore, in this arrangement, only two conversion layers are needed to generate three radiations of different colors, in this particular instance red, green, and blue (which allows for a possible combination of these three radiations for the total emitted radiation, resulting in a color tunable across a wide range of colors).

[0103] However, in alternative solutions, the wavelength range and pixel structure used can differ from those described above.

[0104] For example, electrically pumped LEDs 10R, 10B, and 10G can be configured such that their first emitted radiation is in the near-ultraviolet region (in which case the first device may also have a conversion layer), rather than in the visible light region. Alternatively, they can be configured such that the color of the first emitted radiation corresponds to violet instead of blue.

[0105] Furthermore, a pixel can include multiple basic light-emitting devices other than the three. For example, in addition to the three light-emitting devices mentioned above, it can also include a fourth light-emitting device configured to emit red, of the same type as the second light-emitting device 1R (the pixel then includes two devices for emitting red: one for emitting blue and one for emitting green). Conversely, a pixel can include only two light-emitting devices, such as one for emitting red and one for emitting blue. Generally, another number of subpixels (per pixel), other emission wavelengths, and / or other forms of replication of subpixels within the same pixel can be provided besides those shown above. A group of subpixels forms a pixel, which can be replicated multiple times to form a matrix.

[0106] Furthermore, even when the individual light-emitting devices are combined here to form the colored pixel 100, each of these light-emitting devices can be considered individually and can be further employed separately to obtain a light source. In itself, pixel 100 can also be considered individually and can be employed separately to produce a basic colored light source (in other words, it is not necessarily integrated into the matrix of pixels).

[0107] A color display capable of generating, i.e., displaying, a color image can be produced by combining multiple pixels (e.g., pixel 100 mentioned above). These pixels 100 are then combined in the form of a pixel matrix.

[0108] However, a monochrome display can also be produced by combining multiple identical light-emitting devices (e.g., a second light-emitting device 1R or a third light-emitting device 1G) in a matrix.

[0109] Now refer to Figures 1 to 6 A more detailed description of the structure of pixel 100 will follow. (The following will refer to...) Figures 7 to 14 The method for manufacturing such pixels is described, and more generally, the method for manufacturing a display comprising a pixel matrix 100 of this type.

[0110] Pixel 100, and more generally, displays combining multiple pixels 100 of this type, include a transition stack 90 extending above electrically pumped LEDs 10R, 10G, 10B (above the "blue" LEDs), and the transition stack 90 includes the aforementioned second and third transition layers 20R, 20G (…). Figure 1 The conversion stack 90 has a remarkable structure that enables particularly efficient photoluminescence conversion (here, conversion to green light and conversion to red light).

[0111] First, the lower part of pixel 100 will be described, which includes electrically pumped LEDs 10R, 10G, and 10B ("blue" LEDs). Next, this special conversion stack 90 will be described.

[0112] The first, second, and third electrically pumped LEDs 10B, 10R, and 10G are identical or at least similar to each other. They are arranged on an interconnect stack 70, which allows them to be powered selectively, i.e., independently of each other. The interconnect stack itself is arranged on an integrated control circuit 80.

[0113] Here, each electrically pumped LED 10B, 10R and 10G includes an active layer 15, which includes one or more planar quantum wells for emitting first radiation.

[0114] As described herein, the active layer 15 may be formed of a stack comprising, from top to bottom, the following layers:

[0115] -N-type doped gallium nitride layer 15a,

[0116] -Emitter layer 15b, which includes the aforementioned planar quantum wells or multiple wells, and

[0117] -P-type doped gallium nitride layer 15c layer.

[0118] This stack is deposited on the metal anode contact layer 12. The metal anode contact layer 12 thus extends beneath the p-type doped gallium nitride layer 15c and is in electrical contact with it.

[0119] The first radiation emission is obtained by injecting current into LEDs 10B, 10R, 10G, which then passes through the active layer 15 from the anode region (formed by layer 105c) of the heterojunction to its cathode region (formed by layer 105a).

[0120] "Planar quantum well," or more simply "quantum well," refers here to a structure that includes:

[0121] - A thin central layer, formed of a first semiconductor material having a first bandgap, and

[0122] - Two barrier layers surrounding the central layer, each barrier layer being formed of a second semiconductor material having a second band gap wider than the first band gap (the two barrier layers may be formed of two different semiconductor materials), or, instead of the two barrier layers, a solid semiconductor material surrounding the central layer and having a band gap wider than the first band gap.

[0123] The thin central layer forms a potential well for electrons and / or holes (which explains the name "quantum well" for this structure). The well is planar because the central layer itself is planar, so it extends in a two-dimensional manner across the plane. The thickness of the thin intermediate layer is typically between 3 and 15 nanometers. Its thickness is much smaller than its lateral dimensions, parallel to the layer plane (width and length).

[0124] One or more quantum wells of the emitter layer 15b are configured to emit the aforementioned first radiation, that is, to emit light with a spectrum primarily contained in a wavelength band extending from 400 nm to 490 nm. More precisely, the thickness of the central layer of each of these wells and the thickness of the first semiconductor material forming that central layer are chosen such that the first radiation has the desired spectrum. To obtain blue emission as shown here, for example, indium gallium nitride (InGaN) can be chosen for the first semiconductor material (central layer), and for a portion thereof, the second semiconductor material is gallium nitride (GaN).

[0125] However, the semiconductor material chosen can be different, especially if it is desired to emit the first light radiation in a range different from the wavelengths mentioned above (different from the range of 400 to 490 nanometers).

[0126] Therefore, when it is desired to achieve emission more generally within a specific wavelength range (380 to 700 nm) in the visible light domain, the active layer 15 can be based on a III-V semiconductor material, that is, including elements in the fifth column of the periodic table, such as nitrogen (N), arsenic (As), or phosphorus (P), and related to one or more elements in the third column of the periodic table, such as gallium (Ga), aluminum (Al), and / or indium (In). Thus, III-V semiconductor materials are, for example, gallium nitride (GaN), indium nitride (InN), aluminum nitride (AlN), gallium phosphide (GaP), indium phosphide (InP), gallium nitride (InGaN), aluminum nitride (AlGaN), indium gallium nitride (InGaN), aluminum gallium phosphide (AlGaP), aluminum indium gallium phosphide (AlInGaP), and indium gallium aluminum nitride (InGaAlN). In another case, the first semiconductor material constituting the quantum well or multiple hydrazine central layers can be selected from the following materials: GaN, InP, AlGaN, AlN, AlInGaN, GaP, AlGaP, InGaP, AlInGaP, GaAs, or AlGaAsN, which are binary, ternary, or quaternary compounds of these III-V elements.

[0127] The characteristics of the emitting layer 15b can be further selected so that the half maximum width of the spectrum of the emitted first light radiation is less than a given threshold, for example, less than 30 nanometers.

[0128] In the represented example, each electrically pumped LED, 10R, 10G, 10B includes an upper textured layer, for example made of etched semiconductor, extending above and against the cathode layer 15a. This textured layer forms a photonic crystal, which constitutes a diffraction grating, improving the efficiency of extracting the first radiation from the respective LED.

[0129] In the example represented, the different LEDs 10R, 10G, 10B of pixel 100, and more generally in the display, are separated from each other by substantially vertical trenches (i.e., substantially perpendicular to the plane of the active layer 15 of the LEDs). These trenches are created in the stack (in this particular example, metal anode contact layers 12 and 15a to 15c), which were originally monolithic and have a very large lateral extension. The stack is cut into islands by these trenches to define multiple islands or tiers corresponding to the different LEDs of pixel 100 (or more generally in the display).

[0130] In the example represented, these trenches completely penetrate the metal anode contact layers 12, 15c, 15b, and 15a in the stack. The sides of the trenches are covered with an insulating layer 14, for example, made of silicon oxide. Thus, the sides of each LED 10R, 10G, 10B are covered with this insulating layer 14, here made of silicon oxide. The trenches are then filled with metal, which forms lateral metal walls 16 that optically isolate the LEDs from each other. For each LED, the lateral metal walls 16 laterally restrict the first radiation emitted by the LED, preventing this radiation from propagating to adjacent LEDs. This thus reduces crosstalk between adjacent LEDs, which can affect the purity of the red or green light produced by the second and third light-emitting devices 1R, 1G of pixel 100.

[0131] As shown, the lateral metal wall 16 extends vertically over the entire height of the LED, from the bottom of the metal anode contact layer 12 to the top of the cathode layer 15a. However, in an alternative, the lateral wall could extend, for example, only from the lower surface of the active layer 15 to the upper surface of that layer. In fact, this arrangement also makes it possible to laterally limit the radiation generated by the LED under consideration.

[0132] A contact pickup element 17 is formed at the top of the trench, contacting the lateral metal wall 16. This conductive element has the form of a U-shaped channel, with an opening on its upper side. It is in electrical contact with the 15a layer of N-type doped gallium nitride.

[0133] As previously described, LEDs 10R, 10G, and 10B are fabricated on interconnect stack 70. This stack typically includes one or more insulating layers, one or more metal layers, and one or more conductive via layers.

[0134] Integrated control circuitry 80 is typically formed in and on a semiconductor substrate 81 (e.g., a silicon substrate). For example, control circuitry 80 includes a basic control unit (not detailed) for each LED, electrically connected to the metal anode contact layer 12 of the LED, and includes one or more transistors, enabling control of the current circulating in the LED and / or the voltage applied to the LED. Control circuitry 80 is fabricated, for example, using CMOS technology.

[0135] The cathodes (more precisely, their cathode layers 15a) of different LEDs 10R, 10G, and 10B can be in direct electrical contact with each other, that is, through one or more metal conductors, and each is directly electrically connected to a common contact Ch (Figure 1) forming the electrical contact pads. In other words, from an electrical point of view, different LEDs 10R, 10G, and 10B can have a common cathode.

[0136] In an alternative, the different LEDs of the display can be arranged differently relative to each other. And a different contact system can be used. Thus, for example, it can be specified that the anodes (more precisely, their anode layers) of the different LEDs are in direct electrical contact with each other and directly connected to a common contact Cb, forming a power pad (from an electrical point of view, these different LEDs then have a common anode).

[0137] It should be noted that the 100 pixels of this display are small-sized pixels. In fact, for each of these pixels, the surface area occupied by the photoelectric light-emitting device is less than 3000 micrometers² (3000 square micrometers), or even less than 1000 micrometers² or even less than 100 micrometers².

[0138] Therefore, each of these light-emitting devices has a reduced emitting surface. Thus, it can be specified that, for example, each of these light-emitting devices has dimensions (e.g., width, length, or diameter) on a plane parallel to the mid-plane of the pixel (i.e., parallel to the plane parallel to the substrate 81 that generates the light-emitting diode), each less than 30 micrometers, or even less than 10 micrometers, or even less than 6 micrometers. Each of these devices can, for example, occupy a square surface on the substrate 81 with a side length of 20, 10, or 6 micrometers (e.g.,...). Figure 3(In cases where the dimensions are small, such as 10 micrometers wide and 30 micrometers long (or less), or 6 micrometers wide and 15 micrometers long). The invention is also applicable in a particularly meaningful way to devices of very small dimensions, such as those with side lengths of 3 micrometers or less.

[0139] The transformation stack 90 will now be described in more detail.

[0140] The stack includes:

[0141] -The aforementioned second conversion layer 20R (in red) is located above and aligned with the LED 10R of the second light-emitting device 1R, and

[0142] - The third conversion layer 20G (green conversion layer) is located above and aligned with the LED 10G of the third light-emitting device 1G.

[0143] The second conversion layer 20R of the second light-emitting device 1R is formed by a stack of layers made of semiconductor material. The lateral extension of the stack, that is, the surface it occupies in a plane parallel to the mid-plane of pixel 100, is limited and substantially coincides with the lateral extension of the electrically pumped LED 10R. In other words, the second conversion layer 20R covers LED 10R but does not cover the other LEDs 10B and 10G of pixel 100. The side surface of the second conversion layer 20R, that is, the surface that laterally defines the layer, is marked by reference numeral 24R.

[0144] Similarly, the third conversion layer 20G of the third light-emitting device 1G is formed by stacking layers made of semiconductor material, and its lateral extension is limited to the lateral extension of the LED 10G. The side surface of the third conversion layer 20G, that is, the surface that laterally defines the layer, is marked by reference numeral 24G.

[0145] The second conversion layer 20R extends without direct contact with the LED 10R. On the other hand, the second conversion layer 20R is axially spaced from the upper surface of the LED 10R (i.e., in a direction perpendicular to the midplane of the pixel 100). The space between the LED 10R and the second conversion layer 20R is filled with a transparent filler material 50, such as silicon oxide (SiO2). This filler material 50 also covers and laterally surrounds the second conversion layer 20R. Here, the second conversion layer 20R is thus embedded within the filler material 50.

[0146] In the same manner, the third conversion layer 20G is axially spaced a certain distance from the upper surface of the light-emitting diode 10G, and the space between the light-emitting diode 10G and the third conversion layer 20G is filled with a filling material 50. The third conversion layer 20G is also embedded in itself, that is, surrounded by the filling material 50.

[0147] The structure of the second and third conversion layers 20R and 20G will now be described in more detail.

[0148] The second conversion layer 20R of the second device 1R includes at least one second planar quantum well 21R, which is configured to emit second radiation (here, radiation in the wavelength range of 600 to 700 nm). The semiconductor material used to generate this or these quantum wells may be, for example, aluminum indium gallium phosphide (AlInGaP) and indium gallium phosphide (InGaP).

[0149] The second planar quantum well 21R, or a group of second planar quantum wells 21R of the second conversion layer 20R, is surrounded between the lower layer 22R and the upper layer 23R, forming a cladding of one or more quantum wells. Each of these upper and lower layers 22R, 23R is made, for example, aluminum indium gallium phosphide (AlInGaP).

[0150] A lower layer 22R, located beneath one or more second planar quantum wells 21R, is thus inserted between the LED 10R and these second planar quantum wells 21R. The lower layer 22R has a thickness, for example, between 0.1 and 0.5 micrometers. Similarly, the upper layer 23R may also have a thickness between 0.1 and 0.5 micrometers.

[0151] In practice, the number of second planar quantum wells 21R is typically between 1 and 5. It is important to note that for the conversion from blue to red light, a single quantum well may prove sufficient to convert the majority of the first radiation. During this conversion, the first radiation is absorbed by the barrier layer of the second planar quantum well 21R (or including the upper and lower layers 22R, 23R), and the resulting electron-hole pairs subsequently recombine radiatively in one or more wells, thereby emitting the second radiation.

[0152] The second transition layer 20R is defined at the top by the upper surface 25R (which is the upper surface of the upper layer 23R).

[0153] The upper surface 25R is textured: a second diffraction grating 30R, forming a one-dimensional or two-dimensional photonic crystal, is etched onto this surface. The grating is configured relative to a completely planar interface to improve the efficiency of extracting the second radiation from the second conversion layer 20R without requiring an embossing.

[0154] The second diffraction grating 30R is in Figure 3 And schematically representing the second device 1R in the top view. Figure 2 As can be seen in the text.

[0155] In this example, the second diffraction grating 30R, or in other words, the extraction grating, is a two-dimensional grating. Therefore, it comprises the same pattern 31R that is periodically repeated in two different directions on the grating plane (i.e., the plane of the upper surface 25R). In this particular example, it is a triangular grating (whose basic lattice is, for example, an equilateral triangle). However, in alternatives, it can be a grating with a different lattice, such as a rectangle or hexagon. The periodically repeating pattern 31R can also be apertures or pillars. The diffraction grating in question can also be a one-dimensional grating formed by mutually parallel and regularly spaced (periodically repeated) grooves.

[0156] This grating can improve the extraction efficiency of the second radiation, but it can also give the extracted beam a defined shape and potential polarization. This allows the radiation emitted by the second light-emitting device 1R to be adapted to the target application of the device, providing great flexibility.

[0157] Therefore, for example, if dominant polarization along one axis is desired, a one-dimensional grating is preferably used. With such a grating, the emitted beam will consist primarily of two plates slightly offset from each other by a corner point and having elongated cross-sections that are parallel to each other, such as... Figure 5 As shown in the figure, this figure illustrates the cross-section 200 of a beam extracted by a one-dimensional grating with sides of a few micrometers in the far field and in a plane parallel to the grating plane.

[0158] On the other hand, if it is desired to obtain a beam with a globally isotropic cross-section (e.g., a global disk or toroidal shape), a two-dimensional diffraction grating with a hexagonal or triangular lattice can be used (see [reference]). Figure 4 It is important to note that the fine structure of the far field follows the symmetry of photonic crystals. If the grating is a hexagonal lattice, then a sixth-order symmetry (the pattern repeats every 60°) will be found in the fine structure of the field, even though it has an isotropic shape overall (for a square lattice, there will be a fourth-order symmetry with two orthogonal axes of symmetry). Figure 4 This shows the situation in the far field and in a plane parallel to the grating plane. Figure 3 The cross section 201 of the beam extracted by the triangular grating.

[0159] Based on the desired characteristics of the extracted beam, other types of two-dimensional gratings can be clearly envisioned.

[0160] The size of the diffraction grating also depends on the pattern thickness, the exponential contrast of the grating, and the wavelength to be extracted, which here corresponds to the average wavelength λ of the second radiation. R .

[0161] Generally, the grating period p of the second diffraction grating 30R is... R In λ R / (2.n1) and 2×λR The choice is between n1 and n2, where n1 is the optical index of the high-index material of the grating, which corresponds to the material of the upper layer 23R of the second conversion layer 20R (in this particular example, it is aluminum indium gallium phosphide).

[0162] For example, the grating period p can be specified. R (Or, in the case of a two-dimensional grating, the grating period p) R p R’ ) between λ R / (2.n1) and λ R Between / (n1), or even between λ R / (2.n1) and (3.λ) R Between ) / (4.n1). A relatively small grating period p is chosen in this way. R It is close to the lower limit λ of the conceivable grating period range. R / (2.n1), which allows the diffraction pattern to be repeated more times on the grating surface for a given grating size. This improves the grating extraction efficiency. This improvement is particularly noticeable here because the grating contains a relatively small number of pattern repetitions (e.g., for a one-dimensional grating, this number is between 5 and 50). λ R Here, it is the average wavelength of the second radiation.

[0163] It should be noted that the low-index material of the grating here corresponds to the filler material 50, whose index is indicated as n2.

[0164] from Figure 1 As can be seen, the second and third conversion layers 20R and 20G of the second and third devices have similar structures.

[0165] The third conversion layer 20G of the third device 1G also includes at least one third planar quantum well 21G. In this particular example, the layer includes several third planar quantum wells 21G, typically at least two, or even at least five. It is indeed preferable to use at least two or even at least five third planar quantum wells for this conversion layer because, generally, the blue-to-green conversion is less efficient than the blue-to-red conversion. For certain very specific configurations of the material, such as in the case of converting blue to green using GaN, as shown here, it is even preferable to stack around ten, or even dozens, of InGaN quantum wells in a quasi-totality manner to absorb the incident blue beam.

[0166] Each of these third planar quantum wells 21G is configured to emit third radiation (here corresponding to green, and whose spectrum is mainly in the wavelength range from 490 to 590 nm). The semiconductor material used to generate these third planar quantum wells 21G can be, for example, indium gallium nitride (InGaN) for the well center layer and gallium nitride (GaN) for the barrier layer.

[0167] A set of third planar quantum wells 21G is enclosed between a lower layer 22G and an upper layer 23G made of semiconductor material. The upper layer 23G is, for example, made of gallium nitride (GaN). The lower layer 22G comprises:

[0168] - The first sublayer 22G', optionally made of aluminum gallium nitride (AlGaN) or gallium nitride (GaN), extends immediately below the third planar quantum well 21G located at the lowest position, forming the so-called EBL (electron blocking layer) sublayer, and

[0169] -The second sublayer 22G'' below the first sublayer is made of gallium nitride (GaN) or aluminum gallium nitride (AlGaN).

[0170] The lower 22G and the upper 23G each have a thickness, for example, between 0.1 and 0.8 micrometers.

[0171] For the second device, the third conversion layer 20G of the third device 1G is defined on the upper part by the upper surface 25G (which is the upper surface of the upper layer 23G).

[0172] The upper surface 25G is textured: a third diffraction grating 30G, forming a one-dimensional or two-dimensional photonic crystal, is etched onto this surface. The grating is configured relative to a completely planar interface to improve the efficiency of extracting the third radiation from the third conversion layer 20G without requiring an embossing.

[0173] The third diffraction grating 30G is comparable to the diffraction grating of the second device, except that it is configured to facilitate the extraction of a wavelength λ equal to the average wavelength of the third radiation. G Instead of being configured to favor the extraction of the average wavelength λ of the second radiation. R .

[0174] Specifically, the grating period p of the third diffraction grating 30G G Between λ G / (2.n'1) and 2×λ G Between, rather than between λ R / (2.n1) and 2×λ R Between. n'1 is the optical index of the high-index material of the third diffraction grating 30G, which corresponds to the material of the upper 23G (in this particular instance, gallium nitride GaN).

[0175] In addition to their second and third conversion layers 20R and 20G, the second and third light-emitting devices 1R and 1G each include second and third lateral reflectors 40R and 40G, having second and third reflective surfaces 41R and 41G extending laterally relative to the second and third conversion layers 20R and 20G, and facing at least a portion of the side surfaces 24R and 24G of the second and third conversion layers 20R and 20G under consideration.

[0176] As detailed in the "Summary of the Invention" section, from an optical perspective, the lateral reflector makes it possible to increase the effective surface area of ​​the second and third diffraction gratings 30R and 30G, thereby particularly improving the extraction efficiency of the gratings.

[0177] In the example described here, the second and third lateral reflectors 40R, 40G of each of the second and third light-emitting devices 1R and 1G extend not only relative to a portion of the side surfaces 24R, 24G of their conversion layers, but also around those side surfaces 24R, 24G. In other words, the second and third lateral reflectors 40R, 40G extend around the entire perimeter of the second and third conversion layers 20R, 20G under consideration.

[0178] Here, each of the second and third conversion layers 20R, 20G has a rectangular or optional square cross section in a plane parallel to the intermediate plane of pixel 100.

[0179] In this configuration, the side surfaces 24R and 24G of the second and third conversion layers 20R and 20G are thus formed by four faces, parallel to each other. Furthermore, the perimeter of the second and third conversion layers 20R and 20G, that is, the line that laterally surrounds the layer (in other words, it laterally surrounds the conversion layer), is therefore rectangular. The second and third reflecting surfaces 41R and 41G of the second and third lateral reflectors 40R and 40G extend along this perimeter and also include four faces, in this particular instance, planes surrounding the conversion layer. The second and third reflecting surfaces 41R and 41G also have rectangular cross-sections.

[0180] The different surfaces of the second and third reflective surfaces 41R and 41G extend relative to the different side surfaces of the second and third conversion layers 20R and 20G, respectively, and extend substantially parallel to them (that is, forming an angle of less than 15 degrees or even 5 degrees with them).

[0181] The fact that the second and third reflecting surfaces 41R and 41G surround the entire perimeter of the second and third conversion layers 20R and 20G is advantageous in terms of extraction efficiency. This allows for the lateral confinement of the second or third radiation. Due to this confinement, portions of the second or third radiation propagating in the second and third conversion layers 20R and 20G, parallel to their upper surfaces 25R and 25G, remain within the conversion layers, where they move back and forth. This increases the length of their interaction with the second and third diffraction gratings 30R and 30G, thereby improving extraction efficiency.

[0182] In an alternative, instead of surrounding the conversion layer across its entire perimeter, the reflective surface may surround the layer only over a portion of its perimeter. However, in this case, the reflective surface preferably surrounds the conversion layer over the main portion of its perimeter to achieve good lateral confinement of the second or third radiation as described above.

[0183] The two side surfaces of the second reflective surface 41R of the second light-emitting device 1R are on Figure 1 These are indicated by reference numerals 46R and 47R in the attached drawing. These two surfaces are opposite to the side surfaces 26R and 27R of the second transition layer 20R, respectively.

[0184] like Figure 1 As shown, the two side surfaces 46R and 47R of the second reflecting surface 41R, which are opposite each other to the second conversion layer 20R, are substantially parallel, i.e., they form an angle of less than 15 degrees or even less than 5 degrees. This arrangement allows a portion of the second radiation propagating in the second conversion layer 20R, parallel to its upper surface 25R, to make a significant round trip within the second conversion layer 20R through continuous reflection on the reflecting surfaces 46R and 47R. Therefore, this arrangement enables an effective grating length much greater than its "actual" geometric length l to be obtained. For comparison, for a reflecting surface that would include a single reflecting plane, the effective length obtained would be approximately twice the geometric length l of the second diffraction grating 30R. However, with the two surfaces 46R and 47R substantially parallel to each other, an effective length 10 to 20 times larger than the geometric length l of the second diffraction grating 30R is obtained (depending on the reflection coefficient value of the reflecting surface).

[0185] The two other surfaces (not shown) of the second reflecting surface 41R also form a pair of substantially parallel planar reflecting surfaces. Due to these two reflecting surfaces, the effective width of the grating is halved (here, length is used to specify perpendicularity). Figure 1 The dimensions of the second diffraction grating 30R in the plane.

[0186] The two side surfaces of the reflecting surface 41G of the third device 1G located on either side of the third conversion layer 20G are in Figure 1These are indicated by reference numerals 46G and 47G in the attached diagram. These two surfaces are opposite to side surfaces 26G and 27G of the third conversion layer 20G, respectively. As for the second device, the two side surfaces 46G and 47G are planar and substantially parallel to each other. The same is true for the other two side surfaces (not shown) of the reflecting surface 41G.

[0187] For each of the second and third light-emitting devices 1R and 1G, the second and third lateral reflectors 40R and 40G have rectangular cross sections to accommodate the shapes of the LEDs 10R and 10G and the second and third conversion layers 20R and 20G, which themselves have rectangular cross sections (occupying a rectangular surface in a plane parallel to the substrate 81).

[0188] In alternative solutions, these different components can have shapes other than rectangular, square, or hexagonal. For example, the LED, conversion layer, and side reflector can have triangular or circular cross-sections.

[0189] Regardless of the condition, and regardless of whether the lateral reflector has a rectangular, hexagonal, or other shape, the diffraction grating preferably has the same symmetry as the reflector. For example, for Figures 1 to 3 The second light-emitting device 1R shown, a side reflector with a rectangular cross-section, is reflected relative to... Figure 1 The symmetry of the plane and relative to the perpendicular Figure 1 The symmetry of the vertical second plane remains unchanged. The same applies to the second diffraction grating 30R (see...). Figure 3 It should be noted that the triangular second diffraction grating 30R can also be used with a side reflector with a hexagonal or triangular cross-section, where the reflector and grating have the same symmetry. The latter two cases advantageously make it possible for all gratings to obtain the same number of diffraction elements along the height symmetry direction of the grating (every 60° for a triangular lattice), since the grating or photonic crystal of the triangular or hexagonal lattice fits perfectly to the hexagonal upper surface.

[0190] The fact that gratings and side reflectors have the same symmetry is significant because one or more reflections on the reflective surface of the reflector have the same effect of doubling the diffraction grating from an optical point of view (in other words, by adding the image of the grating onto the grating itself, one can obtain an effective grating with the same structure but a larger extended range).

[0191] Here, each of the second and third lateral reflectors 40R and 40G extends not only relative to the side surfaces 24R, 24G of the respective conversion layers, but also vertically downwards to the upper layer 18 of the LEDs 10R, 10G that generate the first radiation (here, the "blue" radiation). Therefore, the second and third lateral reflectors 40R and 40G extend axially from the upper layer 18 of the LEDs 10R, 10G to the second and third conversion layers 20R, 20G, and even across the entire height of the conversion stack 90 (up to the upper surface 101 of the pixel, which, for example, constitutes a free upper surface).

[0192] The first light-emitting device 1B, which has no conversion layer, also has a lateral reflector 40B. This reflector 40B extends axially from the upper layer 18 of the LED 10B of the device to the same height as the second and third conversion layers 20R and 20G, and even extends to the upper surface 101 of the pixel 100. The area of ​​space surrounded by this first reflector 40B extends above, aligns with, and coincides with the surface of the upper layer 18 of the LED 10B. This area is filled with a filler material 50.

[0193] The different second, third, and first lateral reflectors 40R, 40G, and 40B are produced here in the following manner:

[0194] - By forming substantially vertical grooves in the transition stack 90, and more precisely, by forming grooves in the filler material 50 to serve as support and protection for the second and third transition layers 20R, 20G, and 20b, and

[0195] - By filling these trenches with metal, such as gold (Au), silver (Ag), titanium (Ti), aluminum (Al), or any other metal that can reflect visible light.

[0196] Thus, the trench formed in the conversion stack 90 extends in a straight line with the lateral metal wall 16 that separates the different LEDs 10R, 10G, and 10B from each other.

[0197] Considering the manufacturing methods of the second, third, and first lateral reflectors 40R, 40G, and 40B, certain portions of these reflectors are shared for several different reflectors. For example, the second and third reflectors 40R and 40G include a metal wall shared by these two reflectors at the boundary between the second and third light-emitting devices 1R and 1G. Figure 1 ).

[0198] The metal walls forming the second, third, and first lateral reflectors 40R, 40G, and 40B preferably have a thickness of 0.1 micrometers or more in order to maximize reflectivity and prevent light from passing between two adjacent light-emitting devices.

[0199] The second lateral reflector 40R, together with the lateral metal wall 16 surrounding the active layer 15 of the LED 10R, forms a lateral confinement system for the first radiation emitted by the LED 10R. This confinement system also includes a contact pickup element 17 extending axially from the active layer 15 of the LED 10R to the second conversion layer 20R (and even beyond, given the vertical extension of the second lateral reflector 40R).

[0200] This lateral confinement system guides the first radiation emitted by LED 10R from the active layer 15 to the second conversion layer 20R, where it is converted. This prevents the radiation from reaching the third conversion layer 20G of the third device 1G (which would produce a small amount of parasitic green light instead of red light), or from exiting the pixel 100 without passing through the second conversion layer 20R. Therefore, this lateral confinement of the first radiation improves the purity of the resulting color and limits crosstalk between the different basic light-emitting devices of the pixel 100. Achieving this lateral confinement using a reflective metallic material allows for a greater luminous power at the pixel exit than that obtained through absorptive, non-reflective sidewalls (e.g., made of absorbent black resin).

[0201] In the same manner, the lateral reflector 40G of the third device 1G forms a lateral confinement system for the first radiation with the lateral metal wall 16 surrounding the active layer 15 of the corresponding LED 10G, and the same applies to the lateral reflector 40B of the first device 1B.

[0202] like Figure 1 As shown, for the second and third devices 1R and 1G, the second and third reflecting surfaces 41R and 41G of the second and third lateral reflectors 40R and 40G are separated by a non-zero distance d. R d G It is separated from the side surfaces 24R and 24G of the corresponding conversion layer. The space between these two surfaces is filled with filler material 50 (index n2).

[0203] Here, at a distance of d R Greater than λ R / (n2×40), while the distance d G Greater than λ G / (n2×40). The second and third conversion layers 20R, 20G are separated from the second and third reflective surfaces 41R, 41G (which are metal) to avoid plasma coupling with the metal of the reflector, which would result in power loss (a point confirmed by digital simulation, a difference in finite time domain type).

[0204] Here, at a distance of d R It is still less than λ R / (n2×2), while the distance d G Less than λG / (n2×2). From an optical perspective, this allows for the avoidance of excessive discontinuities between the second and third diffraction gratings 30R, 30G themselves and their images through the reflecting surface (that is, relative to the grating period p). R p G (There is a significant difference). In other words, this makes the obtained effective optical grating close to one with a regular, constant grating period p. R p G The grating is an optically valid grating, and there is no discontinuity between the actual grating and one or more copies formed by it through the second and third reflective surfaces 41R and 41G. This makes sense because the extraction efficiency of the extended and almost completely periodic "effective" grating is better than that of the extended effective grating, which has discontinuities between each copy of the initial grating. To optimize the regularity of grating replication, from an optical point of view, the distance d R and distance d G It can be compared to λ R / (n2×40) to λ R / (n2×2), and λ G / (n2×40) to λ G / (n2×2) is a narrower range of distances. For example, the range in question, determined by digital simulation, can be chosen such that, for patterns located on the edges of the second and third diffraction gratings 30R and 30G, the effective distance between the considered pattern 31R and its image formed by the second and third reflecting surfaces 41R and 41G is equal to the grating period p. R p G The difference does not exceed 20%. Therefore, by adjusting the phase shift caused by reflection, d R and d G Adjusting the distance (during digital simulation) can maximize the power of the extracted radiation (while approaching a near-perfect, efficient reproduction of the grating).

[0205] Figure 6 Another embodiment of the second 'light-emitting device 1R' is schematically and partially shown.

[0206] The second 'light-emitting device 1R' according to the second embodiment is the same as the second device 1R according to the first embodiment, which has been referred to above. Figures 1 to 4 It has been described, except that it further includes first and second multilayer reflective filters 35R and 36R, located below and above the stack including layers 22R, 21R and 23R, respectively.

[0207] A first filter 35R is located between LED 10R and the second planar quantum well 21R. It reflects at least partially the second radiation (red). Its reflectivity, that is, its reflection coefficient for optical power, is, for example, greater than 70%, or even greater than 90%, for the second radiation. The first filter 35R can be further configured to have a high transmittance coefficient, for example, greater than 70%, or even greater than 90%, at the average wavelength of the first radiation (blue) from LED 10R. This allows a significant portion of the first radiation to reach one or more second planar quantum wells 21R, where the first radiation is converted.

[0208] The second filter 36R extends directly above the upper layer 23R of the second conversion layer 20R′. It reflects at least partially the first radiation. For example, its reflectivity for the first radiation is greater than 70%, or even greater than 90%. This prevents a portion of the first radiation that will not be absorbed by the conversion layer from escaping from the second 'light-emitting device 1R′,' thereby improving the color purity of the radiation ultimately emitted by the second 'light-emitting device 1R′. Furthermore, the portion of the first radiation that will not be absorbed and converted is then fed back to the second planar quantum well 21R, where the remaining portion of the first radiation can be absorbed and converted.

[0209] The second filter 36R can also be configured to have a high transmission coefficient for the second radiation emitted by one or more second planar quantum wells 21R, for example, greater than 70% or even greater than 90%, thereby facilitating the extraction of the radiation.

[0210] However, in an alternative, the second filter 36R can be configured to have a moderate transmittance for the second radiation, for example, less than 60%, and a reflectance for the same radiation, for example, between 40% and 70%. In this case, the combination of the first and second filters 35R and 36R forms a Fabry-Perot cavity. The optical thickness of the conversion layer 20R' is then selected such that the cavity reflects the average wavelength λ of the second radiation. R The resonant wavelength exhibits a transmission peak at that wavelength. This arrangement, among other things, allows for fine-tuning of the spectrum of light generated by photoluminescence from the conversion layer 20R'. The tuning and adjustment of the mass coefficient of this Fabry-Perot cavity enable precise definition of the spectral redistribution and the half-maximum width of the extracted radiation.

[0211] According to existing technology, the cavity is adjusted by the phase shift between two consecutive outgoing light beams. It is defined by [the relevant authority / method].

[0212]

[0213] Where n is the refractive index of the cavity; l is the effective thickness of the cavity; and θ is the angle of refraction of light. , where λ R It is the wavelength of the radiation. If the phase shift of continuous light rays... If they are equal to 2π (mod 2π), then they exhibit constructive interference; if their phase shifts... If they are equal to π (mod 2π), they will interfere with each other destructively.

[0214] Each of the first and second filters 35R and 36R is, for example, a Bragg mirror, formed by a combination of thin dielectric layers with a thickness of λ / 4n and high and low indices (where n is the optical index of the thin layer under consideration), such as SiO2 / TiO2 or SiO2 / Si3N4. At 460 nm, the optical indices of these materials are 1.464 / 2.7878 and 1.464 / 2.0751, respectively. The thickness of each of these Bragg mirrors is typically on the order of micrometers, or even less than 1 micrometer. Furthermore, each of the first and second multilayer reflective filters 35R and 36R is, for example, a non-periodic stack with optimized thickness, making it possible to obtain accurate filter transmission profiles.

[0215] It is possible to produce the relevant Fabry-Perot cavity directly by epitaxial growth or by continuous thin-layer deposition (after the epitaxial growth of the grating), which is generally simpler in terms of manufacturing.

[0216] In this embodiment of the second 'light-emitting device 1R', the diffraction grating 30R' that facilitates the extraction of the second radiation can take the form of a hole or pillar passing through a thin layer of the second filter 36R, such as... Figure 6 As shown. Alternatively, this diffraction grating could be fabricated on the surface of the second filter using other methods, such as through less deep etching.

[0217] In an alternative not shown, the third light-emitting device may also be equipped with first and second multilayer reflective filters, arranged in a manner similar to the first and second filters 35R and 36R of the second 'light-emitting device 1R' described above.

[0218] In the case of a third light-emitting device, the first filter (which will subsequently be inserted between the LED 10G and the third planar quantum well 21G) will be configured more precisely to reflect at least a portion of the third radiation, for example, having a reflectivity greater than 70% or even greater than 90% for the third radiation.

[0219] Regarding the second filter that will extend over the third planar quantum well 21G, it will then be configured to operate at the average wavelength λ of the third radiation emitted by one or more third planar quantum wells 21G. GThe second filter has a high transmittance coefficient, for example, greater than 70% or even greater than 90%, thereby facilitating the extraction of the radiation. Furthermore, the second filter can also be configured to have a high reflectivity (greater than 70%, or even greater than 90%) for the first radiation.

[0220] However, the second filter can also be configured to have a moderate transmittance for the third radiation, for example, less than 60%, and a reflectance for the same radiation, for example, between 40% and 70%. In this case, the first and second filter groups will form a Fabry-Perot cavity with an optical thickness such that the cavity transmits light to the average wavelength λ of the third radiation. G It resonates and has a transmission peak for that wavelength.

[0221] Providing these first and second multilayer filters (regardless of whether they form a Fabry-Perot cavity) to the conversion layer of the third device is particularly meaningful because the third planar quantum well 21G, made of indium gallium nitride (InGaN), converts blue to green less efficiently than it converts blue to red. Therefore, utilizing the second filter to recover the first radiation (blue) that was not converted during its first passage through the conversion layer is more meaningful for the third device than for the second device (because this recovery improves the conversion efficiency).

[0222] from Figure 1 As can be seen, the corresponding second and third conversion layers 20R and 20G of the second and third devices 1R and 1G are offset relative to each other. Therefore, they occupy different positions in a direction perpendicular to the plane of the substrate 81. This vertical offset between the second and third conversion layers 20R and 20G is caused by the method of manufacturing the pixel 100.

[0223] In practice, this method first involves fabricating a third conversion layer 20G on a second substrate 300, which is different from the substrate 81 on which the electrically pumped LEDs 10R, 10G, and 10B are fabricated. Next, the third conversion layer 20G is transferred onto the structure of the LEDs 10R, 10G, and 10B. Figure 10 Next, in the same manner, the second conversion layer 20R, which was initially generated on another third substrate 400, is transferred to pixel 100 (Fig. 13), which explains the structure in the successive layers of the conversion stack 90, which is similar to a staircase.

[0224] The method for manufacturing 100 pixels will now be described in more detail.

[0225] As shown in Figure 7, the method includes the following main steps:

[0226] - Step S1: Fabricate the structure of electrically pumped LEDs 10R, 10G, and 10B, then...

[0227] -Step S2: Fabricate the third conversion layer 20G of the third device 1G, then,

[0228] - Step S3: Fabricate the second conversion layer 20R of the second device 1R, then...

[0229] - Step S4: Manufacture the second, third and first lateral reflectors 40R, 40G, 40B.

[0230] Step S1 first includes step S10, which involves fabricating the structure of LEDs 10R, 10G, and 10B as described above. This includes an active layer 15 disposed on an interconnect stack 70, the interconnect stack itself being disposed on an integrated control circuit 80. This circuit is formed on and / or in a substrate 81, which is typically made of silicon and is hereinafter referred to as the first substrate. Optionally, step S10 includes the formation of a textured layer as previously described.

[0231] The method then includes an encapsulation and planarization step S11. This step involves depositing an insulating layer 210, typically formed of a filler material 50 (typically SiO2), either conformally deposited on the textured layer or directly on top of the active layer 15. The method then includes a planarization step of the insulating layer 210, typically by CMP (chemical mechanical polishing). This latter step enables the obtaining of a particularly smooth and flat first contact surface 211, which is ideal for achieving molecular adhesion between this contact surface and another structure. Figure 9 The results obtained at the end of the planarization of insulating layer 210 are shown.

[0232] from Figure 8 As can be seen from this, step S2 in manufacturing the third conversion layer 20G includes the following steps:

[0233] -S20: Fabricate a stack on the second substrate 300, the stack further comprising cladding layers 22G and 23G, and a third planar quantum well 21G enclosed between the two cladding layers (as if sandwiched in the middle) (see Figure 10 ).

[0234] -S21: Transfer the stacked layer onto the structure of LEDs 10R, 10G, 10B, and remove the second substrate 300 ( Figure 11 ).

[0235] -S24: Forming and encapsulation of the third conversion layer 20G.

[0236] The second substrate 300 on which the stack is formed (which will form the third conversion layer 20G after etching) is, for example, a silicon or sapphire substrate. A buffer layer 301 may be deposited on the substrate before the stack is formed.

[0237] Step S21 includes a transfer step S22, followed by a substrate removal step S23.

[0238] In transfer step S22, the free surface of layer 22G contacts the first contact surface 211. Layers 22G, 21G, and 23G are then directly fixed to this surface by molecular adhesion, that is, without the use of additional adhesive material. Alternatively, layers 22G, 21G, and 23G can be fixed to the first contact surface 211 by adhesive bonding.

[0239] In step S23, when the second substrate is made of sapphire, the removal of the second substrate 300 is, for example, by laser lift-off. When the second substrate is made of silicon, it can be removed, for example, by thinning.

[0240] Figure 11 Pixel 100 is shown at an intermediate stage in its manufacturing process, just after removal step S23.

[0241] The next step, S24, involves shaping and encapsulating the third conversion layer 20G, including the following steps:

[0242] -S25: ​​Fabricate the third diffraction grating 30G.

[0243] -S26: A stack of etched layers 22G, 21G, and 23G laterally defines the range of the third conversion layer 20G.

[0244] -S27: Packaging and planarization.

[0245] In step S25, the third diffraction grating 30G is etched onto the free upper surface of layer 23G. This etching typically involves depositing a mask followed by DUV (deep ultraviolet) type etching. Alternatively, this etching can be RIE (reactive ion etching).

[0246] Step S26 may also include depositing a mask followed by DUV lithography, or producing the material via reactive ion etching (RIE). At the end of this step, the third conversion layer 20G forms islands or ridges laterally defined by the aforementioned side surface 24G.

[0247] In step S27, filler material 50 is deposited around and on the islands forming the third conversion layer 20G. The upper surface of the filler layer is then typically polished by CMP (chemical mechanical polishing) to obtain a particularly smooth and flat second contact surface 212.

[0248] Figure 12 The image shows the pixels after step S2 has been performed during the manufacturing process.

[0249] Step S3, which manufactures the second conversion layer 20R, includes steps S30 to S37, which are similar to steps S20 to S27, which have already been described.

[0250] In step S30, a stack of layers 22R, 21R, and 23R, which will form the second conversion layer 20R after etching, is generated on the third substrate 400 formed of gallium arsenide (GaAs).

[0251] In step S32, the stack of layers 22R, 21R, and 23R is transferred and fixed (e.g., by molecular adhesion) onto the second contact surface 212. Figure 13 It shows the pixels that have just completed this transfer step during the manufacturing process.

[0252] During the subsequent removal step S33, selective chemical attack, sulfuric acid (H2SO4), phosphoric acid (H3PO4), or ammonia solution (NH4) can be used. 3(aq) To remove the third substrate 400. In fact, these materials attack gallium arsenide more strongly than aluminum indium gallium phosphide (AlInGaP) or indium gallium phosphide (InGaP).

[0253] Step S35, which manufactures the second diffraction grating 30R, can be performed, like step S25, by depositing a mask and then performing DUV lithography or by reactive ion etching (RIE).

[0254] Step S36, which involves etching the lateral layers 22R, 21R, and 23R, and step S37, which involves encapsulating the second conversion layer 20R, are the same as steps S26 and S27 described above.

[0255] Figure 14 The image shows pixel 100 after step S3 during the manufacturing process.

[0256] Step S4 in manufacturing the second, third, and first lateral reflectors 40R, 40G, 40B includes step S40 of creating trenches that define different light-emitting devices in the conversion stack 90. ​​These trenches can be created by deep RIE (reactive ion etching). Each trench is aligned here with one of the contact pickup elements 17.

[0257] The method then includes step S41, for example, depositing metal in the trench by ECD (electrochemical deposition), followed by a planarization step, for example by CMP, to avoid any short circuits. The method may then include steps of creating contacts Ch (contact pads shared with the cathode) and Cb (contact pads shared with the anode, in other alternatives) at the ends of the anode and cathode lines, typically by means of metal deposition and etching. Figure 1 The result obtained at the end of step S4 is shown.

[0258] Filling trenches with metal in this way is more difficult and more expensive than optically isolating light-emitting devices from each other with black absorbing resin (especially since the metal must be almost liquid to fill the trenches). This production requires specific fabrication methods for nanophotonics (producing quantum wells and photonic crystals) and for microelectronics on silicon, particularly for electrochemical deposition of ECDs.

[0259] In another embodiment, the method of manufacturing pixels may further include the step of manufacturing first and second multilayer reflective filters 35R, 36R, which optionally surround the active portion of the second conversion layer 20R' of the second 'light-emitting device 1R' and / or the third light-emitting device.

[0260] In this configuration, the multilayer reflective filters 35R and 36R of the second 'light-emitting device 1R' can be generated during the fabrication of the stack of layers 22R, 21R, and 23R on the third substrate 400 (therefore, during step S30, before step S31, which involves transferring and removing the third substrate 400). In this configuration, the second filter 36R is generated by deposition on the third substrate 400 before the stack of layers 22R, 21R, and 23R is formed, while the first filter 35R is generated by deposition on layer 22R after the formation of layer 22R. This set of filters 35R and 36R, along with the stack of layers 22R, 21R, and 23R sandwiched between these filters, is then transferred entirely onto the second contact surface 212.

[0261] Similarly, when the third light-emitting device has the first and second reflective filters described above, these filters can be generated during the stacking process of layers 22G, 21G, and 23G on the second substrate 300 (therefore during step S20, before step S21).

[0262] In another scenario, the first and second reflective filters 35R and 36R can be deposited directly above the LED 10R instead of on the third substrate 400 and then transferred. In this case, the first filter 35R can be fabricated by deposition on the second contact surface 221, and then a stack of layers 22R, 21R, and 23R can be transferred onto the filter 35R. The second filter can then be created by deposition on layer 23R, immediately after the transfer and removal step S31 and before the molding and encapsulation step S24 of the second conversion layer 20R.

[0263] In the same manner, the first and second reflective filters of the third device can be deposited directly on the LED 10G, rather than on the second substrate, and then transferred.

[0264] Different implementations and alternatives to the method of manufacturing pixel 100 have been described above. However, it is readily understood that the present invention also relates to methods of manufacturing light-emitting devices, such as second and second 'light-emitting devices 1R, 1R' or third device 1G, considered separately. As described above, such devices can be manufactured by transferring a stack including a converted quantum well onto a first substrate, which has been previously produced on another substrate.

[0265] The present invention also relates to a method for manufacturing the above-described display, the display comprising a matrix of color pixels, such as pixels 100. Such a display can, for example, be manufactured by... Figure 7 The method is similar to the method used to produce it, but in this respect:

[0266] - In step S10, a plurality of triplets of electrically pumped LEDs 10R, 10G, 10B (with the same number of triplets as the matrix) are produced on the first substrate 81.

[0267] - In step S26, in the stack of layers 22G, 21G, 23G (which are initially laterally extended) that have been transferred to the first contact surface 211 beforehand, for example, several islands (as many as pixels) are simultaneously etched, each island forming the third conversion layer 20G of the green light-emitting device.

[0268] Similarly, in step S36, in the stack of layers 22R, 21R, 23R (which initially extend laterally) previously transferred to the second contact surface 212, for example, several islands (as many as pixels) are simultaneously etched, each island forming the second conversion layer 20R of the red light-emitting device, and

[0269] - In step S4, trenches are generated in the conversion stack 90, which laterally define each individual light-emitting device of the pixel matrix, and these trenches are filled with metal.

[0270] It should be noted that, in addition to the methods already mentioned above, different alternatives can be made to the methods described. Therefore, for example, it can be stipulated that step S3 is performed before step S2.

Claims

1. A photoelectric light-emitting device, comprising: - A light-emitting diode, configured to emit a first radiation, and A conversion layer extending above the light-emitting diode and configured to convert at least a portion of the first radiation into second radiation by photoluminescence, the conversion layer being laterally defined by a side surface, the conversion layer including at least one planar quantum well configured to emit the second radiation. A diffraction grating, configured to extract at least a portion of the second radiation from the conversion layer, is etched onto the upper surface of the conversion layer. The device further includes a lateral reflector having a reflective surface that extends laterally relative to the conversion layer, facing at least a portion of the side surface of the conversion layer. The reflective surface of the reflector and the side surface of the conversion layer are separated from each other by a material having an optical index n2, and wherein the distance (d) separating the reflective surface of the reflector from the side surface of the conversion layer is... R d G The value is between λ / (n2×40) and λ / (n2×2), where λ is the average wavelength of the second radiation.

2. The photoelectric light-emitting device according to claim 1, wherein the reflective surface of the reflector surrounds the conversion layer and extends over the main portion of the perimeter of the conversion layer.

3. The photoelectric light-emitting device according to claim 1 or 2, wherein the reflector's reflective surface includes at least a first reflective surface and a second reflective surface, each reflective surface extending laterally relative to the conversion layer and facing a portion of the side surface, the second reflective surface being located opposite the first reflective surface relative to the conversion layer, and the first and second reflective surfaces being substantially parallel to each other.

4. The photoelectric light-emitting device according to claim 1 or 2, wherein the reflector is formed of metal at least on its surface.

5. The photoelectric light-emitting device according to claim 1 or 2, wherein the upper surface of the conversion layer extends to a region of less than 1000 square micrometers, or even less than 100 square micrometers.

6. The photoelectric light-emitting device of claim 1 or 2, wherein the light-emitting diode includes an active layer from which it emits first radiation, and the device further includes a lateral confinement system comprising the lateral reflector extending axially from the active layer to the conversion layer and configured to laterally confine the first radiation.

7. The photoelectric light-emitting device according to claim 1 or 2, comprising a first multilayer reflective filter extending between the light-emitting diode and the planar quantum well and at least partially reflecting the second radiation.

8. The photoelectric light-emitting device according to claim 1 or 2, comprising a second multilayer reflective filter extending above the planar quantum well and at least partially reflecting the first radiation.

9. A pixel, comprising: - The photoelectric light-emitting device according to any one of claims 1-8, and -Additional photoelectric light-emitting devices, including: An additional light-emitting diode is configured to emit the first radiation. An additional conversion layer, extending above the additional light-emitting diode and configured to convert at least a portion of the first radiation into a third radiation by photoluminescence, the additional conversion layer being laterally defined by a side surface, the additional conversion layer including at least one additional planar quantum well configured to emit the third radiation, and an additional diffraction grating configured to extract at least a portion of the third radiation from the additional conversion layer, the additional diffraction grating being etched on the upper surface of the additional conversion layer, and o An additional lateral reflector having a third reflective surface, at least a portion of which extends laterally relative to the additional conversion layer and faces the side surface of the additional conversion layer.

10. The pixel of claim 9, further comprising another photoelectric light-emitting device, said other photoelectric light-emitting device comprising another light-emitting diode, said other light-emitting diode being configured to emit the first radiation, and being: - Either there is no conversion layer. -Or an additional conversion layer is provided, which is configured to convert at least a portion of the first radiation into a fourth radiation that is different from the second and third radiations.

11. A display comprising a matrix of pixels according to claim 9 or 10.

12. A method for manufacturing a photoelectric light-emitting device, comprising the following steps: Step (S10): A light-emitting diode is manufactured, which is configured to emit first radiation. - Step (S2, S3): Fabricate a conversion layer that extends over the light-emitting diode and is configured to convert at least a portion of the first radiation into second radiation by photoluminescence, the conversion layer being laterally defined by a side surface; - The steps (S2, S3) of manufacturing the conversion layer include the steps (S20, S30) of manufacturing a planar quantum well configured to emit the second radiation, and the method further includes the following steps: In step (S25, S35), a diffraction grating is etched on the upper surface of the conversion layer, the diffraction grating being configured to extract at least a portion of the second radiation from the conversion layer. Step (S4) involves manufacturing a side reflector having a reflective surface that extends laterally relative to the conversion layer, facing at least a portion of the side surface of the conversion layer. The reflective surface of the reflector and the side surface of the conversion layer are separated from each other by a material having an optical index n2, and wherein the distance (d) separating the reflective surface of the reflector from the side surface of the conversion layer is... R d G The value is between λ / (n2×40) and λ / (n2×2), where λ is the average wavelength of the second radiation.

13. The method according to the preceding claim, wherein the light-emitting diode is fabricated on a first substrate, the method further comprising encapsulating the light-emitting diode and planarizing a contact surface located above the light-emitting diode, and wherein the step (S2) of fabricating the conversion layer comprises the following steps: - Step (S20): Fabricate a stack including the planar quantum well on the second substrate. - Steps (S22, S32): Place the stacked layers on the contact surface, make contact with the contact surface, and fix the stacked layers on the contact surface. - Step (S22, S32): Remove the second substrate.

14. A method for manufacturing pixels, comprising the following steps: - Step 1 (S10): Fabricate first, second, and third light-emitting diodes, each configured to emit first radiation. - Step two (S2): Fabricating a third conversion layer that extends above the third light-emitting diode and is configured to convert at least a portion of the first radiation into third radiation by photoluminescence, the third conversion layer being laterally defined by a side surface, step two comprising: o Step two (1) (S20), fabricate at least one planar quantum well configured to emit the third radiation, and Step two (2) (S25): Etching a diffraction grating on the upper surface of the conversion layer, the diffraction grating being configured to extract at least a portion of the third radiation from the conversion layer. - Step three (S3): Fabricating a second conversion layer that extends above the second light-emitting diode and is configured to convert at least a portion of the first radiation into second radiation by photoluminescence, the second conversion layer being laterally defined by a side surface, step three comprising: Step three (1) (S30): Fabricate at least one planar quantum well configured to emit the second radiation, and Step 3 (2) (S35): Etch a diffraction grating on the upper surface of the second conversion layer, the diffraction grating being configured to extract at least a portion of the second radiation from the second conversion layer. Step four (S4): For each conversion layer, a side reflector with a reflective surface is manufactured, the reflective surface extending laterally relative to the conversion layer and facing at least a portion of the side surface of the conversion layer, the reflective surface of the reflector and the side surface of the conversion layer being separated from each other by a material having an optical index n2, and wherein the distance (dR, dG) separating the reflective surface of the reflector from the side surface of the conversion layer is between / (n2×40) and / (n2×2), where λ is the average wavelength of the second radiation or the corresponding third radiation.

Images