Optoelectronic device with reduced optical crosstalk

Abstract

The invention relates to an optoelectronic device (1) comprising an array of diodes (100a, 100b, 100c, 100d, 100e) based on a first material. A first diode (100a) and a second diode (100b) of the array have mounted on them a first convex lens (200a) and a second convex lens (200b), respectively. They are separated by a wall (300) having an upper face (301) in contact with the two lenses (200a, 200b). The latter are at a distance from each other, and the wall comprises a stop layer (310) based on a stop material extending from a portion (301*) of the upper face of the wall that is covered neither by the first convex lens nor by the second convex lens. For at least one etch chemistry of the first material, the etch selectivity between the first material and the stop material is greater than or equal to 3:1.

Description

[0001]

[0002] “Optoelectronic device with reduced optical crosstalk”

[0003] TECHNICAL FIELD OF THE INVENTION

[0004] The present invention relates to the field of microelectronics and optoelectronics technologies, in particular the manufacture of arrays of light-emitting or receiving diodes such as LEDs (from the English "Light-Emitting Diode", translating into French as electroluminescent diode), and in particular micro-LEDs. It finds particularly advantageous but non-limiting application in micro-LED-based display systems, or optical communication systems.

[0005] STATE OF THE ART

[0006] Diodes are theoretically a very efficient converter of electrical energy into light energy, or vice versa. They are therefore a promising technology to replace conventional light sources and detectors: high-efficiency diodes would allow considerable energy savings accompanied by other advantages such as robustness, high compactness, long lifetime and good control of color rendering and light intensity.

[0007] One of the crucial challenges in the field of diodes concerns, for emitting diodes, the extraction of the emitted light, and for receiving diodes, the detection of the received light. This extraction or detection is in fact limited by the often very high refractive index of the material constituting the diodes, typically GaN. In the case of emitting diodes, the emitted light remains trapped in the diode by internal reflection. If nothing is done to improve the extraction, only a few percent of the emitted light is actually extracted from the diode.

[0008] Solutions developed in the field of millimeter diodes have achieved satisfactory results. However, these solutions are often difficult or even impossible to implement on micrometer diodes. For example, the fabrication of inverted pyramid-shaped diodes, whose geometry optimizes light extraction, is more complex to achieve on a micrometer scale. The option of integrating a photonic crystal array is not a serious avenue at the micrometer scale because too few diffracting elements can be integrated at this scale to significantly improve extraction. Finally, it is not possible to use structured substrates because the characteristic dimensions of their patterns are incompatible with micrometer diode arrays.

[0009] A first solution applicable to micrometric diodes consists of texturing the surface of the diode in order to break the conservation of the angle at reflection. The protrusions created by texturing act as diffracting elements, which effectively improve light extraction. However, this technique has the significant disadvantage of losing the directivity of the diode, which is particularly disadvantageous when it is coupled to an optical system with reduced numerical aperture.

[0010] A second solution consists of the formation of a microlens above the diode. This solution not only increases light extraction but also improves the diode's directivity.

[0011] Diodes topped with microlenses based on a polymer or dielectric material are described in Eun Kyu Kang et al, IEEE PHOTONICS TECHNOLOGY LETTERS, VOL. 25, NO. 12, JUNE 15, 2013. The publication Dongwoo Kim, Japanese Journal of Applied Physics Vol. 44, No. 1 , 2005, pp. L 18-L 20 particularly shows the effectiveness of GaN microlenses in increasing light extraction.

[0012] However, it has been noted that optical crosstalk occurs between the LEDs described in these documents when they are networked. This phenomenon is very detrimental since it causes a reduction in the contrast ratio between pixels.

[0013] An objective of the present invention is therefore to reduce the phenomenon of optical crosstalk between diodes of the same network surmounted by lenses. Preferably, the solution will make it possible to maintain satisfactory optical properties, particularly in terms of light extraction and directivity.

[0014] SUMMARY OF THE INVENTION

[0015] To achieve this objective, a first object of the invention relates to an optoelectronic device comprising:

[0016] • a substrate having an upper face extending mainly parallel to a plane called the transverse plane,

[0017] • a matrix of light-emitting or photoreceiving diodes arranged on the upper face of the substrate, a first diode of the matrix being surmounted by a first convex lens and a second diode of the matrix being surmounted by a second convex lens, the first convex lens having a lower portion in contact with the first diode and an upper portion on the lower portion, the second convex lens having a lower portion in contact with the second diode, and an upper portion located on the lower portion, the first diode, the second diode and the lower portion of each lens being based on a first material, the upper portion of each lens being based on a second material, the first material having a first refractive index and the second material having a second refractive index, the second refractive index being greater than or equal to the first refractive index,the first diode and the second diode being separated, in the transverse plane, by a wall having an upper face in contact with each of the first lens and the second lens.,

[0018] The first lens and the second lens are spaced apart from each other. The wall preferably comprises a stop layer extending at least from a portion of the upper face of the wall not being covered by the first convex lens or by the second convex lens, the stop layer being based on a material, called a stop material. Furthermore, the first material and the stop material are preferably such that, for at least one etching chemistry of the first material, the etching selectivity between the first material and the stop material is greater than or equal to 3:1, and preferably greater than or equal to 4:1. The first lens and the second lens make it possible to improve light extraction in the case of emitting diodes and light detection in the case of receiving diodes. Their convex shape is ideal for giving the device good light directivity.

[0019] Furthermore, the fact that the first and second diodes are separated by the wall and that the first lens and the second lens are at a distance from each other prevents any optical transmission from one diode to the other. No residual layer that could potentially guide light from one diode to the other creates material continuity between the first diode and the second diode. The two diodes are thus much better optically isolated from each other.

[0020] The invention therefore makes it possible to reduce or even eliminate the phenomenon of optical crosstalk between two neighboring diodes, while maintaining satisfactory optical performance, for example in terms of light extraction and directivity of the source.

[0021] A second subject of the invention relates to a method of manufacturing an electronic device comprising the following steps:

[0022] • Providing a stack comprising: o a substrate having an upper face extending mainly parallel to a plane called the transverse plane, o a matrix of light-emitting or light-receiving diodes arranged on the upper face of the substrate, a first diode of the matrix and a second diode of the matrix being separated, in the transverse plane, by a wall having an upper face, the wall comprising a barrier layer extending at least partially from at least a portion of the upper face of the wall, the barrier layer being based on a material called a barrier material, the first diode and the second diode preferably being based on a first material, o a continuous layer covering at least the first diode, the upper face of the wall and the second diode, preferably the continuous layer comprising a lower portion and an upper portion,the lower portion being in contact with the first diode and with the second diode and being based on the first material, and the upper portion being located on the lower portion and being based on a second material, the first material having a first refractive index and the second material having a second refractive index, the second refractive index being greater than or equal to the first refractive index,

[0023] • Forming in the continuous layer a first convex lens overlying the first diode and a second convex lens overlying the second diode, the first convex lens and the second convex lens being spaced apart from each other, the formation of the first convex lens and the second convex lens comprising selective etching of the first material with respect to the stopping material, the etching selectivity between the first material and the stopping material preferably being greater than or equal to 3:1, and preferably greater than or equal to 4:1.

[0024] The advantages of the device explained above also apply to the method according to the invention. The characteristics of the stop layer also allow the formation of the lenses to be better controlled than in the prior art.

[0025] BRIEF DESCRIPTION OF THE FIGURES

[0026] The aims, objects, as well as the characteristics and advantages of the invention will emerge more clearly from the detailed description of an embodiment thereof which is illustrated by the following accompanying drawings in which:

[0027] Figures 1A to 1L illustrate a first embodiment of the manufacturing method according to the invention in which the barrier layer is metallic.

[0028] Figures 1C, 1E, 1G, 1H, 1I, 1J and 1K refer to a case in which the passivation layer extends over the entire height of the diode sides.

[0029] Figures 1D, 1F, and 1L cover a case in which the passivation layer extends over only part of the height of the diode sides.

[0030] Figure 1M represents the device according to the invention comprising a matrix of five diodes and convex lenses.

[0031] Figures 2A to 2K illustrate a second embodiment of the manufacturing method according to the invention in which the barrier layer is dielectric.

[0032] Figures 2B, 2D, 2F, 2H and 2J show an example in which two neighboring diodes share a common cathode.

[0033] Figures 2C, 2E, 2G, 2I and 2K show an example in which two neighboring diodes have distinct cathodes.

[0034] Figure 3 represents a variant of integration of the common cathode to the two neighboring diodes.

[0035] Figures 4A to 4D illustrate a method of manufacturing secondary walls intended to separate the convex lenses of the device according to the invention.

[0036] Figure 4E is a top view of the device according to the invention when it comprises secondary walls separating the convex lenses.

[0037] Figures 5A to 5C illustrate an embodiment in which the continuous layer is formed by a lower portion and an upper portion based on distinct materials.

[0038] The drawings are given as examples and are not limiting of the invention. They constitute schematic representations of principle intended to facilitate the understanding of the invention and are not necessarily to the scale of practical applications. In particular, the dimensions are not representative of reality.

[0039] DETAILED DESCRIPTION OF THE INVENTION

[0040] Before commencing a detailed review of embodiments of the invention, optional features which may optionally be used in combination or alternatively are set out below:

[0041] According to a preferred example, the first material and the second material are identical, for example are GaN.

[0042] According to a preferred example, the first diode, the second diode, the first convex lens and the second convex lens are based on the same material, for example GaN. This avoids any difference in refractive index between the diodes and the convex lenses. This makes it possible to limit the refraction of light at the interface between the diodes and the lenses. The extraction is thus better than in the case of convex lenses based on a material different from that of the diodes (SiN or a polymer for example). Preferably, the first diode, the second diode and the convex lenses have been formed in the same layer, preferably previously formed by epitaxy. This makes it possible to avoid the presence of an interface between the diodes and the lenses, to avoid any light refraction between these elements and therefore to improve the light extraction.

[0043] According to a preferred example, the first diode and the second diode each have a flank defined by the interface of said diode with the wall, the device further comprising an electrically insulating element extending from the flank of each of the first diode and the second diode.

[0044] According to one embodiment, the device further comprises a second electrical contact in electrical continuity with the first convex lens and with the second convex lens.

[0045] According to one embodiment, the device further comprises a first metal element in electrical continuity with the first convex lens and a second metal element in electrical continuity with the second convex lens, the first metal element and the second metal element being electrically isolated from each other. This makes it possible to single out the control of the two diodes.

[0046] According to one embodiment, the barrier layer is based on a metallic material. By using a metallic material for the barrier layer, this layer is given, in addition to the function of a barrier layer for the formation of the lenses, a function of reconnecting the diodes. This greatly simplifies the structure and manufacturing of the device.

[0047] According to a preferred example, when the barrier layer is based on a metallic material, the barrier layer extends from the upper face of the wall under each of the first lens and the second lens, preferably from the entire extent of the upper face of the wall. This makes it possible to make electrical contact between the barrier layer and the lenses, which is advantageous for making electrical contact resumption of the diodes.

[0048] According to one embodiment, the barrier layer is based on a dielectric material. The etching selectivity between materials commonly used for diodes and lenses and dielectric materials is important. The thickness of the barrier layer can therefore be reduced compared to other materials. This saves space and therefore limits the size of the device.

[0049] According to a preferred example, when the barrier layer is based on an insulating material, a portion of the barrier layer is part of the electrically insulating element.

[0050] According to an advantageous embodiment, the device further comprises a secondary wall extending above the wall, between the first lens and the second lens, the secondary wall extending at least as far as the first convex lens and the second convex lens, preferably beyond the first convex lens and the second convex lens, away from the upper face of the wall in a third direction perpendicular to the transverse plane, the secondary wall being at a distance from the first lens and the second lens. This secondary wall makes it possible to absorb or reflect the light emitted by emitting diodes and which, in the absence of this secondary wall, would be transmitted via the air or the filling material surmounting the diodes. The secondary wall thus makes it possible to reduce the phenomenon of optical cross-talk which may occur in the air or in any material surmounting the diode matrix.

[0051] According to an advantageous embodiment, each convex lens defines a convex surface, said convex surface corresponding to a portion of the external surface of a sphere of diameter D200, each of the first diode and the second diode having a dimension D100, D100 being measured in projection in the transverse plane, perpendicular to the projection of the wall in the transverse plane, with D2oo^

[0052] 2*Dioo. This sizing has been shown to be optimal for maximizing the extraction (or detection) coefficient. The bulk caused by a lens of such a dimension (i.e. with D 2OO >2*DIOO) is also quite reasonable compared to the dimensions of the underlying diode. Furthermore, the center of the sphere is preferably located at the same level in the third direction as the upper face of the wall.

[0053] According to an advantageous embodiment, the first convex lens and the second convex lens are surmounted by an anti-reflection layer. The anti-reflection layer makes it possible to avoid internal refraction of the light emitted by the diodes, in the case of emitting diodes, or refraction towards the medium located above the diodes, in the case of receiving diodes. It therefore makes it possible to improve light extraction (or light detection).

[0054] According to one embodiment of the method according to the invention, the provision of the stack comprises the following steps:

[0055] • Provide an initial stack comprising: o The continuous layer, o A continuous active layer, o The substrate,

[0056] • Carry out a so-called singularization etching from a lower face of the substrate opposite its upper face into the continuous active layer, so as to form the first diode and the second diode in the continuous active layer, and to form a trench between the first diode and the second diode,

[0057] • Form the stop layer in the trench.

[0058] According to one embodiment of the method according to the invention, the first diode and the second diode each have a flank defined by the interface of said diode with the trench, the method further comprising the following steps:

[0059] • Form in the trench an electrically insulating element extending from the sides of the first diode and the second diode,

[0060] • At least partially fill the trench with an electrically conductive material.

[0061] According to one embodiment, the step of providing the stack comprises:

[0062] • provide a primary stack comprising: o the substrate, o the diode matrix, o the lower portion of the continuous layer based on the first material, the lower portion having an upper face opposite the diode matrix,

[0063] • form on the upper face of the lower portion the upper portion of the continuous layer based on the second material, the lower portion and the upper portion together forming the continuous layer.

[0064] According to one embodiment, forming the first convex lens and the second convex lens further comprises selectively etching the second material with respect to the stop material, the etch selectivity between the second material and the stop material being greater than or equal to 3:1.

[0065] According to a preferred example, the formation of the upper portion of the continuous layer is carried out by pulsed laser ablation or by physical vapor deposition.

[0066] According to one embodiment, in a third direction perpendicular to the transverse plane, the lower portion has a thickness e2io and the upper portion has a thickness e 22 o with e 22 o^e2io.

[0067] The term "selective etching with respect to" or "etching exhibiting selectivity with respect to" means an etching configured to remove a material A or a layer A with respect to a material B or a layer B, and exhibiting an etching rate of the material A greater than the etching rate of the material B. The selectivity is the ratio between the etching rate of the material A and the etching rate of the material B. The selectivity between A and B is denoted SA:B.

[0068] This patent application may apply to both a light-receiving diode and a light-emitting diode (as well as to networks of such diodes). The term "diode" is therefore understood to mean either "light-receiving diode" or "light-emitting diode". In this patent application, the terms "light-emitting diode", "light-emitting diode" and "LED" are used synonymously. A "diode" may also be understood to mean a "micro-diode". A "micro-diode" is a diode whose dimensions do not exceed 1 mm (1 mm = 10 -3 m), and typically not 10pm (10 pm = 10 -5 m).

[0069] Two elements are said to be “electrically connected” when they are each in contact with the same continuous electrical connection element having an electrical conduction preferably greater than 10 8 S / m.

[0070] It is specified that, in the context of the present invention, the terms "on", "overcomes", "covers", "underlying", "facing" and their equivalents do not necessarily mean "in contact with". Thus, for example, the deposition, transfer, bonding, assembly or application of a first layer on a second layer does not necessarily mean that the two layers are in direct contact with each other, but means that the first layer at least partially covers the second layer by being either directly in contact with it or by being separated from it by at least one other layer or at least one other element.

[0071] A layer can also be composed of several sub-layers of the same material or of different materials.

[0072] A substrate, a layer, a device, "based" on a material M, is understood to mean a substrate, a layer, a device comprising this material M only or this material M and possibly other materials, for example alloying elements, impurities or doping elements. Thus a material based on a III-N material may comprise a III-N material with added dopants.

[0073] A reference frame, preferably orthonormal, comprising the axes X, Y, Z is shown in figures 1 A, 1 K, 2A, 3 and 4A. This reference frame is applicable by extension to the other figures.

[0074] In this patent application, we will preferably speak of thickness for a layer and height for a structure or device. The height is taken perpendicular to the transverse plane XY. The thickness is taken in a direction normal to the main extension plane of the layer. Thus, a layer typically has a thickness along Z, when it extends mainly along the transverse plane XY, and a projecting element, for example a trench, has a height along Z. The relative terms "on", "under", "underlying" preferably refer to positions taken in the Z direction.

[0075] The terms "substantially", "approximately", "in the order of" mean "to within 10%, preferably to within 5%".

[0076] Optoelectronic device according to the invention

[0077] The device according to the invention will now be described with reference to Figures 1K, 1L, 1M, 2J, 2K, 3, 4D, 4E and 5C.

[0078] The device 1 according to the invention comprises a substrate 10' having an upper face 11' extending mainly in a plane parallel to a transverse plane XY. The transverse plane XY is defined by a first direction X and a second direction Y. The substrate 10' comprises or even corresponds to the dielectric layer 10 which will be described further with reference to the method according to the invention.

[0079] The device 1 further comprises a matrix of diodes 100a, 100b, 100c, 100d, 100e arranged on the upper face 11' of the substrate 10'. A matrix of five diodes 100a, 100b, 100c, 100d, 100e is shown in Figure 1M. The other figures illustrate only two diodes 100a, 100b, called first diode 100a and second diode 100b, but it is understood that the characteristics described below with reference to these two diodes 100a, 100b and to the elements surrounding them can perfectly apply to the other diodes of the matrix, if the device 1 comprises more than two diodes.

[0080] The diodes 100a, 100b are based on a semiconductor material such as GaN. Each diode 100a, 100b comprises an active region 110a, 110b being the site of radiative recombinations of electron-hole pairs making it possible to obtain or absorb light radiation. An active region 110a, 110b typically comprises a plurality of quantum wells, for example formed by layers based on GaN, InN, InGaN, AIGaN, AIN, AllnGaN, GaP, AIGaP, AHnGaP, AIGaAs, GaAs, InGaAs, AIIIAs, or a combination of several of these materials. Each diode 100a, 100b also comprises a P-doped region 120a, 120b, located above the active region 110a, 110b, and making it possible to generate hole-type carriers, and an N-doped region 130a, 130b located above the active region 110a, 110b, and making it possible to generate electron-type carriers.

[0081] The device 1 preferably comprises a first electrode 15a, 15b under each diode 100a, 100b. The first electrodes 15a, 15b are typically at least partially included in the substrate 10'. Each of the first electrodes 15a, 15b may be in contact with an underlying first metal pad 16a, 16b. The first electrodes 15a, 15b may for example be copper-based. The first metal pads 16a, 16b may for example be copper-based.

[0082] Each first electrode 15a, 15b is electrically connected to the diode 100a, 100b which surmounts it when the device 1 is functional. The first electrodes 15a, 15b may be separated from the diodes 100a, 100b, and in particular from the layer 120a, 120b by one or more conductive layers, such as for example a secondary mirror layer 63, a secondary metal layer 61 and the contact layer 30, which may itself be composed of a primary contact layer 30' and a primary mirror layer 30”.

[0083] Each of the diodes 100a, 100b is surmounted by a convex lens 200a, 200b. The assembly consisting of a diode 100a, 100b and the convex lens surmounting it 200a, 200b is designated module 1000a, 1000b.

[0084] Each convex lens 200a, 200b may be formed of a lower portion 210a, 220a and an upper portion 210b, 220b, as illustrated in FIG. 5C. Within each lens 200a, 200b, the lower portion 210a, 210b and the upper portion 220a, 220b may be made from the same material or from distinct materials.

[0085] The lower portions 210a, 210b of the lenses 200a, 200b are preferably based on the first material, already forming the diodes 100a, 100b. The fact that the diodes 100a, 100b and the lower portions 210a, 210b of the lenses 200a, 200b are made of the same material makes it possible to avoid the presence of an optical interface and therefore to reduce optical losses. This makes it possible to obtain high-performance devices.

[0086] The upper portions 220a, 220b are based on a second material, which may or may not be identical to the first material. Generally, preferably, the refractive index of the second material is greater than or equal to the refractive index of the first material. In the case where the first material and the second material are identical, their refractive indices are of course equal. There is therefore no optical interface between the lower 210a, 210b and upper 220a, 220b portions of the lenses 200a, 200b, which limits or even completely eliminates optical losses. This makes it possible to obtain high-performance devices.

[0087] When the first material and the second material are distinct, it is advantageous for the effective index of the second material to be greater than that of the first material because this facilitates the extraction of the light emitted by the diodes 100a, 100b (or the capture of the light received by the diodes 100a, 100b). The lower portions 210a, 210b of the convex lenses 200a, 200b are preferably based on the same material as the diodes 100a, 100b. Advantageously, each diode 100a, 100b and the lower portion 210a, 210b of the convex lens 200a, 200b surmounting it have been formed simultaneously, for example during the same epitaxial growth step. In other words, preferably, no interface is present between each diode 100a, 100b and the convex lens 200a, 200b surmounting it.It is also possible that each diode 100a, 100b and the entire convex lens 200a, 200b surmounting it were formed simultaneously, for example during the same epitaxial growth step. Each module 1000a, 1000b thus preferentially forms a continuous layer.

[0088] The first diode 100a and the second diode 100b are separated in the transverse plane XY by a wall 300. The wall 300 has an upper face 301 extending preferably in a plane parallel to the transverse plane XY. It also has flanks 303 extending at least partially against the diodes 100a, 100b. The wall 300 may extend partially into the substrate 10'.

[0089] Each of the first convex lens 200a and the second convex lens 200b extends above the wall 300. These two convex lenses 200a, 200b are in contact with the upper face 301 of the wall 300. The two convex lenses 200a, 200b are entirely at a distance from each other. In other words, the first convex lens 200a and the second convex lens 200b are not in contact. They do not touch each other at any point. In particular, a portion 301* of the upper face 301 of the wall 300 is not surmounted by either the first convex lens 200a or the second convex lens 200b.

[0090] The wall 300 comprises in particular a stop layer 310. This stop layer 310 extends at least from the portion 301* of the upper face 301 of the wall 300 not being covered by the first lens 200a nor by the second lens 200b.

[0091] The barrier layer 310 can be made from different types of materials.

[0092] According to a first example, the barrier layer 310 is based on a metallic material. This is particularly the case in the embodiments illustrated in FIGS. 1 K, 1 L, 1 M, 3 and 4 D. The metallic material may for example be chosen from TiN, AISi and indium tin oxide (commonly referred to as ITO). The metallic barrier layer 310 may also be composed of a stack of layers of several metallic materials, for example an AISi / Ti N stack. In this last example, the metallic barrier layer 310 may optionally comprise a Ti bonding film or an ITO contact film.

[0093] According to a second example, the barrier layer 310 is based on a dielectric material. This is particularly the case in the embodiments illustrated in FIGS. 2J and 2K. The dielectric material may for example be chosen from the following materials: SiO2, SiN, SiON.

[0094] Whatever the material chosen for the stop layer 310, called the stop material, and whatever the material forming the convex lenses 100a, 100b, called the first material, for at least one etching chemistry of the first material, the etching selectivity between the first material and the stop material is preferably greater than or equal to 3:1, preferably greater than or equal to 4:1.

[0095] Examples of possible pairs of first material and stop material as well as associated etching chemistries are given further on, in the description of the method according to the invention. The dimensions of the stop layer 310 are also described in more detail further on, in the description of the method according to the invention. In addition to separating the diodes 100a, 100b, the wall 300 advantageously has the function of allowing electrical contact to be reestablished between the diodes 100a, 100b. Thus, the wall 300 advantageously comprises at least one electrically conductive element intended to be electrically connected, when the device is in operation, to at least one of the convex lenses 200a, 200b. It will appear in the different embodiments described below that the two diodes 200a, 200b can be electrically connected to a common electrically conductive element or to separate electrically conductive elements.

[0096] The device 1 further advantageously comprises an electrically insulating element at least partially separating the diodes 100a, 100b from the electrically conductive element(s). This electrically insulating element extends for example from the flank 103a, 103b of each diode 100a, 100b towards a central region of the wall 300. In other words, the electrically insulating element extends from the flank 103a, 103b of each of the first diode 100a and the second diode 100b and in a direction perpendicular to this flank 103a, 103b. The different shapes that can be taken by this electrically insulating element will appear in the examples presented below.

[0097] Several examples of the device 1 will now be described. These examples differ from each other in particular by the structure of the wall 300.

[0098] A first example of the device 1 will first be described with reference to FIG. 1K. In this example, the wall 300 comprises a metallic barrier layer 310. Advantageously, this layer 310 extends not only from the uncovered portion 301* of the upper face 301 of the wall 300 but also from other portions of this face 301, and is therefore in contact with the convex lenses 200a, 200b. Thus, a portion of the layer 310 is formed by the portion 301* and at least a portion of the remainder of the upper face 301 of the wall 300. This layer 310 extends towards a central region of the wall 300. Thanks to this, the barrier layer 310 is electrically connected to the diodes 100a, 100b. Preferably, as illustrated in Figure 1K, the barrier layer 310 extends from the entirety of the upper face 301 of the wall 300. The barrier layer 310 extends toward a central region of the wall 300.This makes it possible to maximize the electrical contact recovery area by the stop layer 310.

[0099] The wall 300 further comprises a main metal layer 350. This main metal layer 350 is the electrically conductive element mentioned above. It forms part of the electrode connected to the two convex lenses 200a, 200b via the contact layer 310. The main metal layer 350 may be in contact with a second underlying metal pad 17. The main metal layer 350 may, for example, be copper-based. The second metal pad 17 may, for example, be copper-based.

[0100] The main metal layer 350 typically has an upper face 351 parallel to the upper face 301 of the wall 300. It also has sides 353, for example perpendicular to its upper face 301.

[0101] The metallic barrier layer 310 may cover, for example entirely, the upper face 351 and the sides 353 of the main metallic layer 350. It may be separated from the main metallic layer 350 by one or more conductive layers, such as for example a layer called a mirror layer 330 based on a metallic material such as AISi. The mirror layer 330 preferably also covers the upper face 350 and the sides of the main metallic layer 350.

[0102] The wall 300 further comprises a passivation layer 320 extending from each of its sides 303 towards the main metal layer 350, and typically towards the metal barrier layer 310. The passivation layer 320 is based on an electrically insulating material. It may for example be based on one of the following materials: AI2O3, SiN, AIN, or a combination thereof. It makes it possible to electrically isolate the diodes 100a, 100b in the transverse plane XY.

[0103] In the example illustrated in Figure 1K, the passivation layer 320 extends as much along the third direction Z as the stop layer 310. In other words, it completely covers the sides 313 of the stop layer 310. It is thus flush with the upper face 301 of the wall 300.

[0104] A second example of the device 1 is illustrated in FIG. 1 L. This example differs from the first in that the passivation layer 320 is set back in the third direction Z relative to the stop layer 310. It is also set back relative to the main metal layer 350. In other words, in this example, the passivation layer 320 does not extend over the entire height of the sides 303. Thus, the sides 313 of the metal stop layer 310 are partially in contact with the modules 1000a, 1000b.

[0105] In these first two examples, the electrically insulating element comprises, or even consists of, the passivation layer 320.

[0106] A third example of the device 1 will now be described with reference to FIG. 2J. In this example, the wall 300 comprises a dielectric barrier layer 310. In order to ensure the electrical connection of the modules 1000a, 1000b at the upper face 301 of the wall 300, the barrier layer 310 does not extend from the entirety of this upper face 301. As illustrated in FIG. 2J, typically, the barrier layer 310 forms a discontinuous layer and thus comprises several portions, one of which extends at least partially from the portion 301* not covered by the lenses 200a, 200b. The barrier layer 310 extends towards a central region of the wall 300. This portion is designated barrier portion 310*. It is also conceivable that the barrier layer 310 is continuous and only comprises the barrier portion 310*.

[0107] The wall 300 comprises, as in the first example, a main metal layer 350. This main metal layer 350 is electrically connected on the one hand to the first lens 200a and on the other hand to the second lens 200b. The main metal layer 350 thus typically comprises a first protuberance 350a' and a second protuberance 350b', both electrically conductive. These two protuberances 350a', 350b' are separated, over at least part of their height in the third direction Z, by the stop portion 310*.

[0108] Again, the main metal layer 350 may be in contact with an underlying second metal pad 17. The second metal pad 17 is then electrically connected to both the first lens 200a and the second lens 200b. In this example, the electrical contact resumption at the wall 300 of the modules 1000a, 1000b is therefore done in a common manner.

[0109] The main metal layer 350 has flanks 353 which can be covered with a metal layer 340 and / or a mirror layer 330. The metal layer 340 and the mirror layer 330 can also extend between the protrusions 350a', 350b' and the convex lenses 200a, 200b. The electrical connection between the main metal layer 350 and the convex lenses 200a, 200b is thus made via these layers 340, 330.

[0110] Advantageously, the wall 300 comprises a passivation layer 320 as described in the context of the first example.

[0111] The dielectric barrier layer 310 may comprise portions, called lateral portions 315, extending between the sides 303 of the wall 300 and the second electrical contact 350. These lateral portions 315 typically extend over the entire height of the wall 300.

[0112] In order to ensure electrical insulation between the anode and the cathode of the diode, it is necessary that at least one electrically insulating element separates in the transverse plane XY the lenses 100a, 100b from the second electrical contact 350. A single electrically insulating layer, sufficiently thick, can play this role. The simultaneous presence of the passivation layer 320 and the lateral portions 315 of the stop layer 310 is therefore not necessarily necessary.

[0113] In this third example, the electrically insulating element therefore comprises either the passivation layer 320, or a part of the barrier layer 310, or both.

[0114] A fourth example of the device 1 will now be described with reference to Figure 2K.

[0115] This example differs from the previous example in that the electrically conductive element comprises a first metal element 350a and a second metal element 350b electrically insulated from each other. The first metal element 350a is electrically connected to the first lens 200a and the second metal element 350b is electrically connected to the second lens 200b. The metal elements 350a, 350b may each be in contact with an underlying second metal pad 17a, 17b.

[0116] The first metal element 350a and the second metal element 350b are separated over their entire height by the insulation portion 310* of the dielectric barrier layer 310. Preferably, the insulation portion 310* extends over the entire height of the wall 300 in the third direction Z. The barrier portion 310* thus plays in this example not only a role of barrier layer during the manufacture of the device 1 but also a role of electrical insulator making it possible to individualize the operation of the two diodes 100a, 100b.

[0117] The first metal element 350a and the second metal element 350b each have an upper face 351a, 351b, preferably parallel to the upper face 301 of the wall 300, and sides 353a, 353b, preferably perpendicular to their upper faces 351a, 351b.

[0118] A metal layer 340 may cover, and preferably is in contact with, the upper faces 351a, 351b and the sides 353a, 353b of the metal elements 350a, 350b. The electrical connection between the metal elements 350a, 350b and the convex lenses 200a, 200b may thus be made via the metal layer 340.

[0119] The stop layer 310 may also comprise lateral portions 315 extending between the flanks 303 of the wall 300 and the first metal element 350a on the one hand and the second metal element 350b on the other hand. These lateral portions 315 typically extend over the entire height of the wall 300.

[0120] Advantageously, the wall 300 comprises a passivation layer 320 as described in the context of the first example.

[0121] The wall 300 may also comprise a mirror layer 330 extending between the passivation layer 320 and the lateral portions 315 of the dielectric barrier layer 310. The mirror layer 330 also typically extends over the entire height of the wall 300 in the third direction Z and is thus flush with the upper face 301 of the wall 300.

[0122] The lateral portions 315 of the barrier layer 310 typically completely separate the passivation layer and the mirror layer from the metal layer 340 and the metal elements 350a, 350b.

[0123] In this fourth example, the electrically insulating element comprises, or even consists of, the passivation layer 320. If the mirror layer 330 is omitted, it will also comprise the lateral portions 315 of the barrier layer 310.

[0124] In the third and fourth examples, no metal layer is present between the diodes 100a, 100b and the mirror layer 330. This improves the reflectivity of the mirror layer 330, which contributes to increasing the light extraction. It is optionally possible to add a second metal layer between the diodes 100a, 100b and the mirror layer 330, which will improve the quality of the contact at the expense of the light extraction.

[0125] Furthermore, it should be noted that unlike what was the case in the first two examples, in the third and fourth examples, the electrical connection and stop layer functions are decorrelated.

[0126] Figure 3 illustrates another example structure for the wall 300. In this example, the barrier layer 310 is metallic. The wall 300 includes a second electrical contact 350 electrically connected and preferably in contact with the metallic barrier layer 310.

[0127] The wall 300 further comprises a passivation layer 320 extending from the flanks 103a, 103b of the diodes 100a, 100b towards a central region of the wall 300. It also comprises a mirror layer 330 extending between the passivation layer 320 and the metal stop layer 310. The mirror layer 330 preferably extends to the upper face 301 of the wall 300. It is thus electrically connected to the convex lenses 200a, 200b when the device 1 is in operation.

[0128] In this example, the mirror layer 330 and the metallic barrier layer 310 are further partially separated by an insulating layer 360.

[0129] The resumption of contact of the modules 1000a, 1000b is therefore done both by means of the stop layer 310 and the second electrical contact 350 and by means of the mirror layer 330.

[0130] This embodiment allows, in the case where the stop layer 310 is metallic, that it is not located between the diodes 100a, 100b and the mirror layer 330. This makes it possible to improve the reflectivity of the mirror layer 330, which contributes to increasing the light extraction.

[0131] As illustrated in FIG. 3, the wall 300 may have sides 303 inclined relative to the third direction Z. This is applicable to the other embodiments.

[0132] Various configurations of the 300 wall were presented, but it is understood that other configurations are possible.

[0133] According to an example also illustrated in FIG. 3, the device 1 further comprises an anti-reflection layer 400 covering the upper faces 201a, 201b of the convex lenses 200a, 200b. This is typically a quarter-wave layer known to those skilled in the art. It makes it possible to avoid destructive interference on the surface 201a, 201b of the lenses 200a, 200b.

[0134] The anti-reflective layer 400 has a thickness e 4O o typically measured radially, in a direction passing through the center of the sphere of which the convex surface 201a, 201b of the convex lens 200a, 200b corresponds to a portion. e4oo is chosen as a function of the wavelength emitted or received by the diodes 200a, 200b, the refractive index of the material composing the lenses 200a, 200b and the refractive index of the material surmounting the anti-reflection layer 400 (air, filling material, etc.). e 400must not be too large to avoid modes being guided into the anti-reflection layer, which would be detrimental to light extraction. e400 is typically between 100 and 150 nm depending on the emitted wavelength.

[0135] It is understood that the use of an anti-reflective layer 400 can be combined with any previously described embodiment and does not apply solely to the embodiment of the device shown in FIG. 3.

[0136] According to an example illustrated in Figure 4D, the device 1 further comprises a secondary wall 300' surmounting the wall 300.

[0137] The secondary wall 300' is intended to absorb the radiation emitted by the diodes 100a, 100b which could be transmitted to a higher level, for example an optical element located above the neighboring diode. It thus makes it possible to limit the cross-talk phenomenon which can take place between the diodes and their surroundings by means of the material surmounting the diodes (air, filling material). For example, if a single diode 100a, 100b (corresponding to a pixel) is lit, the presence of the secondary wall 300' makes it possible to render a more intense black in the areas adjacent to this diode. The arrow 3 shown in Figure 4D illustrates the path of radiation emitted by the second diode 100b and reflected by the secondary wall 300'. The arrow 3 is shown for explanatory purposes only. The angles of reflection of the radiation at the different interfaces are not representative of reality.

[0138] The secondary wall 300' may be based on a metallic material absorbing at the emission wavelength for emitting diodes or at the received wavelength for receiving diodes, such as TiN for the visible, or a charged organic material such as those used to produce black matrices (commonly referred to by the English term “black matrix”).

[0139] The secondary wall 300' is located at a distance from the convex lenses 200a, 200b. In other words, it is not in contact with the convex lenses 200a, 200b.

[0140] The secondary wall 300' has a height H300' measured along the third direction Z. It also has a width hoo measured in the transverse plane XY between its two main flanks 303' located opposite the lenses 200a, 200b (in FIG. 4D, along the first direction X). In the classic case of a secondary wall 300' in the shape of a straight block, hoo is invariant along the third direction Z. However, it is also conceivable that the secondary wall 300' has other shapes, such as for example a trapezoidal shape, or that its main flanks 303' have a parabolic profile. In these latter cases, the width of the secondary wall 300' varies along the third direction Z. hoo is then assimilated to the smallest width of the secondary wall 300', typically taken at the level of the upper face 301 of the wall 300.

[0141] The convex lenses 200a, 200b protrude relative to the plane in which the upper face 301 of the wall 300 extends over a height H200, measured along the third direction Z.

[0142] Preferably, the secondary wall 300' extends at least as far as the convex lenses 200a, 200b away from the upper face 301 of the wall 300: H3oo^H2oo. In other words, preferably, the first convex lens 200a and the second convex lens 200b are separated in planes parallel to the transverse plane XY over the entire height H200 by the secondary wall 300'.

[0143] Typically, H300 > 2 pm. Alternatively, preferably, hoo 0.3 pm.

[0144] A shape ratio r=H3oo / hoo of the secondary wall 300' is defined. Preferably, r is greater than 5, preferably greater than 7.

[0145] Converters may optionally be housed in the spaces defined by the secondary wall 300' and the convex lenses 200a, 200b.

[0146] A method of manufacturing such a secondary wall 300' is described further.

[0147] Method of manufacturing the optoelectronic device according to the invention

[0148] A first example of embodiment of a device according to the invention will be described with reference to Figures 1A to 1L. Figures 1A to 1J illustrate an example of a method for obtaining the stack provided at the start of the method according to the invention.

[0149] For clarity, the steps illustrated in Figures 1A to 1L illustrate the fabrication of a two-diode device only. Naturally, these steps can simultaneously produce a diode array comprising many lens-covered diodes from a single initial stack.

[0150] Figure 1A illustrates the provision of an initial stack 1' comprising a support substrate 20, a continuous layer 200, a continuous active layer 100 and a dielectric layer 10. The continuous layer 200 and the continuous active layer 100 are in contact with each other.

[0151] The support substrate 20 has an upper face 21 extending mainly in a plane parallel to the transverse plane XY. The support substrate 20 also has a lower face 22 opposite and preferably parallel to its upper face 21. The support substrate 20 is typically a GaN on Si, or GaN on sapphire growth substrate.

[0152] The continuous active layer 100 is based on a semiconductor material. The continuous active layer 100 comprises an active region 110. This active region 110 is the site of radiative recombinations of electron-hole pairs making it possible to obtain or absorb light radiation. The active region 110 typically comprises a plurality of quantum wells, for example formed by layers based on GaN, InN, InGaN, AIGaN, AIN, AllnGaN, GaP, AIGaP, AHnGaP, AIGaAs, GaAs, InGaAs, AIIIAs, or a combination of several of these materials. The continuous active layer 100 further comprises a P-doped layer 120 and an N-doped layer 130 located on either side of the active region 110.

[0153] Preferably, the continuous layer 200 and the continuous active layer 100 are based on the same material. They have advantageously been formed simultaneously, for example by epitaxial growth on the support substrate 20. They together form a layer called the epitaxial layer 1000, which, when the continuous layer 200 and the continuous active layer 100 have been formed during the same growth step, forms a single continuous layer. The epitaxial layer 1000 is preferably based on GaN. It may also be based on InN, InGaN, AIGaN, AIN, AllnGaN, GaP, AIGaP, AHnGaP, AIGaAs, GaAs, InGaAs, AIIIAs, or a combination of several of these materials. The epitaxial layer 1000 may in particular comprise sub-layers such as buffer layers based on AIGaN.

[0154] The epitaxial layer 1000 has a thickness e1000 along the third direction Z. This thickness e1000 is typically between 2 and 6 pm.

[0155] The continuous layer 200 can also be formed in two stages. For example, a lower portion 210 of the continuous layer 200 can be formed by epitaxy, for example on the support substrate 20 and preferably at the same time as the continuous active layer 100. An upper portion 220 of the continuous layer 200 can then be deposited on the upper face 211 of the lower portion 210. The lower portion 210 and the upper portion 210 thus form the continuous layer 200.

[0156] The deposition of the upper portion 220 can for example be carried out by pulsed laser ablation (in English Pulsed Laser Deposition, PLD). This technique has the advantage of being able to be carried out at low temperature, typically at temperatures less than or equal to 400°C. This is advantageous when metal is present in the stack. This technique can also be carried out at high pressure, typically at pressures greater than or equal to 0.01 mbar, which is advantageous because it allows a deposition to be carried out with little damage to the underlying lower portion and also allows better control of the composition of the deposited material. This deposition can also be carried out by physical vapor deposition (in English Physical vapor deposition, PVD).

[0157] The lower portion and the upper portion forming the continuous layer 200 may or may not be based on the same material. The lower portion 210 is formed from the first material, and the upper portion from a second material which may be identical to or different from the first material. For example, the two portions 210, 220 may be based on GaN. According to another example, the first material is based on GaN and the second material is based on TiO2 or InGaN, preferably with a high concentration of indium. Generally, preferably, the second material has a refractive index higher than that of the first material. The second material being above the first material within the lenses at the end of the process, this allows for better light extraction.

[0158] Figure 5A illustrates the continuous layer 200 formed by two portions 210, 220 based on distinct materials. The step illustrated in Figure 5A is thus a variant of that illustrated in Figure 1A, where the two portions 210, 220 are based on identical materials (and were obtained by a single manufacturing step or two distinct steps).

[0159] The advantage of forming the continuous layer 200 in two stages, i.e. by forming a lower portion and then an upper portion, is to achieve a thickness of the continuous layer 200 greater than what could have been achieved conventionally with a single deposition, typically with an epitaxial step. Indeed, epitaxial growth from a support substrate is highly subject to constraints and it is difficult to obtain a thick epitaxial layer. In the typical case of GaN growth on a Si substrate, it is difficult to obtain a GaN layer of reasonable quality going beyond a thickness of 3 or 4 pm, 4.5 pm being a maximum. Consequently, taking into account that part of the epitaxy is used to form the active zones of the diodes (here to form the continuous active layer 100), the height of the lenses formed is only of the order of 3 pm.

[0160] By using a PLD or PVD technique following the epitaxy of the lower portion, it is possible to form an upper portion providing additional thickness to the continuous layer 200.

[0161] The portion formed by epitaxy, that is to say typically the continuous active layer 100 and the lower portion 210 of the continuous layer 200, has, along the third direction Z, a thickness typically between 2 pm and 3 pm.

[0162] The layer formed by PLD or PVD, that is to say the upper portion 220 of the continuous layer 200, has, along the third direction Z, a thickness 6220 typically between 2 pm and 3 pm.

[0163] The assembly formed by the continuous active layer 100 and the continuous layer 200 thus has a thickness at least equal to 4 pm, preferably greater than or equal to 5 pm, preferably greater than or equal to 6 pm. In this way, it is possible to form lenses having a very significant height (typically greater than or also 4 pm or even 6 pm).

[0164] PLD and PVD techniques also do not require a major thermal budget and make it possible to maintain good crystallinity. The dielectric layer 10 has an upper face 11 extending mainly in a plane parallel to the transverse plane XY. The dielectric layer 10 also has a lower face 12 opposite and preferably parallel to its upper face 11. The dielectric layer 10 may for example be based on SiO2 or SiN. It has a thickness eio along the third direction Z. This thickness eio may for example be substantially equal to 300 nm.

[0165] The initial stack may comprise other layers, which may in particular be interposed between the dielectric layer 10 and the continuous active layer 100. For example, an electrically conductive layer called a contact layer 30 may be inserted between the dielectric layer 10 and the continuous active layer 100. The contact layer 30 may for example be based on AISi or may be a stack of sub-layers of AISi and ITO. As illustrated in FIG. 1B, an etching is then carried out from the lower face 12 of the dielectric layer 10. This etching makes it possible to form a trench 50 passing entirely through the dielectric layer 10 in the third direction Z and extending into the epitaxial layer 1000. The trench 50 does not pass entirely through the epitaxial layer 1000 in the third direction Z. This etching step may comprise several successive etching sub-steps.A first etching sub-step with a stop on the contact layer 30 can firstly make it possible to create an opening in the dielectric layer 10. A second etching sub-step can then make it possible to remove an underlying portion of the contact layer 30 and of the epitaxial layer 1000, thus completing the formation of the trench 50.

[0166] An optional chemical treatment step, for example a wet chemical treatment based on tetramethylammonium hydroxide (commonly referred to as TMA) or potassium hydroxide (KOH) for example. This makes it possible to reduce the defects of the flanks 53 of the trench 50.

[0167] The formation of the trench 50 makes it possible to single out the continuous active layer 100 into a first diode 100a and a second diode 100b. These two diodes 100a, 100b are separated in the transverse plane XY by the trench 50. Each diode 100a, 100b has a flank 103a, 103b merged with a flank 53 of the trench 50. The formation of the trenches 50 also makes it possible to single out the active region 110, the P-doped layer 120 and the N-doped layer 130.

[0168] The sides 53 of the trench 50 are shown in the figures parallel to the third direction Z, but it is understood that they can be inclined relative to this direction Z, for example by an angle between 75° and 85°.

[0169] As illustrated in Figure 1B, a passivation layer 320 is then deposited against the sides 53 and on the bottom 51 of the trench 50. The passivation layer 320 can be deposited in a conformal manner. It can for example be based on one of the following materials: AI2O3, SiN, AIN.

[0170] As illustrated by the transition from Figure 1 B to Figure 1 C, the portion of the passivation layer 320 covering the bottom 51 of the trench 50 is then removed. This makes it possible to again expose an intermediate surface 1001 of the epitaxial layer 1000. The portions of the passivation layer 320 extending from the sides 103a, 103b of the diodes 100a, 100b are retained.

[0171] According to a first example, the epitaxial layer 1000 does not undergo an over-etching step at this stage and the rest of the method takes place from the assembly as illustrated in FIG. 1C (option 1). According to a second example illustrated in FIG. 1D, an over-etching step of the epitaxial layer 1000 is implemented so as to extend the trench 50 and expose internal flanks 1003 of the epitaxial layer 1000 (option 2). The internal flanks 1003 of the epitaxial layer 1000 are then in the extension of the passivation layer 320.

[0172] The passivation layer 320 has a thickness e32o, measured perpendicular to the sides 103a, 103b of the diodes 100a, 100b. e32o is measured from the sides 103a, 103b of the diodes 100a, 100b to the internal sides 323 of the passivation layer 320 opposite the sides 103a, 103b of the diodes 100a, 100b. The role of the passivation layer 320 is to electrically insulate the anode and the cathode of the same diode 100a, 100b. It must therefore be sufficiently thick to withstand the potential difference between the anode and the cathode of the diodes 100a, 100b during operation of the device 1 (typically greater than 4V). To limit the size of the device, however, we will avoid unnecessarily increasing the thickness of the passivation layer 320. Thus, typically, e32o is between 50 and 100 nm.

[0173] The intermediate surface 1001 and the internal sides 1003 of the epitaxial layer

[0174] 1000 preferably undergo at this stage of the process a chemical surface preparation step, for example based on hydrogen fluoride (HF), buffered oxide etch (commonly referred to as BOE), or tetramethylammonium hydroxide (TMAH).

[0175] As illustrated in Figures 1 E and 1F, a stop layer 310 based on a metallic material is then deposited in this embodiment against the internal flank 323 of the passivation layer 320 and on the intermediate surface 1001 of the epitaxial layer 1000. The metallic stop layer 310 is preferably deposited conformally. It may for example be based on TiN or ITO. It may also be composed of a stack of sub-layers of AISi and TiN. In this last example, the metallic stop layer 310 may optionally comprise a Ti bonding film or an ITO contact film.

[0176] The metallic barrier layer 310 has an external flank 314 located opposite and preferably in contact with the internal flank 323 of the passivation layer 320. It also has an internal flank 313, opposite its external flank 314.

[0177] The metallic barrier layer 310 has a thickness e3io, measured along the third direction Z for its portion covering the intermediate surface

[0178] 1001 of the epitaxial layer 1000. Typically, its thickness is substantially the same or less in its portions covering the internal sides 323 of the passivation layer 320.

[0179] The thickness e3io of the metallic barrier layer 310 is chosen according to the material(s) composing it and in particular their selectivity with respect to the material(s) of the continuous layer 200.

[0180] For example, in the case of an ITO-based metal 310 barrier layer, e3io may be substantially equal to 60 nm. When the metal 310 barrier layer is TiN-based, e3io may be substantially equal to 200 nm.

[0181] The value of e3io for a given barrier material is determined in particular as a function of the total thickness variation (commonly referred to as TTV) of the continuous layer 200 appearing during the manufacturing steps, as will be discussed further. The thicknesses mentioned above were determined for a TTV value of 300 nm.

[0182] Figure 1 E and Figure 1 F correspond respectively to the deposition of the stop layer 310 on the assemblies illustrated in Figures 1C and 1 D.

[0183] In the example illustrated in Figure 1E, the metal barrier layer 310 is in contact with the epitaxial layer 1000 only at its intermediate surface 1001 (option 1). In the case of an epitaxial layer 1000 based on raw GaN from a GaN support substrate 20 on Si, the contact between the metal barrier layer 310 and the intermediate surface 1001 of the epitaxial layer 1000 is made in Ga polarity.

[0184] In the example illustrated in Figure 1F, the metal stop layer 310 is furthermore in contact with the internal flanks 1003 of the epitaxial layer 1000 previously exposed by the over-etching (option 2). This creates an additional contact zone, which can be designated lateral contact, between the metal stop layer 310 and the epitaxial layer 1000. This example thus makes it possible to increase the contact surface between the metal stop layer 310 and the epitaxial layer 1000.

[0185] When the metallic 310 barrier layer is ITO-based, option 1 is preferred.

[0186] Figures 1G to 1K illustrate the continuation of the process in the case of option 1, but it is understood that the same steps apply when the second option is implemented.

[0187] As illustrated in FIG. 1G, an opening 60 is then formed in the dielectric layer 10 above each diode 100a, 100b. These openings 60 are intended to accommodate the first electrodes 15a, 15b electrically connected to the diodes 100a, 100b. The openings 60 advantageously open onto the contact layer 30. They are typically formed by lithography.

[0188] Following the formation of the openings 60 in the dielectric layer 10, a secondary metal layer 61 is advantageously deposited against the sides and in the bottom thereof, followed by a secondary mirror layer 63. The secondary metal layer 61 is preferably based on the same metallic material as the metallic stop layer 310. In this case, it can advantageously be provided that, during the deposition of the secondary metal layer 61, this material is again deposited in the trench 50 in order to increase the thickness of the metallic stop layer 310. It is also conceivable that the stop layer 310 is entirely formed after the formation of the openings 60, at the same time as the secondary metal layer 61.

[0189] As illustrated in Figure 1H, a layer called mirror layer 330 is advantageously deposited in the trench 50, preferably in a conformal manner. The mirror layer 330 has external flanks 334 located opposite and preferably in contact with the internal flanks 313 of the metallic barrier layer 310. It also has internal flanks 333, opposite its external flanks 334. The mirror layer 330 is based on a metallic material such as AISi.

[0190] The trench 50 and the openings 60 are then filled with a conductive material in order to form a main metal layer 350 in the trench 50 and a first electrode 15a, 15b in each opening 60. For this, it is first possible to deposit seed layers (not shown, typically of the Ti / TiN / Cu type), then a damascene copper brick produced by filling with copper by electrochemical deposition (commonly referred to as ECD). Chemical-mechanical polishing of the Cu can then be implemented in order to free the lower face 12 of the dielectric layer 10 from any trace of Cu outside the trench 50 and 60. For this, the deposition of the stop layer in two stages as illustrated in FIG. 1G / 1H can allow better control of the CMP stop.

[0191] A passivation bilayer 10', for example of the SiN / SiO2 type, is then advantageously deposited, then a lithography and etching step is carried out in order to define through openings in the passivation bilayer 10'. The etching step may comprise a first sub-step of etching the SiO2 with a stop on the SiN, then a second sub-step of etching the SiN with a stop on the material constituting the main metal layer 350 and on the first electrodes 15a, 15b (typically copper). These openings open onto the main metal layer 350 and onto the first electrodes 15a, 15b. A gentle deoxidation of the copper may be carried out at this stage of the process. The openings in the bilayer 10' are then filled with an electrically conductive material such as Cu, preferably again using a damascene-type filling strategy. Once these openings are filled, we obtain the assembly illustrated in Figure 11.

[0192] This step makes it possible to extend the main metal layer 350 and the first electrodes 15a, 15b by ensuring that their dimensions in the transverse plane XY at the level of the passivation bilayer 10' are smaller than their dimensions in the dielectric layer 10. This makes it possible to ensure the possibility of hybrid bonding of the assembly with a support substrate 20.

[0193] As illustrated by the transition from Figure 11 to Figure 1J, a substrate of interest 5 is bonded by hybrid bonding to the passivation bilayer 10', the assembly is turned over and the support substrate 20 is removed. The removal of the support substrate 20 can for example be done by a polishing step then a finishing step by chemical etching.

[0194] Buffer layers present in the epitaxial layer 1000 may also be removed at this stage of the process. This removal may be carried out by dry etching or by more selective processes involving electropolishing or stopping on a selective buried layer.

[0195] The step of etching the buffer layers induces variations in the thickness of the continuous layer 200 (and therefore of the epitaxial layer 1000) along the third direction Z. It is thus possible to determine a total thickness variation (TTV) of the continuous layer 200. The TTV of a continuous layer 200 after removal of AIN-based buffer layers is typically of the order of 300 nm.

[0196] Figure 5B illustrates the equivalent of Figure 1J when the continuous layer 200 is formed from a lower portion 210 and an upper portion 220 having distinct materials.

[0197] Finally, the lenses 200a, 200b are formed in the continuous layer 200 (transition from FIG. 1J to FIG. 1K or from FIG. 5B to FIG. 5C). To do this, a resin can be deposited on the continuous layer 200 and then structured by a grayscale lithography step or a thermal reflux step. In both cases, the structuring of the resin into a convex shape is followed by a step of etching the continuous layer 200 through the resin mask in order to transfer the pattern of the resin into the continuous layer 200 and thus form the convex lenses 200a, 200b.

[0198] The etching is for example a dry etching. This is a selective etching between the material(s) constituting the continuous layer 200 and the stop material. The etching can for example be based on Ch, a Ch / Ar mixture, a CI2 / BCI3 mixture, or even CH4. All these chemistries make it possible to etch GaN-based lenses 200a, 200b with very high selectivity compared to a dielectric stop layer 310, whether based on an oxide or a nitride. The chemistries based on Ch, a Ch / Ar mixture and a CI2 / BCI3 mixture make it possible to etch GaN-based lenses with satisfactory selectivity compared to a metal stop layer 310 based on AISi, TiN, ITO and AISi / TiN. A CH4-based chemistry can also be used in the case of a metallic 310 barrier layer based on AISi, TiN or ITO. In the case of a metallic 310 barrier layer based on AISi, a CH4-based chemistry is even particularly preferred.

[0199] The etching parameters other than the chemistry (bias voltage, gas flow, etc.) are adapted in a conventional manner depending on the materials constituting the lenses 200a, 200b and the stop layer 310.

[0200] As mentioned previously, the thickness e3io of the stop layer 310 is chosen according to the anticipated TTV for the continuous layer 200. The greater the TTV, the greater the thickness e3io of the stop layer 310. This allows, during the etching of the formation of the lenses 200a, 200b, for the stop layer 310 to absorb this TTV. The stop layer 310 thus makes it possible to compensate for the TTV of the continuous layer 200, which results in a good definition of the convex shape of the lenses 200a, 200b and in a good uniformity of height of the convex lenses 200a, 200b, without residual continuous layer between two lenses.

[0201] In the case described above where the continuous layer 200 is formed of a lower portion 210 and an upper portion 220, the etching of the lenses 200a causes the singulation of these two portions. Thus, each lens 200a, 200b is itself formed of a lower portion 210a, 210b and an upper portion 220a, 220b. This is what is illustrated in FIG. 5C.

[0202] When the lower portion 210 and the upper portion 220 of the continuous layer 200 are both formed from the same material, then the etching of these two portions 210, 220 is done during the same etching step.

[0203] When the upper portion 220 is based on a second material distinct from the first material, the step of forming the lenses 200a, 200b further typically comprises selective etching of the second material relative to the stop material. The etching characteristic conditions described above apply mutatis mutandis to the etching of the second material of the upper portion 220 of the continuous layer 200. Whether or not the lower portion and the upper portion of the lenses are formed from the same material, each assembly formed from a diode 100a, 100b and the lens 200a, 200b which surmounts it is a monolithic assembly. In other words, it is an assembly based on one or more deposited and / or raw materials without a transfer or bonding step. This makes it possible to avoid the presence of an optical interface within these assemblies, thus limiting or even eliminating optical losses and therefore improving the performance of the devices.

[0204] A second example of a method for manufacturing a device according to the invention will be described with reference to Figures 2A to 2K. Figures 2A, 2B, 2D, 2F, 2H and 2J illustrate the steps for obtaining a device according to the third example described above. Figures 2A, 2C, 2E, 2G, 2I and 2K illustrate the steps for obtaining a device according to the fourth example described above.

[0205] Figures 2A to 2I illustrate an example of a method for obtaining the stack provided at the start of the method according to the invention.

[0206] Again, for clarity, the steps illustrated in Figures 2A to 2K illustrate the fabrication of a two-diode device only. Naturally, these steps can simultaneously produce a diode array comprising many lens-covered diodes from a single initial stack.

[0207] Figure 2A illustrates the provision of an assembly similar to that shown in Figure 1C. It is obtained by the same steps as those previously described with reference to Figures 1A to 1C.

[0208] As illustrated in Figures 2B and 2C, in this embodiment, a stop layer 310 based on a dielectric material is then deposited in the trench 50. This dielectric material can in particular be chosen from the following materials: SiO2, SiN.

[0209] The dielectric barrier layer 310 has external flanks 314 located opposite the internal flanks 323 of the passivation layer 320.

[0210] According to a first example illustrated in FIG. 2B, the dielectric barrier layer 310 does not completely fill the trench. It can in particular be deposited in a conformal manner against the internal flanks 323 of the passivation layer 320 and on the intermediate surface 1001 of the epitaxial layer 1000. The dielectric barrier layer 310 then also has internal flanks 313, opposite its external flanks 314. In this example, the external flanks 314 of the dielectric barrier layer 310 are preferably in contact with the internal flanks 323 of the passivation layer 320. In this example, the portion of the barrier layer 310 covering the intermediate surface 1001 has a thickness e300, measured along the third direction Z. In the case of a barrier layer 310 based on SiO2 or SiN, e300 can be substantially equal to 100 nm.

[0211] According to a second example illustrated in FIG. 2C, the stop layer 310 is deposited at this step so as to completely fill the trench 50.

[0212] In this example, a mirror layer 330 and a second metal layer 340' may have previously been deposited against the internal sides 323 of the passivation layer 320 and possibly partially on the intermediate surface 1001 of the epitaxial layer 1000. Thus, in this example, the stop layer 310 is typically separated from the passivation layer 320 by the mirror layer 330 and the second metal layer 340'.

[0213] In the context of the first example, openings 70 are then formed in the portion of the stop layer 310 covering the intermediate surface 1001 (FIG. 2D). They make it possible to partially expose this surface 1001. The openings 70 are separated by a portion of the stop layer 310 left in place, called stop portion 310*.

[0214] The openings 70 typically have in the transverse plane XY a width l7o of the order of 1 pm.

[0215] In the context of the second example, openings 80 are then formed in the stop layer 310 (FIG. 2E). These openings 80 pass entirely through the stop layer 310 in the third direction Z and make it possible to partially expose the intermediate surface 1001. The openings 80 are also separated by a portion of the stop layer 310 left in place, called stop portion 310*.

[0216] The 80 openings typically have in the transverse XY plane an Iso width of the order of 0.5 pm to 1 pm.

[0217] As illustrated in Figures 2D and 2E, in both examples, openings 60 are also formed in the dielectric layer 10, as in the first embodiment.

[0218] As illustrated in FIG. 2F, a mirror layer 330 and a metal layer 340 can be deposited in a conformal manner in the trench 50 and in particular in the openings 70. These layers 330, 340 thus completely cover the internal flanks 313 of the stop layer 310 and its flanks defined by the etching of the openings 70. They also cover the portions of the intermediate surface 1001 accessible by the openings 70. Furthermore, it is possible to deposit at the same time against the flanks and in the bottom of the openings 60 in the dielectric layer 10 a secondary metal layer 61 then a secondary mirror layer 63.

[0219] The trench 50, including in particular the openings 70, as well as the openings 60 are then filled with a conductive material in order to form a main metal layer 350 in the trench 50 and a first electrode 15a, 15b in each opening 60. For this, it is first possible to deposit seed layers (not shown), then a damascene copper brick (typically of the Ti / TiN / Cu type), and finally to fill with copper by electrochemical deposition (commonly referred to as ECD, from the English “Electrochemical Deposition”).

[0220] As illustrated in FIG. 2G, in the context of the second example, it is also possible to deposit in a conformal manner a mirror layer 330 (not shown) and a metal layer 340 in the openings 80 provided in the barrier layer 310. These openings 80 and the openings 60 in the dielectric layer 10 are then filled with a conductive material in order to form a metal element 350a, 350b in each of the openings 80 and a first electrode 15a, 15b in each of the openings 60.

[0221] This step can be followed by a chemical-mechanical polishing step with a stop on the dielectric layer 10.

[0222] As in the first embodiment, the main metal layer 350 or the metal elements 350a, 350b, and the first electrodes 15a, 15b are extended in a passivation bilayer 10' deposited on the dielectric layer 10, a substrate of interest 5 is bonded by hybrid bonding to the passivation bilayer 10', the assembly is turned over and the support substrate 20 is removed. The assembly illustrated in FIG. 2H (first example) or in FIG. 2I (second example) is thus obtained. The characteristics of these steps described in the context of the first embodiment can be applied mutatis mutandis in this second embodiment.

[0223] Finally, the continuous layer 200 is structured so as to form the convex lenses 200a, 200b, as illustrated by the transition from Figure 2H to Figure 2J (first example) and from Figure 2I to Figure 2K (second example). The formation of the lenses 200a, 200b is done in the same way as in the first embodiment, this time with a dielectric stop layer 310. The same remarks on the presence of a lower layer 210 and an upper layer 220 apply.

[0224] According to an advantageous variant, the method according to the invention further comprises the formation of a secondary wall 300' surmounting the wall 300 and separating the first lens 200a from the second lens 200b, as described previously.

[0225] Figures 4A to 4D illustrate steps for forming such a secondary wall 300'. These figures represent the device 1 according to an embodiment in which the barrier layer is based on a metallic material, but it is understood that the addition of a secondary wall 300' can be combined with any other embodiment of the device 1 just described. The manufacturing method described below does not present any incompatibility with another embodiment of the device.

[0226] As illustrated by the transition from Figure 4A to Figure 4B, the device 1 is first covered with a sacrificial layer 500 based on a sacrificial material, typically dielectric (SiO2, low dielectric constant material (in English “ultra-low k”), ...). A trench is then formed in the sacrificial layer 500 by a lithography process. This trench completely crosses the sacrificial layer 500 and typically opens onto the upper face 301 of the wall 300. The trench does not extend in particular into the lenses 100a, 100b. The trench is then filled with an absorbent material such as TiN or Cu. The filling is typically carried out by an electrochemical process.

[0227] As illustrated in FIG. 4D, the sacrificial layer 500 is then removed by a wet chemical process. For example, in the case where the sacrificial material is SiO2, the removal can be carried out using a so-called buffered oxide etch (BOE) solution or a hydrogen fluoride (HF) solution.

[0228] This manufacturing process makes it possible to form a 300' secondary wall with a high aspect ratio between its height and its width.

[0229] Again, it is understood that the steps just described for obtaining a secondary wall 300' can be implemented so as to simultaneously form several secondary walls, each separating two convex lenses from an array of numerous convex lenses.

[0230] All the characteristics described with reference to the device according to the invention are applicable to the method according to the invention, and vice versa.

[0231] Through the different embodiments described above, it appears clearly that the invention proposes an effective solution for reducing optical crosstalk in a micro-LED matrix while retaining good optical properties of the device 1, in particular good light extraction and good directivity.

[0232] The invention is not limited to the embodiments previously described and extends to all embodiments covered by the invention.

Claims

CLAIMS 1. Method of manufacturing an electronic device (1) comprising the following steps: • Providing a stack (2) comprising: o a substrate (10') having an upper face (11') extending mainly parallel to a plane called the transverse plane (XY), o a matrix of light-emitting or photoreceiving diodes (100a, 100b, 100c, 100d) arranged on the upper face (11') of the substrate (10'), a first diode (100a) of the matrix and a second diode (100b) of the matrix being separated, in the transverse plane (XY), by a wall (300) having an upper face (301), the wall (300) comprising a barrier layer (310) extending at least partially from at least a portion of the upper face (301) of the wall (300), the barrier layer (310) being based on a material called a barrier material, the first diode (100a) and the second diode (100b) being based on a first material, o a continuous layer (200) covering at least the first diode (100a), the upper face (301) of the wall (300) and the second diode (100b),the continuous layer (200) comprising a lower portion (210) and an upper portion (220), the lower portion (210) being in contact with the first diode (100a) and with the second diode (100b) and being based on the first material, and the upper portion (220) being located on the lower portion (210) and being based on a second material, the first material having a first refractive index and the second material having a second refractive index, the second refractive index being greater than or equal to the first refractive index, • Forming in the continuous layer (200) a first convex lens (200a) overlying the first diode (100a) and a second convex lens (200b) overlying the second diode (100b), the first convex lens (200a) and the second convex lens (200b) being spaced apart from each other, the formation of the first convex lens (200a) and the second convex lens (200b) comprising selective etching of the first material with respect to the stopping material, the etching selectivity between the first material and the stopping material being greater than or equal to 3:

1.

2. Method according to the preceding claim in which the first material and the second material are identical.

3. A method according to any one of the two preceding claims wherein providing the stack comprises the following steps: • Provide an initial stack comprising: o The continuous layer (200), o A continuous active layer (100), o The substrate (10'), • Carry out a so-called singularization etching from a lower face (12') of the substrate (10') opposite its upper face (11') into the continuous active layer (100), so as to form the first diode (100a) and the second diode (100b) in the continuous active layer (100), and to form a trench (50) between the first diode (100a) and the second diode (100b), • Form the stop layer (310) in the trench (50).

4. Method according to the preceding claim, in which the first diode (100a) and the second diode (100b) each have a flank (103a, 103b) defined by the interface of said diode (100a, 100b) with the trench (50), the method further comprising the following steps: • Forming in the trench (50) an electrically insulating element extending from the sides (103a, 103b) of the first diode (100a) and the second diode (100b), • At least partially fill the trench (50) with an electrically conductive material.

5. Method according to any one of the preceding claims in which the step of providing the stack (2) comprises: • provide a primary stack comprising: o the substrate (10'), o the diode matrix (100a, 100b, 100c, 100d), o the lower portion (210) of the continuous layer (200) based on the first material, the lower portion (210) having an upper face (211) opposite the diode matrix (100a, 100b, 100c, 100d), • forming on the upper face (211) of the lower portion (210) the upper portion of the continuous layer (200) based on the second material, the lower portion (210) and the upper portion (220) together forming the continuous layer (200).

6. Method according to the preceding claim wherein the formation of the first convex lens (200a) and the second convex lens (200b) further comprises selective etching of the second material with respect to the stopping material, the etching selectivity between the second material and the stopping material being greater than or equal to 3:

1.

7. Method according to any one of the two preceding claims in which the formation of the upper portion (220) of the continuous layer (200) is carried out by pulsed laser ablation (PLD) or by physical vapor deposition (PVD).

8. Method according to any one of the four preceding claims in which in a third direction (Z) perpendicular to the transverse plane (XY), the lower portion (210) has a thickness e210 and the upper portion (220) has a thickness 6220 with 6220^6210.

9. Optoelectronic device (1) comprising: • a substrate (10') having an upper face (11') extending mainly parallel to a plane called the transverse plane (XY), • a matrix of light-emitting or light-receiving diodes (100a, 100b, 100c, 100d, 100e) arranged on the upper face (11') of the substrate (10'), a first diode (100a) of the matrix being surmounted by a first convex lens (200a) and a second diode (100b) of the matrix being surmounted by a second convex lens (200b), the first convex lens (200a) having a lower portion (210a) in contact with the first diode (100a) and an upper portion (220a) on the lower portion (210a), the second convex lens (200b) having a lower portion (210b) in contact with the second diode (100b), and an upper portion (220b) located on the lower portion (210b), the first diode (100a), the second diode (100b) and the lower portion (210a, 210b) of each lens (200a, 200b) being based on a first material, the upper portion (220a, 220b) of each lens (200a, 200b) being based on a second material, the first material having a first refractive index and the second material having a second refractive index, the second refractive index being greater than or equal to the first refractive index, the first diode (100a) and the second diode (100b) being separated, in the transverse plane (XY), by a wall (300) having an upper face (301) in contact with each of the first lens (200a) and the second lens (200b),characterized in that the first lens (200a) and the second lens (200b) are spaced apart from each other, and in that the wall (300) comprises a stop layer (310) extending at least from a portion (301*) of the upper face (301) of the wall (300) not being covered by either the first convex lens (200a) or the second convex lens (200b), the stop layer (310) being based on a material, called a stop material, the first material and the stop material being such that, for at least one etching chemistry of the first material, the etching selectivity between the first material and the stop material is greater than or equal to 3:1., 10. Device (1) according to the preceding claim in which the first material and the second material are identical, for example are GaN.

11. Device (1) according to any one of the two preceding claims, wherein the first diode (100a) and the second diode (100b) each have a flank (103a, 103b) defined by the interface of said diode (100a, 100b) with the wall (300), the device (1) further comprising an electrically insulating element (320; 315) extending from the flank (103a, 103b) of each of the first diode (100a) and the second diode (100b).

12. Device (1) according to any one of the three preceding claims, further comprising a main metal layer (350) in electrical continuity with the first convex lens (200a) and with the second convex lens (200b).

13. Device (1) according to any one of the four preceding claims, further comprising a first metallic element (350a) in electrical continuity with the first convex lens (200a) and a second metallic element (350b) in electrical continuity with the second convex lens (200b), the first metal element (350a) and the second metal element (350b) being electrically insulated from each other.

14. Device (1) according to any one of the five preceding claims, wherein the barrier layer (310) is based on a metallic material.

15. Device (1) according to the preceding claim wherein the stopping layer (310) extends from the upper face (301) of the wall (300) under each of the first lens (200a) and the second lens (200b), preferably from the entire extent of the upper face (301) of the wall (300).

16. Device (1) according to any one of claims 9 to 13, in which the barrier layer (310) is based on a dielectric material.

17. Device (1) according to the preceding claim in combination with claim 11 in which a part of the barrier layer (310) forms part of the electrically insulating element.

18. Device (1) according to any one of claims 9 to 17 further comprising a secondary wall (300') extending above the wall (300), between the first lens (200a) and the second lens (200b), the secondary wall (300') extending at least as far as the first convex lens (200a) and the second convex lens (200b), preferably beyond the first convex lens (200a) and the second convex lens (200b), away from the upper face (301) of the wall (300) in a third direction (Z) perpendicular to the transverse plane (XY), the secondary wall (300') being at a distance from the first lens (200a) and the second lens (200b).

19. Device (1) according to any one of claims 9 to 18 wherein each convex lens (200a, 200b) defines a convex surface (201a, 201b), said convex surface (201a, 201b) corresponding to a portion of the external surface of a sphere of diameter D200, each of the first diode (100a) and the second diode (100b) having a dimension D100, D100 being measured in projection in the transverse plane (XY), perpendicular to the projection of the wall (300) in the transverse plane (XY), with D200>2*D100.

20. Device (1) according to any one of claims 9 to 19 in which the first convex lens (200a) and the second convex lens (200b) are surmounted by an anti-reflection layer (400).

Images