OPTICAL SYSTEM WITH A PHOTOELECTRIC TRANSDUCER COUPLED TO A WAVE GUIDE AND METHOD FOR ITS MANUFACTURING

Description

DOMAINE TECHNIQUE DE L'INVENTION

[0001] The technical field of the invention is that of photonic devices for intra- and inter-chip communication.

[0002] The present invention relates to an optical system using a semiconductor photoelectric transducer coupled to a waveguide. ARRIERE-PLAN TECHNOLOGIQUE DE L'INVENTION

[0003] In the field of inter- or intra-chip communication, the use of light to replace electrical signals makes it possible to overcome physical constraints related to the impedance and size of metallic interconnections, and thus, to meet the need to transfer large amounts of data at very high speed.

[0004] The integrated photonic devices or platforms developed for this purpose include a light source, usually of micrometer dimensions, electrically excited and capable of being efficiently coupled to a photonic waveguide, which conducts the light and, thus, carries the information.

[0005] In this context, light-emitting diode (LED) architectures, also called LED or micro-LED sources, have recently been used as light sources. Compared to laser sources, they have the advantage of being low-cost and consuming little power. It is also easier to reduce their size to the micrometer scale.

[0006] Such LED architectures are generally formed from a heterostructure based on a III-V semiconductor material, that is, a material composed of an element from group III and an element from group V of Mendeleev's periodic table. For example, LED sources are made from gallium nitride (or GaN). The heterostructure is further connected to two metallic electrodes which, when a voltage is applied between them, excite the III-V semiconductor material and generate light emission.

[0007] However, the use of LED sources poses the problem of efficiently coupling the emitted light in a waveguide.

[0008] A primary difficulty arises from the fact that an LED source emits waves that are incoherent with each other. Therefore, it is not possible to use a surface coupling technique, which relies on phase-matching phenomena to select the propagation mode through which optical power is transferred.

[0009] A second difficulty arises from the fact that it is often necessary to choose a waveguide material with a lower refractive index than the LED. Since light tends to propagate through the material with the highest refractive index, it can only pass very weakly from the LED to the waveguide, even if the LED and the waveguide are directly adjacent to each other or integrated within one another.

[0010] More specifically, the confinement of light in the LED is due to the refractive indices and the respective dimensions of the LED source and the waveguide.

[0011] Thus, the III-V semiconductor material of the LED has a higher refractive index than the material used for the waveguide: at 532 nm, the III-V GaN semiconductor material has a refractive index of about 2.4, while the silicon nitride SiN generally used for the waveguide has an index close to 2.

[0012] Furthermore, the LED typically has much larger dimensions than the waveguide. For example, the LED source may have a transverse extent of approximately 1 µm x 0.2 µm and the waveguide an extent of approximately 150 µm x 0.4 µm.

[0013] To overcome this difficulty and improve the efficiency of coupling the radiation emitted by the LED source into the waveguide, one solution is to form the LED source and the guide from the same layer of material, according to a so-called monolithic integration.

[0014] The paper "On-Chip photonic system using suspended pn junction InGaN / GaN multiple quantum wells device and multiple waveguides", by Wang Y. et al., Applied Physics Letters, 2016 and patent US20180267238 describe a planar monolithic integration of an LED source and a waveguide from an n-doped GaN layer.

[0015] The paper "Waveguide-coupled nanopillar metal-cavity light-emitting diodes on silicon" by Dolores-Calzadilla V. et al., Nature Communications, 2017, describes a vertical monolithic integration of an LED source and a waveguide from a p-doped InP layer. The LED source is then arranged on a segment of the waveguide.

[0016] However, using monolithic integration does not solve the problem that a large portion of the radiation emitted by the LED source is not transmitted to the waveguide. This is because the LED also emits radiation in regions that are not adjacent to the waveguide's end. Without a means of collecting and directing this radiation back to the waveguide, it is lost.

[0017] However, it is not easy to efficiently collect this "lost" radiation because it has a high angle range, or, in other words, a large divergence.

[0018] The previously cited US patent 20180267238 describes a cavity formed beneath a portion of the LED source and waveguide. This cavity creates an air / GaN interface that optimizes the use of waves emitted vertically by the LED source. However, waves emitted laterally in the direction opposite to the waveguide remain unused.

[0019] The previously cited paper, "Waveguided-coupled nanopillar metal-cavity light-emitting diodes on silicon" by Dolores-Calzadilla V. et al., Nature Communications, 2017, describes adapting the LED source structure by integrating the source into a metallic cavity. In this structure, the cavity consists of a III-V semiconductor nanopillar that defines the nanometric LED source. This nanopillar is coated with metal to channel a large portion of the spontaneous emission into the fundamental mode of an InP waveguide.

[0020] The paper "Inverse design optimization for efficient coupling of electrically injected optical antenna-LED to a single-mode waveguide," by Andrade et al., Optics Express, 2019, also describes an LED structure integrated into a metal-enclosed cavity. This cavity acts as a resonant cavity, increasing the spontaneous emission of the source. The structure is coupled vertically or laterally to a waveguide. The waveguide walls near the LED structure are then coated with the same metal used for the metal cavity. However, the presence of metal (from the LED structure or electrodes) in this area is generally undesirable because it introduces more optical losses in the LED or waveguides.

[0021] Publication WO 2021 / 053096 A1 discloses a micro-optical interconnect component comprising an optical platform with a substrate supporting an optical alignment structure for fixing or aligning an optical component. The platform also includes a very small volume light deflection element, made of a high-index material, having a curved reflective surface for deflecting a light beam entering or exiting the substrate at an angle within a predefined range.

[0022] In the field of integrated optical interconnects, there is therefore always a need for an LED source capable of being efficiently injected into a waveguide. RESUME DE L'INVENTION

[0023] The invention provides a solution to the problems mentioned above by enabling the rearward lateral radiation emitted by an electrically controlled micrometric LED source to be collected and redirected to the front of the LED source, where the waveguide is located. Thus, this "rear" lateral radiation, which normally never reaches the end of the waveguide, is channeled into the waveguide's numerical aperture. Conversely, if the electrically controlled micrometric receiver is a photodiode, the invention enables the forward radiation emitted by the waveguide to be collected and redirected to the micrometric receiver.

[0024] For this purpose, the LED source is integrated into a cylindrical-parabolic reflective structure configured to reflect light rays from the rear face of this LED source and make them converge at a focal point located in front of the LED source and in front of the end of the waveguide.

[0025] One aspect of the invention relates to an optical system comprising a substrate, a reflective structure, a photoelectric transducer and a waveguide, the reflective structure being formed by a parabolic-cylindrical pillar disposed on the substrate, the parabolic-cylindrical pillar being delimited by a cylindrical surface comprising a parabolic lateral surface and a flat lateral surface, and by two flat faces parallel to a plane of the substrate, and comprising: a first internal reflection face of parabolic profile disposed on the parabolic lateral surface of the cylindrical surface of the reflective structure, the first internal reflection face being defined by a plurality of parabolic sections parallel to the plane of the substrate, each parabolic section being defined by a parabola having an optical axis and a focus and extending from the vertex of the parabola to the line perpendicular to the optical axis and passing through the focus, a second internal face and a third internal reflection face being flat disposed on the flat lateral surface of the cylindrical surface of the reflective structure,the second and third internal reflective faces being orthogonal to the plane of the substrate and to the plane containing the optical axes of the plurality of parabolic sections and situated on either side of said plane containing the optical axes at the level of the foci of the plurality of parabolic sections, the photoelectric transducer comprising an active region configured to emit light waves or configured to receive light waves, and arranged so that the active region is positioned in the reflective structure, at the level of a part of the foci of the plurality of parabolic sections, the material of the reflective structure being chosen to be transparent to light waves, the waveguide having a proximal end, a distal end and a longitudinal axis parallel to the optical axes of the plurality of parabolic sections, and arranged on the substrate so that the proximal end is abutted to the reflective structure,between the second and third reflective surfaces and at the level of the active region.

[0026] Thanks to the invention, the reflective structure has a first internal reflection face of parabolic profile defined by a plurality of parabolic sections parallel to the plane of the substrate, each parabolic section being defined by a parabola having an optical axis and a focus and extending from the apex of the parabola to the line perpendicular to the optical axis and passing through the focus.

[0027] Thus, this first internal reflective face acts as a two-dimensional parabolic mirror and obeys in each plane parallel to the plane of the substrate the laws of geometric optics specific to the parabolic mirror.

[0028] Thus, in each plane parallel to the plane of the substrate, the first internal reflective surface provides, at its focus (located on the optical axis), a perfect image of a point at infinity. This means that rays from an isotropic light source placed at the focus are reflected (on this first internal reflective surface) parallel to the optical axis. Conversely, light rays coming from parallel to the optical axis are reflected by the first internal reflective surface at a point located at the focus.

[0029] The set of foci of the plurality of parabolic sections defines a line segment forming the locus of foci of the first internal reflection face.

[0030] Since the photoelectric transducer includes an active region configured to emit or receive light waves, the active region behaves as either an isotropic light source or a light receiver. "Light waves" refers to waves having at least one wavelength in the spectral band between 200 nm and 1200 nm.

[0031] This active region is positioned within the reflective structure at a portion of the foci of the plurality of parabolic sections. Thus, the active region has a lateral face (called the back face) oriented towards the first internal reflection face and located near its focal point. In each plane parallel to the substrate plane, the active region therefore corresponds to a point located at the focus of the parabolic section. According to the above, light rays originating from this region are reflected by the first internal reflection face, parallel to the optical axis. Conversely, light rays originating on the first internal reflection face parallel to the optical axis are reflected towards the active region.

[0032] Furthermore, the reflective structure has a second internal face and a third internal reflection face that are planar, orthogonal to the plane of the substrate and to the plane containing the optical axes of the plurality of parabolic sections.

[0033] Thus, the second and third internal reflecting surfaces act as plane mirrors positioned in front of the two-dimensional parabolic mirror in a plane perpendicular to the substrate plane. This means that, in each plane parallel to the substrate plane, light rays reflected by the first internal reflecting surface parallel to the optical axis are reflected in the same direction by one of the second or third internal reflecting surfaces. Accordingly, these light rays originating from either the second or third internal reflecting surface are then directed toward the active region by a further reflection on the first internal reflecting surface. In this way, the light rays are concentrated into an area with dimensions equivalent to those of the active region.This allows the use of a waveguide which has dimensions (in cross-section) comparable to those of the active region without introducing coupling losses (compared to systems in which the region where the rays emitted by the active region are concentrated has dimensions much larger than those of the active region, or much larger than the typical dimensions of waveguides used in integrated optics).

[0034] The fact that the dimensions of the waveguide section are comparable to those of the active region opens the way to an optical system comparable to a single-mode waveguide, and therefore having improved performance, particularly in terms of guidance.

[0035] The fact that the second and third internal reflective surfaces are located on either side of the plane containing the optical axes at the foci of the multiple parabolic sections allows the waveguide to be attached to the reflective structure at the active region and as close as possible to the foci of the parabolic reflective structure. Thus, the active region face opposite the rear face, called the front face, is attached to, or at least located near, the proximal end of the waveguide. This improves the coupling of light between the active region and the waveguide.

[0036] It should be added that using a material transparent to light rays emitted or received by the active region for the reflective structure allows the internal reflective surfaces to be fabricated from a solid pillar. The advantage is that such a pillar can be produced using standard microelectronic fabrication techniques (e.g., layer structuring). In other words, the entire optical system is compatible with microelectronic fabrication processes.

[0037] Another advantage of using a solid, transparent parabolic cylinder incorporating the active region is the continuity of refractive index within the reflective structure. This continuity reduces optical losses and thus improves the coupling of light between the active region and the waveguide.

[0038] The combination of these different characteristics thus makes it possible to obtain an integrated optical interconnect (or integrated photonic system) in which the photoelectric transducer has dimensions comparable to the cross-section of the waveguide while being efficiently coupled to it, since the back and front rays emitted by the active region are collected and concentrated by the reflective structure at the entrance of the waveguide. The term "efficiently coupled" refers to the fact that the fraction of optical power transmitted to the waveguide (when the photoelectric transducer is a light source) or to the photoelectric transducer (when the photoelectric transducer is a light receiver) is increased.

[0039] Advantageously, the photoelectric transducer comprises an upper electrode and a lower electrode, the upper electrode being positioned on an upper face of the reflective structure directly above the predetermined area for the active region, and having lateral dimensions corresponding to the predetermined dimensions for the active region.

[0040] Positioned in this way, the upper electrode is located at a distance from the active region and the proximal end of the waveguide. The metal from which it is formed causes little or no disturbance to the light rays, which reduces optical losses and promotes better coupling efficiency.

[0041] Thus, thanks to the two flat faces parallel to the substrate of the reflective structure, the photoelectric transducer electrodes are positioned outside the reflective structure. This reflective structure can therefore be formed from a single transparent material (the active region is defined solely by the photoelectric transducer electrodes). This material continuity, which implies a continuity of refractive index within the reflective structure, minimizes optical losses within this structure, and therefore improves light coupling in the waveguide.

[0042] It is worth noting that the position of the upper electrode preserves the transparency of the reflective structure. Furthermore, since the size and position of the active region are defined by the size and position of the upper electrode, no etching of the transducer's sides is required, thus enabling a simple manufacturing process.

[0043] Advantageously, a first medium is disposed outside the reflective structure in a space adjacent to the first internal reflection face and a second medium is disposed outside the reflective structure and the waveguide in the space adjacent to the second internal reflection face and in the space adjacent to the third internal reflection face, the first and second medium having a refractive index strictly lower than the refractive index of the transparent material chosen for the reflective structure.

[0044] Thus, the internal reflective surfaces are obtained through an interface between two media. The advantage is that these internal reflective surfaces can be produced from a solid pillar, simply and in a manner compatible with standard microelectronic fabrication processes.

[0045] According to a first embodiment, the first medium is made of a dielectric material and the second medium comprises air.

[0046] The dielectric material allows the first internal reflective face to be defined while also electrically isolating the sides of the reflective structure.

[0047] Air allows the second and third internal reflective surfaces to be defined without requiring a metallic material that could disrupt the propagation of light rays near the active region and / or the waveguide. Indeed, unlike the first internal reflective surface, the second and third internal reflective surfaces are located near the waveguide and the active region, and it is preferable to avoid adding any metal to this area.

[0048] According to a second variant, the first and second media are made of a dielectric material.

[0049] The dielectric material allows the internal reflective faces to be defined while also electrically isolating the sides of the reflective structure.

[0050] Advantageously and in a manner compatible with the two previous embodiment variants, a wall of metallic material is placed in the first middle and near the first internal reflective face.

[0051] Thus, the metallic wall allows the first internal reflective face to operate in specular reflection or total internal reflection. The fraction of reflected radiation is increased compared to a transparent material / dielectric medium interface. Because this face is far from the active region and the waveguide, the metallic wall does not affect the confinement of light rays at the active region or the proximal end of the waveguide.

[0052] In addition to the characteristics mentioned in the preceding paragraph, the optical system according to one aspect of the invention may have one or more complementary characteristics from among the following, considered individually or according to all technically possible combinations.

[0053] The photoelectric transducer is formed of a PN junction and quantum wells arranged in the PN junction.

[0054] The PN junction material is a III-V material such as a GaN-based material.

[0055] The transparent material chosen for the reflective structure is the PN junction material.

[0056] Thus, the reflective structure and the photoelectric transducer form a single, solid piece with a homogeneous refractive index. The advantage is that the figure of merit of the reflective structure is improved, leading to better coupling efficiency.

[0057] The waveguide is formed from the PN junction material and quantum wells.

[0058] Thus, the waveguide is attached to the reflective structure without the need to etch the quantum wells of the photoelectric transducer. The advantage is that electrical problems are avoided.

[0059] A layer of metallic material is disposed in the second medium, near the second internal reflective face and near the third internal reflective face.

[0060] Thus, the second and third internal reflective faces behave as total reflective surfaces.

[0061] The reflective structure has a lateral height dimension in the plane of the substrate strictly greater than a size of the waveguide.

[0062] Preferably, the height is 18 times greater than the size of the guide.

[0063] Thus, even highly divergent light rays (or those making a large angle with respect to the optical axis of the first internal reflective surface) are efficiently collected and redirected towards the active region or waveguide. The reflective structure also remains compact.

[0064] The waveguide includes a proximal end that is enlarged relative to the dimensions of the waveguide.

[0065] This waveguide configuration allows for more efficient coupling of the light emitted by the active region and redirected towards the proximal end of the waveguide by the reflective structure.

[0066] Another aspect of the invention relates to a method for implementing an optical system according to the first aspect of the invention, comprising the steps of: Provision of a substrate comprising a substrate layer, a buffer layer arranged on the substrate layer and a PN junction arranged on the buffer layer; Three-dimensional structuring of the PN junction so as to define a top face of the waveguide as well as a portion of the parabolic-cylindrical pillar forming the reflective structure, the structuring comprising lithography substeps; Vertical etching of the PN junction on either side of the parabolic-cylindrical pillar formed to complete the definition of the waveguide and the parabolic-cylindrical pillar forming the reflective structure, and to define a bonding layer; Passivation by conformal deposition of the first dielectric medium outside the parabolic-cylindrical pillar, at the first internal reflection face, to obtain the reflective structure; Formation of a top electrode on the top face of the PN junction at the foci of the plurality of parabolic sections.and a lower electrode on the bonding layer to complete the photoelectric transducer and optical system.

[0067] The invention and its various applications will be better understood by reading the following description and examining the accompanying figures. BREVE DESCRIPTION DES FIGURES

[0068] The figures are presented for illustrative purposes only and are in no way limiting to the invention. There figure 1 schematically represents, in top view, a first embodiment of the optical system, The figure 2 shows an enlarged top view of the reflective structure and photoelectric transducer of the optical system of the figure 1 , There figure 3 is a cross-sectional view of the optical system of the figure 1 , There figure 4 schematically represents, in top view, a second embodiment of the optical system, The figures 5A à 5E represent, in cross-section, the stages or sub-stages of manufacturing the optical system of the figure 1 .

[0069] Unless otherwise specified, the same element appearing on different figures has a unique reference. DESCRIPTION DETAILLEE

[0070] The first aspect of the invention relates to the field of integrated photonics. In particular, the first aspect of the invention concerns a guided optics system based on a photoelectric transducer formed by a PN junction and laterally integrated with a waveguide. The optical system according to the invention is notable, in particular, for its ability to improve the lateral coupling of this photoelectric transducer with the waveguide.

[0071] A first embodiment of the optical system will be described with reference to the figure 1 , to the figure 2 and to the figure 3 .

[0072] A second embodiment will then be described with reference to the figure 4 .

[0073] Common to the first and second embodiments, the optical system 1 includes a photoelectric transducer 30, a waveguide 40 and a reflective structure 20 carried by a substrate 10.

[0074] These elements 10, 20, 30 and 40 are described individually below.

[0075] With reference to the figure 3 , the substrate 10 comprises a substrate layer 11, for example a sapphire substrate layer, a buffer layer 12, for example aluminum nitride, arranged on the substrate layer 11. This buffer layer 12 carries the photoelectric transducer 30, the waveguide 40 and the reflective structure 20.

[0076] Still referring to the figure 3 , the photoelectric transducer 30 comprises a lower region 31, an active region 32 and an upper region 33 formed in a PN 13 junction, as well as two electrodes 34,35, one, called upper electrode 34, being connected to the P 13c layer of the PN 13 junction and the other, called lower electrode 35, being connected to the N 13a layer of this PN 13 junction via a bonding layer 16.

[0077] The photoelectric transducer 30 is a receiving optical system when the active region 32 is capable of transforming an optical signal (illustrated by rays r 6 and r 8 on the figure 1 ) received from the waveguide 40 as an electrical signal, or an optical transmitting system when the active region 32 is capable of generating an optical signal (illustrated in particular by rays r 5 and r 7 on the figure 1 ) towards the waveguide 40 under the effect of an electrical signal. The electrical signals are conducted / applied by the electrodes 34,35.

[0078] In the following description, we will consider a photoelectric transducer 30 of the transmitter type, thus having an active emitting region 32. The general principle of the invention, however, naturally applies to a photoelectric receiver transducer comprising an active receiving region 32.

[0079] It should be noted that the terms "optical signal" or "light waves" refer to electromagnetic waves comprising at least one wavelength in the spectral band from 200 nm to 1200 nm.

[0080] The material of the photoelectric transducer corresponds to the material 13 forming the PN junction.

[0081] This material 13 is a semiconductor material such as a III-V semiconductor material or a IV-IV semiconductor, or even a II-VI semiconductor, exhibiting a high refractive index in the spectral band 200 nm - 1200 nm, for example greater than 2.

[0082] Preferably, the material 13 of the photoelectric transducer is a III-V semiconductor material.

[0083] As used here, the term "III-V semiconductor material" means and includes any semiconductor material that is at least primarily composed of one or more elements from group IIIA of the periodic table (B, Al, Ga, In, and Ti) and one or more elements from group VA of the periodic table (N, P, As, Sb, and Bi). For example, III-V semiconductor materials include, but are not limited to, GaN, GaP, GaAs, InN, InP, etc.

[0084] For example, the photoelectric transducer material is based on gallium nitride (GaN). The refractive index of such a material is 2.4 at a wavelength of 532 nm.

[0085] The lower 31, active 32 and upper 33 regions are respectively formed in the N 13a layer, the quantum wells 13b and the P 13c layer of the PN 13 junction. They are connected successively from bottom to top.

[0086] For example, the N 13a layer is an nGaN layer, the P 13c layer is a pGaN layer and the quantum wells 13b are formed from an alternation of InGaN and GaN layers.

[0087] The active region 32 is defined by the upper electrode 34 and the PN junction configuration.

[0088] Thus, the position and lateral dimensions of the upper electrode 34 determine the position and dimensions of the active region 32 and of the upper and lower regions 31,32 on the substrate 10. In other words, the upper electrode 34 defines the lateral dimensions and the position of the photoelectric transducer 30.

[0089] These lateral dimensions Le1, Le2 of the upper electrode 34 are represented in figure 2 Preferably, they are on the order of a few hundred nanometers, for example 200 nm or 400 nm. For simplicity, the first lateral dimension Le1 can be taken to be equal to the second lateral dimension Le2. As will be described later, the lateral dimensions Le1 and Le2 of the upper electrode 34 are related to the dimensions of the reflective structure 20 and are therefore determined based on the dimensions of this reflective structure 20.

[0090] The configuration of the PN junction determines the optical and photoelectric properties of the active region 32.

[0091] The configuration of the PN junction is thus chosen so that the active region behaves, under the effect of an electrical excitation supplied by the electrodes 34,35, as a point source of isotropic light, in each plane parallel to the plane of the substrate 10. The wavelength of the radiation is for example located in the green, around 532 nm.

[0092] The radiation emitted by the active region 32 in a plane i parallel to the plane of the substrate 10 is illustrated in figure 1 using a geometric optical description. The radiation is emitted with a large angular range Θ, as illustrated by the rays r10, r20, r5, r7 shown in figure 1 . Part of the rays are emitted (for example the r 10 , r 20 rays) by a lateral face called the rear face 321 (cf. figure 3 ) of the active region 32. Another part is emitted by a lateral face opposite to the rear face 321 (cf. figure 3 ), said front face 322 of the active region 32.

[0093] With reference to the figure 1 , the reflective structure 20 surrounds the photoelectric transducer 30 at the stack formed by the lower region 31, active region 32 and upper region 33.

[0094] Specifically, the reflective structure 20 forms with the lower 31, active 32 and upper 33 regions of the photoelectric transducer 30 a solid pillar 20 of cylindrical-parabolic shape. In other words, the pillar 20 is a truncated right cylinder, with axis and generatrices perpendicular to the plane of the substrate 10, and having as its directrix a parabola with axis Xoi parallel to the plane of the substrate 10 and having as its focus a point Fi located on the axis Xoi. This right cylinder is truncated by a plane parallel to the generatrices and including the focus Fi. The pillar 20 is thus delimited by a cylindrical surface comprising a parabolic lateral surface 25 and a flat lateral surface 26 and by two flat faces 27, 28 (cf. figure 3 ) parallel to the plane of the substrate 10.

[0095] The two flat faces 27,28 allow the electrodes 34,35 of the photoelectric transducer to be placed outside the reflective structure 20, directly above the active region of the photoelectric transducer.

[0096] The material 21 chosen for the reflective structure 20 is transparent to light waves emitted by the active region 32. It has a refractive index close to or equal to the refractive index of the photoelectric transducer material in the wavelength range of 400 nm - 800 nm. By "close to" we mean that the difference in index is less than 0.2.

[0097] Preferably, the transparent material 21 chosen for the reflective structure 20 is identical to the material 13 of the photoelectric transducer 30. When the latter is GaN-based, the refractive index of the reflective structure is 2.4. In other words, the reflective structure 20 and the active region 32 of the photoelectric transducer then have the same refractive index. This reduces optical losses in the reflective structure and thus improves the coupling of light between the active region and the waveguide.

[0098] The reflective structure 20 includes, near the parabolic lateral surface 25, a first internal reflection face 22 and, near the flat lateral surface 26, a second internal reflection face 23 and a third internal reflection face 24.

[0099] More specifically, as shown by the figure 1 , the first internal reflective face 22 rests on, or is disposed on, the parabolic lateral surface 25 of the cylindrical surface of the reflective structure.

[0100] As further shown by the figure 1 The second internal reflection face 23 and the third internal reflection face 24 rest upon, or are arranged upon, the flat lateral surface 26 of the cylindrical surface of the reflective structure. These second and third internal reflection faces are also arranged on either side of the foci Fi.

[0101] The first internal parabolic reflecting face 22 exhibits geometric characteristics analogous to those of the parabolic surface 25 of the cylindrical surface of the reflective structure. With reference to the figure 1 The first internal reflection face 22 has a parabolic profile defined by a plurality of parabolic sections parallel to the plane of the substrate 10. Such a section is illustrated in figure 1 . Each parabolic section is defined by the parabola presenting the optical axis X oi and the focus F i , and extends from the vertex S i of the parabola to the line y perpendicular to the optical axis X oi and passing through the focus F i .

[0102] The second internal face 23 and the third internal reflective face 24 have geometric characteristics analogous to the flat lateral surface 26 of the cylindrical surface of the reflective structure. The second internal face 23 and the third internal reflective face 24 are planar faces, orthogonal to the plane of the substrate 10 and to the plane containing the optical axes Xoi of the plurality of parabolic sections. These faces 23, 24 are further located on either side of said plane containing the optical axes Xoi at the foci Fi of the plurality of parabolic sections.

[0103] The internal reflective faces 22, 23, 24 are capable of reflecting light waves emitted by the active region inside the reflective structure 20 into the interior of the reflective structure 20. In other words, the internal faces 22, 23 and 24 act as semi-reflective surfaces.

[0104] To this end, a first medium 51 is arranged outside the reflective structure 20 in a space adjacent to the first internal reflection face 22. In other words, the second medium 51 encloses the parabolic lateral surface 25 of the reflective structure 20 along its entire height. Furthermore, a second medium 52 is arranged outside the reflective structure on the side of the flat lateral surface 26, in the space adjacent to the second internal reflection face 23 and in the space adjacent to the third internal reflection face 24.

[0105] The first and second media 51,52 have a refractive index strictly lower than the refractive index of the transparent material 21 chosen for the reflective structure (or of the material 13 of the photoelectric transducer 30).

[0106] Preferably, the first medium 51 is made of a dielectric material and the second medium 52 comprises air.

[0107] The dielectric material is, for example, silicon oxide.

[0108] Thus the refractive index of the first and second media 51,52 is close to 1, while the refractive index of the reflective structure 20 and the photoelectric transducer 30 is 2.4. The first dielectric medium 51 then forms a dielectric wall 51 surrounding the parabolic lateral surface of the reflective structure 20.

[0109] Another possibility is that the first and second media 51,52 are both made of a dielectric material (not shown in the figures 1, 2 Or 3 ).

[0110] To complete the description of the reflective structure 20, it should be specified that it has lateral dimensions (cf. figure 2 ) of length Lsr1 along the optical axis X oi , and of height Lsr2 along the line y, as well as a height Lsr3 (along an axis perpendicular to the plane of the substrate 10, cf. figure 3 ).

[0111] For example, the lateral length dimension Lsr1 is 1 micron and the lateral height dimension Lsr2 is 4 microns. The lateral height dimension Lsr2 is related to the lateral length dimension Lsr1 by the equation of the parabola.

[0112] With reference to the figure 1 The waveguide 40 has a proximal end 41 and a distal end 42 (not shown in the figure 1 ), and a longitudinal axis X g.

[0113] The waveguide 40 is made of a semiconductor material 43 having a refractive index, called the refractive index of the guide, of the same order of magnitude as the refractive index of the material 13 of the photoelectric transducer 30. The waveguide 40 is further surrounded by one or more media 52, 12 having a refractive index strictly lower than the refractive index of the waveguide 40 (and of the material 13 of the photoelectric transducer).

[0114] The waveguide 40, for example, is made of silicon nitride SiN.

[0115] Preferably, the waveguide 40 is made of the same material as the material 13 of the photoelectric transducer 30 and includes the quantum wells 13b. This allows a continuity of index between the active region, the reflective structure and the waveguide which promotes better coupling of light.

[0116] The arrangement of the photoelectric transducer 30, the reflective structure 20 and the waveguide 40 in the optical system 1 relative to each other is described below.

[0117] First, the photoelectric transducer 30 is arranged so that the active region 32 is positioned in the reflective structure 20, at a portion of the foci F i of the plurality of parabolic sections. The lateral dimensions Le1 and Le2 of the upper electrode 34 (and therefore of the active region 32) are at least ten times smaller than the lateral length and height dimensions Lsr1 and Lsr2 of the reflective structure 20.

[0118] In figure 1 , the active region 32 occupies an area corresponding to the upper electrode 34. This is positioned laterally to or near the focus F i of the first internal reflection face 22.

[0119] Thus, the rear lateral face 321 of the active region is oriented towards the parabola formed by the first internal reflection face 22.

[0120] Secondly, the waveguide 40 is arranged so that its longitudinal axis Xg is parallel to the optical axes Xoi of the plurality of parabolic sections and that its proximal end 41 is attached to the reflective structure 21 between the second and third internal face 23,24 of reflection (laterally cf. figure 1 ) and at the level of the active region 32 (vertically, cf. figure 3 ). The term "attached" refers to the fact that the proximal end 41 of the waveguide 40 is in contact with the reflective structure and the active region 32.

[0121] When the waveguide 40 is formed of the same material 13 as the photoelectric transducer 30 and includes the quantum wells 13b, these are aligned with the longitudinal axis Xg of the guide (cf. figure 3 ). In this case, the waveguide 30 is joined to the reflective structure 20 and to the active region 32 by monolithic integration.

[0122] Thirdly, the lateral dimensions Lsr1 and Lsr2 (cf. figure 2 The lateral height dimension (Lsr2) of the reflective structure 20 is strictly greater than the waveguide size Lg2. For example, the lateral height dimension Lsr2 of the reflective structure 20 is 4 microns for a waveguide size Lg2 of 400 nm. In another example, the lateral height dimension Lsr2 of the reflective structure 20 is 4 microns for a waveguide size Lg2 of 220 nm. The lateral height dimension Lsr2 is then 18 times greater than the waveguide size Lg2 of the waveguide 40.

[0123] Thanks to the elements just described and their particular arrangement, the optical system 1 allows, in each plane parallel to the substrate and over the entire height Lsr3 of the reflective structure 20, the channeling of the light rays coming from the rear face 321 of the active region 32 and directing them towards the proximal end 41 of the waveguide 40. It should be noted that in the absence of the reflective structure 20, these rays do not reach the proximal end 41 of the waveguide 40. Thus, the brightness (or optical power) at the output of the assembly formed by the photoelectric transducer and the reflective structure is increased.

[0124] The path followed by the light rays from the rear face 321 of the active region 32 is illustrated using rays r10 and r20 shown in figure 1 .

[0125] Since the active region is located at the foci of the first internal reflection face 22, rays r10 and r20 are reflected onto the first internal reflection face 22 parallel to the optical axis Xoi. This first reflection gives rise to rays r11 and r21 shown in figure 1 .

[0126] Since the second and third reflection faces 23,24 are planar and orthogonal to the plane of the substrate and to the plane containing the optical axes of the plurality of parabolic sections, the rays reflected once r 12 and r 21 are reflected a second time and redirected in the same direction, which is therefore parallel to the optical axis X oi . This second reflection gives rise to rays r 13 and r 23 .

[0127] Finally, these rays from two successive reflections are reflected a third time on the first internal reflection face 22. Since they arrive at the first internal reflection face 22 parallel to the optical axis Xoi, the reflected rays r14 and r24 are directed towards the focal point Fi, in front of which lies the proximal end 41 of the waveguide 40. It should be noted that the focal point area Fi, in which the reflected rays r14 and r24 are concentrated, has dimensions comparable to those of the active region of the photoelectric transducer. This allows the use of a waveguide with a cross-section of the same dimensions as the active region, without compromising coupling efficiency. The ability to use a waveguide with a cross-section of the same dimensions as the active region enables those skilled in the art to design a single-mode waveguide while maintaining high coupling efficiency.

[0128] The fact that the lateral height dimension Lsr2 of the reflective structure 20 is much greater, for example 18 times greater, than the size Lg2 of the waveguide 40 allows to channel rays emitted with a large angle Θ with respect to the optical axis X oi.

[0129] The second embodiment illustrated in figure 4 differs from the first embodiment in two points.

[0130] First, the first internal reflective face 22 is obtained using a wall of metallic material 60 arranged in the first medium 51 and close to the first internal reflective face 22. The first internal reflective face 22 thus operates in total reflection.

[0131] Next, the waveguide 40 has an enlarged proximal end 41 and a funnel shape 44 at this end 41. In this configuration, the lateral dimensions Le1, Le2 of the active region 32 are adapted to the size of the enlarged end 41, via the size of the upper electrode 34. This configuration is favorable to better coupling of the incident optical power at the end 41 of the waveguide.

[0132] It should be noted that a layer of metallic material (not shown in figure 3 ) can also be placed in the second medium 52, near the second internal reflection face 23 and near the third internal reflection face 24 so that the second and third faces 24,25 also work in total reflection.

[0133] A second aspect of the invention relates to a method 500 for manufacturing the optical system 1.

[0134] THE figures 5A à 5E illustrate, in cross-section, the stages or sub-stages of manufacturing the optical system of the figure 1 , according to a preferred embodiment.

[0135] Process 500 begins at step S501 illustrated in figure 5A with the provision of a substrate 10 comprising the substrate layer 11, the buffer layer 12 and the PN junction 13, the latter comprising a lower N layer 13a, quantum wells 13b and an upper P layer connected successively from bottom to top. The PN junction 13 is obtained for example by epitaxial growth and doping.

[0136] This step S501 continues with step S502, represented in figure 5B . Step S502 consists of structuring in three dimensions by lithography the PN 13 junction to create recesses and thus define a top face 45 of the waveguide 40 and delimit a part of the lateral surfaces of the reflective structure 20.

[0137] Following this initial three-dimensional structuring S502, a cylindrical-parabolic pillar 20a, conforming to the pillar forming the reflective structure 20, is created in the PN junction. This pillar 20a thus comprises the parabolic lateral face 25 and the planar lateral face 26 described previously.

[0138] Step S503 follows step S502. It is illustrated in figure 5C . Step S503 is a vertical etching step, also called mesa etching, carried out on either side of the pillar 20a formed in step S502 to define the waveguide 40 at the level of the flat lateral face 26 of the pillar 20a, to complete the definition of the cylindro-parabolic pillar 20b forming the reflective structure 20, and to define the bonding layer 16.

[0139] The vertical etching step S503 is followed by a passivation step S504, illustrated in figure 5D , aimed at conformally depositing a dielectric layer 51 on the parabolic lateral surface 25 of the pillar 20b. The flat lateral surface 26 as well as the lateral and upper faces of the waveguide are in contact with the air 52.

[0140] The dielectric layer 51 electrically isolates the parabolic lateral surface 26 from the reflective structure 20 while simultaneously creating an interface between the material 13 of the PN junction and the dielectric, these having very contrasting refractive indices (2.4 for example for a PN junction in GaN, and 1 for the dielectric).

[0141] At the end of step S504, the substrate 10 carries the waveguide 40 and the reflective structure 20. The reflective structure also includes the first internal reflection face 22 and the second and third internal reflection faces 23, 24. Finally, the waveguide is attached laterally to the reflective structure 20 at its flat lateral surface 26, between the second and third internal faces 24, 25 (not shown in the diagram). figure 5D ) and at the quantum wells 13b, these being distributed along its longitudinal axis Xg. A bonding layer 16 linked to the lower region 13a has also been defined on the side of the parabolic lateral surface 25 of the reflective structure 20.

[0142] Step S505, which follows step S504 of passivation, is illustrated in figure 5E This step S505 consists of the formation of the upper electrodes 34 and lower electrodes 35.

[0143] For this purpose, a first metallic pattern 34 is formed on the upper face of the cylindrical-parabolic pillar 20b, near the flat lateral surface 26, which corresponds to the level of the foci of the first internal face 22 of reflection.

[0144] A second metallic wall-shaped pattern 35 is also formed on the bonding layer 16. The metallic wall 35 may or may not take the shape of the parabolic lateral surface 25.

[0145] A combination of photolithography, resin deposition and removal can be used at this S505 step.

[0146] At the end of this S505 step, the photoelectric transducer is obtained and the optical system 1 is finalized.

[0147] The 500 process just described concerns a monolithic integration of the waveguide, the reflective structure and the photoelectric transducer.

[0148] Naturally, this process can also be adapted when the waveguide 40 and the reflective structure are made of materials different from the photoelectric transducer material. In this case, the S502 structuring step comprises a combination of structuring and epitaxial substeps to define 1) the photoelectric transducer, 2) the parabolic-cylindrical pillar 20b, and 3) the upper face 45 of the waveguide 40.

Claims

1. An optical system (1) comprising a substrate (1), a reflective structure (20), a photoelectric transducer (30) and a waveguide (40), - the reflective structure (20) being formed by a cylindro-parabolic pillar (20b) disposed on the substrate (10), the cylindro-parabolic pillar being delimited by a cylindrical surface comprising a parabolic side surface (25) and a planar side surface (26), and by two planar faces (27, 28) parallel to a plane of the substrate (10), and comprising: ∘ a first inner reflection face (22) with a parabolic profile disposed on the parabolic side surface (25) of the cylindrical surface of the reflective structure, the first inner reflection face (22) being defined by a plurality of parabolic sections parallel to the plane of the substrate (10), each parabolic section being defined by a parabola having an optical axis (Xoi) and a focus (Fi) and extending from the vertex (Si) of the parabola to the straight line perpendicular (y) to the optical axis (Xoi) and passing through the focus (Fi), ∘ a second inner planar reflection face (23) and a third inner planar reflection face (24) disposed on the planar side surface (26) of the cylindrical surface of the reflective structure, the second and third inner reflection faces (23, 24) being orthogonal to the plane of the substrate (10) and to the plane including the optical axes of the plurality of parabolic sections and located on either side of said plane including the optical axes at the foci (Fi) of the plurality of parabolic sections, - the photoelectric transducer (30) comprising an active region (32) configured to emit light waves or configured to receive light waves, and arranged so that the active region (32) is positioned in the reflective structure (20), at a part of the foci (Fi) of the plurality of parabolic sections, the material (21) of the reflective structure (20) being chosen to be transparent to light waves, - the waveguide (40) having a proximal end (41), a distal end (42) and a longitudinal axis (Xg) parallel to the optical axes (Xoi) of the plurality of parabolic sections, and arranged on the substrate (10) so that the proximal end (41) is adjoining the reflective structure (20), between the second and third reflection faces (23, 24) and at the active region (32).

2. The optical system (1) according to claim 1, characterised in that a first medium (51) is disposed outwardly of the reflective structure (20) in a space adjacent to the first inner reflection face (22) and a second medium (52) is disposed outwardly of the reflective structure (20) and the waveguide (40) in the space adjacent to the second inner reflection face (23) and in the space adjacent to the third inner reflection face (24), the first and second media (51, 52) having a refractive index strictly lower than the refractive index of the transparent material (21) chosen for the reflective structure (20).

3. The optical system (1) according to claim 2, characterised in that the first medium (51) is formed by a dielectric material and the second medium (52) comprises air.

4. The optical system (1) according to claim 2, characterised in that the first and second media (51,52) are formed by a dielectric material.

5. The optical system according to one of claims 3 to 4, characterised in that a wall (60) of metal material is disposed in the first medium (51) and in proximity to the first inner reflection face (22).

6. The optical system (1) according to one of claims 1 to 5, characterised in that the photoelectric transducer (30) is formed by a P-N junction (13) and quantum wells (13b) disposed in the P-N junction (13) and the transparent material (21) chosen for the reflective structure (20) is the material (13) of the P-N junction.

7. The optical system (1) according to claim 6, characterised in that the waveguide (40) is formed by the material (13) of the P-N junction and the quantum wells (13b).

8. The optical system according to one of claims 6 to 7, characterised in that the material (13) of the P-N junction is a III-V material such as a GaN-based material.

9. The optical system (1) according to one of claims 1 to 8, characterised in that the photoelectric transducer (30) comprises an upper electrode (34) and a lower electrode (35), the upper electrode (34) being positioned on the upper face (27) of the reflective structure (20) in vertical alignment with the predetermined zone (32) for the active region, and having lateral dimensions (Le1, Le2) corresponding to the predetermined dimensions for the active region (32).

10. The optical system (1) according to one of claims 1 to 9, characterised in that the reflective structure (20) has a lateral height dimension (Lsr2) in the plane of the substrate (10) strictly greater than a size (Lg2) of the waveguide.

11. The optical system (1) according to one of claims 1 to 10, characterised in that the waveguide (40) comprises a proximal end (41) enlarged relative to the dimensions (Lg2) of the waveguide.

12. A method (500) for making an optical system (1) according to one of claims 1 to 11, comprising the steps of: - Providing (S501) a substrate (10) comprising a substrate layer (11), a buffer layer (12) arranged on the substrate layer (11) and a P-N junction (13) arranged on the buffer layer (12), - three-dimensionally structuring (S502) the P-N junction (13) so as to define an upper face (45) of the waveguide (40) as well as a part of the cylindro-parabolic pillar (20a) forming the reflective structure (20), structuring (S502) comprising lithography sub-steps, - Vertically etching (S503) the P-N junction (13) on either side of the cylindro-parabolic pillar (20a) formed to complete definition of the waveguide (40) and of the cylindro-parabolic pillar (20b) forming the reflective structure (20), and to define a bonding layer (16), - Passivating (S504) the first dielectric medium by conformal deposition (51) outwardly of the cylindro-parabolic pillar (20b), at the first inner reflection face (22), to obtain the reflective structure (20), - forming (S505) an upper electrode (34) on the upper face (27) of the P-N junction at the foci (Fi) of the plurality of parabolic sections, and a lower electrode (35) on the bonding layer (16) to complete the photoelectric transducer (30) and the optical system (1).