ARRANGEMENT OF AN ACTIVE SEMICONDUCTOR COMPONENT AND A PASSIVE OPTICAL COMPONENT ON A SILICONE BASE
Patent Information
- Application Number
- DE602020062089
- Authority / Receiving Office
- DE · DE
- Patent Type
- Patents
- Current Assignee / Owner
- Priority Date
- 2019-07-05
- Filing Date
- 2020-07-03
- Publication Date
- 2025-11-12
- Estimated Expiration
- 2040-07-03
AI Technical Summary
Achieving optical coupling between silicon photonics (SiPho) waveguides and active semiconductor components with minimal coupling losses and low cost is challenging due to differences in optical mode geometry and alignment tolerance issues, with existing methods being complex, costly, and unreliable.
A method involving evanescent coupling between silicon-based passive optical components and active semiconductor components, utilizing double evanescent coupling through intermediate waveguides with specific geometry and alignment, allowing efficient transfer of optical modes despite differing sizes and tolerating misalignments, and using standard manufacturing processes.
Enables efficient coupling with reduced losses, improved alignment tolerance, and cost-effective manufacturing by ensuring evanescent coupling even with different optical modes, facilitating integration of active functions on SiPho circuits.
Description
Domaine technique de l'invention
[0001] The present invention relates to an assembly of an active semiconductor component and a passive silicon-based optical component, a method for manufacturing such an assembly and a method for coupling between waveguides of components of such an assembly. Etat de la technique
[0002] Silicon photonics (SiPho) encompasses the study and applications of optical components (SiPho components) that use silicon as the optical propagation medium. SiPho components can be fabricated using semiconductor manufacturing technologies, and since silicon is used as a substrate in most integrated circuits, it is possible to create hybrid devices in which optical and electronic components are integrated on the same substrate, making SiPho technology particularly attractive. Notably, SiPho components are used in a variety of applications, such as passive waveguides, photodiodes, and modulators.SiPho components thus have silicon as their main component and may also include, but are not limited to, silicon oxides (SiO2), silicon nitrides (SiN, Si2N3, Si2N, Si3N4), silicon oxynitrides (SiON).
[0003] Due to the nature of the materials used (indirect bandgap semiconductors), SiPho technology does not, however, allow for the generation of active functions, such as optical amplification. Therefore, the goal is to integrate active semiconductor components onto SiPho circuits; these components are capable of modifying an optical property of a light beam when subjected to a current or voltage source. Active semiconductor components include, for example, materials capable of generating optical gain, such as semiconductor crystals epitaxially grown on substrates, like gallium arsenide (GaAs), gallium nitride (GaN), or indium phosphide (InP).
[0004] One difficulty, however, is achieving the optical coupling ratio between waveguides fabricated using these two technologies (SiPho and active components) with minimal coupling losses and at low cost. Indeed, depending on the materials constituting the components, the optical modes propagating in the waveguides have different geometries, which complicates coupling. For example, the geometry of a mode in a Si waveguide is typically 0.7 x 0.7 µm²; it is typically 1 x 1.3 µm² in a waveguide on an InP substrate and more than 4 x 4 µm² in a SiO₂ waveguide.
[0005] One initial approach is direct coupling between waveguides (or " butt coupling (according to the Anglo-Saxon expression). To try to increase the overlap between modes and reduce coupling losses, it is known to implement mode-size converters on both waveguides (see for example Janiak et al. [Ref 1] and Hirohito [Ref 2]). To limit the six degrees of alignment freedom (three translations and three rotations), most solutions involve mounting the two components flipped relative to each other on a common substrate (a "flip-chip" mounting). Vertical alignment is then ensured by support points at a predetermined height, and horizontal alignment is ensured by specific high-precision alignment equipment (approximately 1 µm). In the optimal case, coupling losses of -2 dB are achieved.
[0006] However, this type of coupling exhibits a very low alignment tolerance, due to mode mismatch, with typically 3dB of additional coupling losses for every 1µm of misalignment.
[0007] To overcome these limitations, alternative solutions have been proposed, at a much higher cost, using optical devices placed in the free space between the waveguides to be coupled to mitigate mode mismatch. See, for example... Gao et al. [Ref 3]. The FIG. 1A This illustrates an example of direct coupling between a SiPho component 10 comprising a waveguide 12 (e.g., Si, SiN, or SiON in SiO2) and an active semiconductor component 11 comprising a waveguide 13 that is to be optically coupled to the waveguide 12. The two components are assembled on a basic element 16, for example, made of silicon. In this example, an optical device 14, for example, a lens, facilitates coupling in the case of a significant difference in the size of the optical modes between the two waveguides. Although effective, these technologies require expensive equipment and very long and precise alignment times, which limits their use.
[0008] A second approach, called "silicon-based hybrid III-V integration," is based on the evanescent coupling between the waveguides of the two components. This type of coupling is schematically illustrated in the diagram. FIG. 1B In this example, the SiPho component comprising the waveguide 12 is fabricated on the silicon base plate 16, and then layers of active semiconductor materials are arranged on the SiPho component 10 using bonding techniques, for example, molecular bonding, direct epitaxy, or a polymer interface, to form the semiconductor component 11. The waveguide 13 is then defined within the stack of semiconductor material layers by lithography. The alignment accuracy between the waveguides 12 and 13 obtained by this method is approximately 0.2 µm (see, for example, Duan et al. [Ref 4] or the published patent application US2013195137 [Ref 5]). The method described above for obtaining active semiconductor waveguides precisely aligned with SiPho waveguides ensures evanescent coupling between the waveguides, thus limiting coupling losses even if the mode sizes differ. Furthermore, the active material can be positioned wherever required.
[0009] However, evanescent coupling techniques between waveguides in known III-V hybrid integration technologies on silicon require a complex and specific hybrid technology that is not yet mature. In particular, these techniques present efficiency and reliability problems, as well as limited reproducibility, notably due to the necessary bonding step between the SiPho component and the semiconductor material layers. Thus, today, very few laboratories have the capacity to develop this approach, and even fewer to demonstrate the feasibility of an industrial device.
[0010] Patent application WO2016077499 [Ref. 7] describes evanescent coupling techniques between waveguides for coupling a SiPho integrated circuit to an optical fiber or vice versa. However, as in [Ref. 4] and [Ref. 5], the active semiconductor components are coupled to the silicon waveguides of the SiPho integrated circuit by bonding (“ wafer bonding ".
[0011] One objective of this description is to propose a new coupling method based on evanescent coupling between SiPho and semiconductor components previously designed for this purpose, thus offering, compared to known methods, better reliability and greater simplicity in the process. Résumé de l'invention
[0012] According to a first aspect, the present description relates to an assembly of an active semiconductor component and a passive optical component based on silicon, said assembly being configured for the coupling of waveguides formed respectively in said active semiconductor component and passive optical component.
[0013] An assembly according to the first aspect is described in claim 1 and comprises a support, said active semiconductor component and said silicon-based passive optical component arranged on said support.
[0014] According to this description, said active semiconductor component comprises: a first set of layers of semiconductor materials comprising at least a first waveguide configured to guide in a first section of the assembly, at least a first optical mode; a second set of layers of semiconductor materials, superimposed and in contact with said first set of layers and comprising at least a second waveguide configured to guide at least a second optical mode.
[0015] Furthermore, according to this description, at least a portion of said layers of the first set of layers and of the second set of layers are doped to form in a first region of the active semiconductor component a PIN-type diode; said at least one first waveguide and said at least one second waveguide are configured to permit evanescent coupling between them in a second section of the assembly; and said first set of layers is etched to form in a second region of the active semiconductor component a first surface flush with said at least one second waveguide.
[0016] According to this description, said silicon-based passive optical component of said assembly comprises: a substrate; a set of layers of materials composed of silicon formed in a first region of the substrate and comprising at least one waveguide configured to guide at least one optical mode, said at least one waveguide being flush with a first surface of said set of layers opposite a second surface of said set of layers in contact with a surface of the substrate.
[0017] In the assembly according to the first aspect, said first surface of the passive optical component is in contact with said first surface of the active semiconductor component to allow evanescent coupling between said at least one waveguide of the passive optical component and said at least one second waveguide of said active semiconductor component.
[0018] In this description, "flush" means that the distance between at least one second waveguide and the first surface of the active semiconductor component, and between at least one waveguide and the first surface of the passive optical component, is sufficiently small to allow evanescent coupling between the second waveguide of the active semiconductor component and the waveguide of the passive optical component after assembly. In other words, during operation, there is a transfer of the at least one second optical mode propagating in the at least one second waveguide of the active semiconductor component to the at least one waveguide of the passive optical component. This distance depends on the geometry of the waveguides and the refractive indices of the materials used to form the waveguides.According to one or more embodiment examples, said distance is between 0 and about 1 µm, for example between 0 and about 500 nm.
[0019] In an assembly as described herein, the active semiconductor component can be designed and tested before assembly with the passive optical component. It enables very good coupling quality with a SiPho component by means of two evanescent couplings: firstly, between the first waveguide, which propagates the optical mode generated in the active region of the active semiconductor component, and an intermediate waveguide (the second waveguide); and secondly, between the intermediate waveguide and the waveguide of the SiPho component, even if the optical modes are of different sizes.
[0020] According to one or more exemplary embodiments, said semiconductor materials of the active semiconductor component include III-V semiconductor materials, for example GaAs, InP, GaN, ternary derivatives, for example InGaAs, InAlAs, InGaP, InGaN, or quaternary derivatives, for example InGaAsP, InGaAlAs.
[0021] According to one or more embodiment examples, the semiconductor material layers of the first set of layers of the active semiconductor component are configured to generate one or more optical functions chosen from: optical amplification, laser emission, phase modulation, electro-absorption modulation, photodetection.
[0022] In the case of optical amplification and laser emission, the first waveguide may comprise an optically gain semiconductor material and the active semiconductor component configured for current injection into said first waveguide. Depending on the optical functions (photodetection, modulation, amplification, emission) generated by the active semiconductor component and the passive optical component, the direction of light propagation will be either from the active semiconductor component to the passive optical component or from the passive optical component to the active semiconductor component.
[0023] III V semiconductor materials enabling the generation of these optical functions (apart from phase modulation) generally exhibit high optical propagation losses (greater than 20db / cm) at the propagation wavelength and are thus described as absorbing.
[0024] According to one or more exemplary embodiments, the respective geometries of said at least one first waveguide and at least one second waveguide of the active semiconductor component are configured such that an effective index of said at least one first optical mode is strictly greater than an effective index of said at least one second optical mode in the first section of the assembly and that said effective index of said at least one first optical mode becomes strictly less than the effective index of said at least one second optical mode in the second section of the assembly.
[0025] This inversion of the effective index of the optical modes allows the evanescent coupling transfer of the optical mode propagating in said at least one first waveguide into said at least one second waveguide of the active semiconductor component.
[0026] According to one or more exemplary embodiments, said at least a first waveguide and second waveguide of the active semiconductor component have variable widths in the direction of light propagation.
[0027] According to the present description, the first waveguide of the active semiconductor component is formed from a semiconductor material chosen from the ternary or quaternary derivatives of GaAs, InP or GaN and preferably from InGaP, InGaAsP or InGaAlAs possibly in the form of multi-quantum wells, buried in InP.
[0028] According to the present description, the second waveguide of the active semiconductor component is formed from a semiconductor material chosen from the ternary or quaternary derivatives of GaAs, InP or GaN and preferably from massive InGaAsP transparent at the propagation wavelength, buried in InP.
[0029] The transparency of the material is characterized in this description by low losses at the propagation wavelength, for example less than 10 dB / cm. With a transparent material for the second waveguide of the active semiconductor component, the applicant has shown that the propagation losses in the active semiconductor component can be substantially reduced, for example by a factor of 2 compared to a case where the material for the second waveguide of the active semiconductor component is absorbing and exhibits losses at the propagation wavelength greater than 20 dB / cm.
[0030] According to one or more exemplary embodiments, said first set of layers of the active semiconductor component comprises a first layer of semiconductor material arranged between a second layer of semiconductor material and a third layer of semiconductor material having refractive indices lower than the refractive index of the first layer of semiconductor material, said first layer forming said at least one first waveguide.
[0031] According to one or more exemplary embodiments, said second set of layers of the active semiconductor component comprises a fourth layer of semiconductor material arranged between said third layer of said first set of layers and a fifth layer of semiconductor material, said third and fifth layers having lower refractive indices than the refractive index of the fourth layer of semiconductor material, said fourth layer forming said at least a second waveguide.
[0032] According to one or more exemplary embodiments, a layer of the second set of layers of the active semiconductor component forms a substrate of the active semiconductor component.
[0033] According to one or more embodiment examples, said substrate of the active semiconductor component has a thickness of between about 200 µm and about 1 mm.
[0034] According to one or more exemplary embodiments, said substrate of the active semiconductor component is formed in a semiconductor material with a crystal lattice of the same size as that of the material arranged in contact, for example a binary III-V semiconductor material, for example InP, GaAs, GaSb.
[0035] According to one or more exemplary embodiments, said active semiconductor component comprises a plurality of N first waveguides arranged in parallel, and as many second waveguides, each second waveguide being coupled to a first waveguide by evanescent coupling in the second section of the assembly.
[0036] According to one or more exemplary embodiments, the active semiconductor component further comprises one or more additional waveguides arranged in series with the first waveguide(s). These additional waveguides can enable various functionalities, such as a laser waveguide upstream of a first waveguide forming an electro-absorption modulator. Only the first waveguide(s) are coupled to the second waveguide(s) by evanescent coupling in the second section of the component.
[0037] According to one or more embodiment examples, the substrate of the passive optical component is made of silicon.
[0038] According to one or more embodiment examples, the substrate of the passive optical component forms the support for the assembly.
[0039] According to one or more exemplary embodiments, the substrate of the passive optical component comprises a first region on which said set of layers of the passive optical component is arranged and a second region on which at least one electrical contact pad is arranged; said active semiconductor component comprises at least one electrical contact pad in contact with said first set of layers of semiconductor material; and said active semiconductor component is fixed on said substrate of the passive optical component such that said electrical contact pad of the passive component is in contact with said electrical contact pad of the active semiconductor component.
[0040] According to one or more embodiments, the electrical contact pads are metal pads glued or welded to secure the assembly. These metal pads may also provide power to the active semiconductor component.
[0041] The assembly thus created, which implements double evanescent coupling thanks to the presence of the second waveguide(s) as intermediate waveguides, offers the advantages of evanescent coupling, namely efficient coupling even if the optical modes are of different sizes, tolerance to misalignments in x and y, and extremely tolerant to the z direction. It also offers the advantages of direct coupling with a standard manufacturing process and efficient heat dissipation of the active semiconductor component fixed to the silicon substrate, for example by soldering.
[0042] According to one or more exemplary embodiments, said set of layers of the passive optical component comprises a first layer of silicon composite material arranged between a second layer and a third layer of silicon composite materials, having refractive indices lower than the refractive index of the first layer, said first layer forming said at least one waveguide of the passive optical component. The wafers, referred to as SOI (for " Silicon On Insulator " which are commonly used in Silicon microelectronics and Silicon Photonics are an example of a layer set that can be used to realize the passive optical component.
[0043] According to one or more embodiment examples, the said second and third layers are formed in silicon oxide (SiO2).
[0044] According to one or more exemplary embodiments, said first layer forming said at least one waveguide of the passive optical component comprises a material selected from doped SiO2, SiN, SiON, Si derivatives of Si, having a higher optical index than said second and third layers to ensure optical guidance.
[0045] According to one or more exemplary embodiments, the geometry of said at least one waveguide of the passive optical component is configured to exhibit an effective index of said at least one optical mode propagating in said at least one waveguide increasing in the direction of propagation to permit the transfer by evanescent coupling of said at least one optical mode into said at least one second waveguide of the active semiconductor component, after assembly.
[0046] In practice, an inversion of the effective index of the optical modes propagating respectively in said at least one waveguide of the passive optical component and said at least one second waveguide of the active semiconductor component will be observed to allow the transfer by evanescent coupling of said at least one optical mode of the passive optical component into said at least one second waveguide of the active semiconductor component.
[0047] According to one or more exemplary embodiments, the silicon-based passive optical component includes at least one of the following functions: wavelength multiplexing, polarization multiplexing, filtering, modulation, photodetection, beam combining or splitting, routing, and other functionalities suitable for use with known state-of-the-art Silicon Photonics (SiPho) circuits. According to one or more exemplary embodiments, the assembly comprises a plurality of N waveguides in the passive optical component and an equal number of first and second waveguides in the active semiconductor component, N being greater than or equal to 2, for example, N being between 2 and 20.In operation, each first waveguide of the active semiconductor component is coupled, by evanescent coupling, with a second waveguide of the active semiconductor component, said second waveguide of the active semiconductor component being coupled, by evanescent coupling, with a waveguide of the passive optical component.
[0048] According to a second aspect, the present description relates to a method of coupling between waveguides in an assembly according to the first aspect, the method comprising: in said first section of the assembly, the propagation of at least one optical mode in said at least one first waveguide of the active semiconductor component; in said second section of the assembly, the transfer, by evanescent coupling, of said at least one optical mode in said at least one second waveguide of the active semiconductor component; in a third section of the assembly, the propagation of said at least one optical mode in said at least one second waveguide of the active semiconductor component; in a fourth section of the assembly, the transfer, by evanescent coupling, of said at least one optical mode in said at least one waveguide of the passive component; in a fifth section of the assembly, the propagation of said at least one optical mode in said at least one waveguide of the passive component.
[0049] According to a third aspect, the present description relates to a manufacturing process for an assembly according to the first aspect, the process comprising: the supply of said active semiconductor component; the supply of said silicon-based passive component; the assembly of said active semiconductor component and said passive component by bringing said first surface of said set of layers of the passive component into contact with said first surface of the active semiconductor component.
[0050] According to one or more exemplary embodiments, the substrate of said passive component comprises a first region on which said set of layers is arranged and a second region on which at least one electrical contact pad is arranged; the active semiconductor component comprises at least one electrical contact pad in contact with said first set of layers of semiconductor material; and the assembly of said active semiconductor component and said passive component comprises bringing said electrical contact pad of the passive optical component into contact with said electrical contact pad of the active semiconductor component.
[0051] According to one or more embodiment examples, said electrical contact pads of the passive optical component and of the active semiconductor component are metallic pads and the assembly includes the fixing of said active semiconductor component and said passive component by welding or gluing said pads. Brève description des figures
[0052] Other advantages and features of the invention will become apparent upon reading the description, illustrated by the following figures: [ Fig. 1A ], already described, represents a diagram illustrating the direct coupling between an active semiconductor component and a silicon-based (SiPho) component, according to the prior art; [ Fig. 1B ], already described, represents a diagram illustrating the evanescent coupling between an active semiconductor component and a silicon-based (SiPho) component, according to the prior art; Fig. 2A ] represents a schematic diagram of an active semiconductor component according to an example from this description, in a cavalier perspective view; [ Fig. 2B ] represents a schematic of the active semiconductor component shown on the Fig. 2A , according to a section (section AA); [ Fig. 2C ] represents a schematic of the active semiconductor component shown on the Fig. 2A , according to a cut (BB cut); [ Fig. 2D ] represents a schematic of the active semiconductor component shown on the Fig. 2A , according to a section (CC section); [ Fig. 2E ] represents a schematic of the active semiconductor component shown on the Fig. 2A , according to a section (DD section); [ Fig. 3A ] represents a schematic diagram of a silicon-based passive component (SiPho) according to an example from this description, in a cavalier perspective view; [ Fig. 3B ] represents a diagram of the silicon-based passive component shown on the Fig. 3A , according to a section (section AA); [ Fig. 3C ] represents a diagram of the silicon-based passive component shown on the Fig. 3A , according to a cut (BB cut); [ Fig. 3D ] represents a diagram of the silicon-based passive component shown on the Fig. 3A , according to a section (CC section); [ Fig. 4A ] represents a diagram illustrating an assembly of an active semiconductor component as shown on the Fig. 2A with a passive component as illustrated on the Fig. 3A , according to a cross-sectional view, and a schematic representation of optical propagation in different sections of the assembly; [ Fig. 4B ] represents a diagram illustrating the evolution of the effective indices of modes propagating in the different waveguides as a function of the different sections of the assembly shown in Fig. 5A; [ Fig. 5 ] illustrates with diagrams the manufacturing steps of an active semiconductor component according to an example from this description; [ Fig. 6A ] represents a diagram showing the doping in an active semiconductor according to an example of an embodiment of this description; [ Fig. 6B ] represents a diagram showing the doping in an active semiconductor according to an example of an embodiment of this description; [ Fig. 6C ] represents a diagram showing the doping in an active semiconductor according to an example of an embodiment of this description; [ Fig. 7 ] illustrates with diagrams the manufacturing steps of a silicon-based passive component according to an example from this description; [ Fig. 8 ] represents a diagram of an assembly according to an example of the realization of this description. Description détaillée de l'invention
[0053] In the figures, the elements are not drawn to scale for better visibility. In particular, the horizontal and vertical axes are shown at different scales for improved clarity. Also, along each of the vertical and horizontal axes, the heights and widths of the different structures and layers can vary considerably, depending on the specific characteristics and constraints of the implementation.
[0054] THE FIGS. 2A à 2E represent an example of an active semiconductor component 200 configured for assembly with a passive silicon-based optical component, for example a component 300 as described on the FIGS. 3A - 3D .
[0055] The active semiconductor component 200 in this example includes a first set of layers 210 of semiconductor materials comprising at least a first waveguide 212 configured to guide at least a first optical mode, for example an optical mode generated in a gain region of the active component 200.
[0056] The active semiconductor component 200 further comprises a second set of layers 220 in contact with the first set of layers 210. The second set of layers 220 includes at least one second waveguide 222 configured to guide at least one second optical mode. The second set of layers may include a substrate 224, for example, an InP substrate.
[0057] As will be explained in more detail later by means of the FIGS. 6A - 6C , at least part of the layers of the first and second sets of layers are configured to form a PIN type diode, in a region 240 of the component, by means of contacts schematically represented in a simplified way by an electrical contact pad 213 in contact with a surface 211 of the first set of layers 210.
[0058] In the example of FIGS. 2A - 2E Only one first waveguide 212 and one second waveguide 222 are shown; in practice, as will be illustrated later in an example implementation (see FIG. 8 ), there can be several first waveguides arranged in parallel and as many second waveguides configured to guide a plurality of optical modes in parallel.
[0059] The first waveguide 212 and the second waveguide 222 are configured to allow, in operation, an evanescent coupling transfer of an optical mode propagating in the first waveguide 212 to the second waveguide 222.
[0060] Furthermore, as can be seen on the FIGS. 2A -2B , the first set of layers 210 is etched to form in a region 230 of the component 200 a first surface 231 flush with the second waveguide 222. The first surface 231 is configured to be in contact with a surface of the passive optical component for the evanescent coupling transfer of an optical mode propagating in the second waveguide into a waveguide of the passive optical component, as will be explained in more detail later.
[0061] For example, and not limited to, the first waveguide 212 is formed of a layer of quaternary semiconductor material, for example in bulk InGaAsP with a photoluminescence of 1.1 µm, encapsulated between two layers 216, 218 in InP to form, in the optical gain region, an InP laser.
[0062] The active semiconductor component as described herein can be configured to exhibit an optical amplification function or other optical functions.
[0063] Thus, depending on the desired optical function of the active component 200, the first waveguide 212 can be made of optically gaining material (vertical structure of the laser) and thus be subjected to electrical current to amplify an incident optical wave. It can also be made of a transparent or absorbent material at the operating wavelength, generally used under negative electrical voltage to modify the phase of the signal (e.g., Mach-Zehnder phase modulator, tunable laser section) or to absorb it (photodiode, electro-absorption modulator).
[0064] The second 222 waveguide can comprise a bulk material with photoluminescence close to the propagation wavelength, or "operational wavelength" (for example, about 1.3 µm for the O-band and 1.5 µm for the C-band of telecommunications). Advantageously, the second 222 waveguide is transparent at the propagation wavelength, that is, it exhibits losses at said wavelength of less than about 10 dB / cm.
[0065] THE FIGS. 3A à 3D represent an example of a passive silicon-based component 300 configured for assembly with an active semiconductor component, for example a component 200 as described on the FIGS. 2A - 2E .
[0066] The passive component 300 in this example comprises a substrate 310, for example a silicon substrate, and a set of layers 320 made of silicon-based materials, arranged in this example in a first region 330 of the substrate. The set of layers 320 includes at least one waveguide 312 configured to guide at least one optical mode. For example, the waveguide 312 is obtained by encapsulating a layer between two layers 316, 318 of lower refractive index. Multiple combinations of materials can be used to form the set of layers 320, such as, for example, but not limited to, a Si waveguide embedded in SiO2, or a SiN or doped SiO2 waveguide embedded in SiO2.
[0067] As before, in the example of FIGS. 3A - 3D , only one waveguide 312 is shown; in practice, as will be illustrated later in an example implementation (see FIG. 8 ), there may be several waveguides arranged in parallel for coupling with several second waveguides of the active semiconductor component.
[0068] As illustrated on the FIGS. 3A - 3D , the waveguide 312 is flush with a first surface 321 of the layer set 320 opposite a second surface 322 in contact with a surface 311 of the substrate.
[0069] The first surface 321 is configured to be in contact with the first surface (231, Figs. 2A - 2E ) of the active component for the evanescent coupling between the waveguide 312 and the second waveguide 212 of the active semiconductor component, as will be detailed later.
[0070] Furthermore, the active passive component in this example includes an electrical contact pad, advantageously a metallic pad, arranged in a second region 340 of the substrate 310.
[0071] There Fig. 4A represents a diagram illustrating an assembly 400 of an active semiconductor component 200 as illustrated on the FIGS. 2A - 2E with a passive component as illustrated on the FIGS. 3A - 3D .
[0072] As illustrated on the FIG. 4A The active semiconductor component 200 is inverted and positioned on the substrate 310 of the passive component 300, which in this example forms a support for the assembly 400. Alignment marks and mechanical stops can be implemented for the xy alignment of the waveguides of the active and passive components. During assembly and alignment, the first surface 321 of the layer assembly 320 of the passive component 300 is brought into contact with the first surface 231 of the active component. This ensures evanescent coupling between the second waveguide 222 of the active component and the waveguide 312 of the passive component.
[0073] Once the alignment has been achieved and the surfaces 321 and 231 are in contact, the assembly can be secured by means of gluing or welding between the metal pads on the two components (213, FIG.2A and FIGS. 7A / 7B and 313, FIGS. 3A - 3D ), according to known methods. These metallic welds can also provide power to the active component 300 and create an efficient heat dissipation path for the active semiconductor component.
[0074] One welding method uses metal balls, such as gold-tin, as pads 313. These balls are welded to the metal contact 213 of the active component by mechanical compression and heating above the melting point of the balls. Another method uses gold pads 313 onto which the metal contacts 213 of the active component are compressed. Regardless of the welding method used, the geometries of the two components 200 and 300, as well as the thicknesses of the weld pads and the metal contacts of the active component, can be chosen so that vertical alignment is ensured by bringing the surfaces 231 and 321 into contact, thereby ensuring mode transfer by evanescent coupling.
[0075] To avoid any optical reflection, a medium with an index of 410 close to that of SiO2 (such as BCB or an index matching polymer for example) can be inserted in the free space located in section 3 of the assembly, between the active and passive components, thus avoiding an abrupt change in the effective index in the vicinity of the propagating optical mode.
[0076] As illustrated on the FIG. 4A The evanescent coupling between the intermediate waveguide (second waveguide 222) and the waveguide 312 of the SiPho 300 component is achieved through surface contact. Therefore, the waveguides are arranged close to the surface, but no further vertical alignment with specific stop layer thicknesses is required. This technique is compatible with multiple technologies and paves the way for passive alignment in mass production using standard production equipment.
[0077] There FIG. 4A illustrates a technical advantage of the assembly as described herein. In the case of a vanishing coupling according to the state of the art, as described for example on the FIG. 1B The active area 13 is bonded directly to the passive component over a thick layer (approximately 2 µm) of SiO2. This material has very poor thermal conductivity, resulting in poor dissipation of the heat generated on the active component and therefore degraded performance (gain and reliability in particular). In the assembly described herein, the solder area of the active component 200 is offset from the mode transition area by evanescent coupling. This allows the use of standard metallic solder, providing excellent heat dissipation. Furthermore, this solution is compatible with any electrical trace design on the passive component, especially traces suitable for high-frequency signal transmission.
[0078] Note that in the example of the Fig. 4A The substrate 310 of the passive optical component 300 forms the support for assembly. In other embodiments, depending on the application, the passive optical component 300 and the active semiconductor material component 200 may be arranged on an independent support, which may make the assembly and alignment of the components more complex but the fabrication of the passive optical component easier.
[0079] A coupling method between the first waveguide 212 of the active semiconductor component 200 and the waveguide 312 of the silicon-based passive component 300 thus includes in a first step the design and manufacture of the active semiconductor component and passive component conforming to this description, a common design between the two components allowing to ensure proper operation of the whole.
[0080] More specifically, since each component has layers of material with a refractive index known to those skilled in the art, the common design involves modifying the geometry of the various waveguides of both components to enable evanescent coupling and optimal optical mode transfer (maximum optical power transferred without reflection). To achieve this, well-known software, such as, but not limited to, Fimmwave, OptimBPM, Beamprop, and Lumerical, can be used to determine the propagation of optical modes in the different waveguides. This software is based on known calculation methods (e.g., Finite Difference Time Domain (FDTD), Beam Propagation Method (BPM), and Eigen Mode Expansion (EME)). For example, pointed waveguides such as those described in Chapter 2 of Marco Lamponi's thesis [Ref. 6] could be used.
[0081] The active and passive components can be manufactured using conventional methods, examples of which are given with reference to FIGS. 5 , 6A - 6C And 7 .
[0082] The assembly is then carried out as described, for example, with reference to the FIG. 4A In operation, the coupling between the waveguides then occurs in the following way. The first optical mode generated, for example, in the active region of the semiconductor component 200 and guided in the first waveguide 212 (section 1 of the assembly, FIG. 4A ) is transferred by evanescent coupling in the second waveguide 222 (section 2). This coupling is controlled by the manufacturing process and can be made even more efficient by the tip of the first waveguide 212 (possibly down to 0.7µm in lithography, 0.2µm in e-beam technology) and the inverted tip shape of the second waveguide 222. The tip shapes can be adjusted by simulation for optimal coupling and manufacturing tolerance, as described previously.
[0083] In a third section of the assembly (section 3, FIG. 4A ), the optical mode remains in the second waveguide 222. To avoid any optical reflection, a medium with an index of 410 close to that of SiO2 (such as BCB for example) can be inserted in the free space between the components.
[0084] The optical mode is then evanescently coupled (section 4) from the second waveguide 222 into the Si-based waveguide 312 (Si, SiO2, SiN...) again using optimized waveguide shapes (pointed and / or inverted point) (as described in [Ref. 6]) and then propagates (section 5) into the SiPho component waveguide 312.
[0085] The two components, semiconductor and SiPho, can simply be in surface contact, ensured by the pressure applied during the soldering of the semiconductor component. The distance between the Si-based waveguide and the second 222 waveguide is controlled by the thickness of the etch in the SiPho component and the SiO2 overlay layer. By choosing a highly guiding Si-based waveguide (high contrast ratio), it is possible to obtain a strong modal attraction for the Si-based waveguide, allowing for high tolerance to y-axis alignment and full tolerance to x-axis alignment when using sufficiently long tips.
[0086] There FIG. 4B represents a diagram illustrating, by way of example, the evolution of the effective indices of modes propagating in the different waveguides as a function of the different sections of the assembly shown on the FIG. 4A . Curves 401, 403, 405 thus represent respectively the effective mode index in waveguides 212, 222, 312.
[0087] The index and geometry of the first waveguide 212 in section 1 are defined by the desired function and performance of the laser component or the InP circuit upstream of section 1. In this section 1, the aim is to ensure that the effective index of the mode (calculated taking into account both the geometry of the waveguide and the indices of the materials forming and surrounding it) propagating in the first waveguide 212 (curve 401) is greater than the effective index of an optical mode that could propagate in the second waveguide 222 (curve 403). This requirement of a higher effective index in the first waveguide compared to the second waveguide ensures that the optical mode propagates only in the first waveguide in section 1.
[0088] In Section 2, we vary the geometry of the two waveguides in this example (decreasing the width of one guide and increasing the width of the other, for example) in order to modify the effective indices of the modes propagating in the guides and to achieve, at the end of Section 2, a reversal of the previous condition. In other words, we aim for the effective index of the mode propagating in the first waveguide 212 (curve 401) to be lower than the effective index of an optical mode that could propagate in the second waveguide 222 (curve 403). This reversal of the condition will allow the optical mode to transfer from the first waveguide to the second waveguide.
[0089] Once the mode transfer has been completed at the end of section 2, the optical mode will propagate and stabilize in section 3. Then in section 4, the optical mode is transferred from the second waveguide 222 to the waveguide 312 of the SiPho component.
[0090] Once the optical mode has been transferred into the waveguide 312 at the end of section 4, the optical mode will stabilize in section 5 and will be able to propagate into the rest of the SiPho component.
[0091] There Fig. 5 illustrates manufacturing steps of an example of an active semiconductor component according to an example of this description.
[0092] A structure comprising a stack of layers of semiconductor materials is fabricated by epitaxial growth on a substrate 524 (step 501). The layer stack includes, for example, in addition to the substrate 524 (e.g., InP), layers 522 and 512 for the formation of waveguides 212 and 222, for example, InGaAsP, encapsulated between layers 518 and 516. Depending on the type of active components desired, those skilled in the art will use the materials necessary to achieve the intended functionality. By way of example, III-V semiconductor materials, for example GaAs, InP, GaN, with some of their ternary derivatives, for example InGaAs, InAlAs, InGaP, InGaN, or quaternary derivatives, for example InGaAsP, InGaAlAs, may be used.
[0093] A series of photolithography, dielectric deposition, and dielectric and semiconductor etching steps defines the two waveguides (step 502). The first waveguide 212 is then etched (step 503) to create a tip called a mode converter within the first waveguide. This promotes evanescent coupling to the second waveguide 222, forming an open region 230 without the first waveguide. A thin etch stop layer a few nm above the second waveguide can be used to control the depth of this etch.
[0094] The entire structure is then embedded (step 504) in a top coating (for example, p-doped InP in the case of InP laser formation). For this, either the open region 230 is locally protected during the re-epitaxial process, or it is opened by etching, reusing the same stop layer as for the initial etching. Finally, the fabrication is completed (step 505), with the standard steps of depositing metallization pads 213, which will later be used for component assembly and power supply. Passivation and thinning steps are generally performed at the end of this fabrication process.
[0095] In the example of the FIG. 5 The active semiconductor component is fabricated using a technology of optical ribbons embedded in semiconductor material. The invention can also be extended to other types of technology depending on the semiconductor technology used.
[0096] The doping of the active component 200 to form a PIN diode can be done in various ways known from the state of the art.
[0097] THE FIGS. 6A - 6C This represents different examples of doping in which 601 represents an N-doped region, 602 represents a P-doped region, 603 represents an undoped region, and 604 represents an electrically insulating region. The P- and N-doped regions are in contact with contacts 213 as described previously.
[0098] More specifically, the FIGS. 6A - 6C These represent three possible implementation schemes for a PIN diode-type heterostructure for the active component. FIG. 6A features a vertical PIN diode obtained using conventional ribbon-type technology, the FIG. 6B features a vertical PIN diode obtained by burying the optical strip and the FIG. 6C A lateral PIN diode is obtained by burying an optical strip and ion implantation to define the lateral diode. Other PIN diode configurations known to those skilled in the art may be compatible with the invention and the manufacturing process described with reference to the FIG. 5 .
[0099] There FIG. 7 illustrates manufacturing steps of an example of a silicon-based passive component according to an example of this description.
[0100] The manufacturing steps of a SiPho component are standard and may include the formation of a structure (step 701) by growth or by the Smartcut process (SOI wafer fabrication method) on the silicon substrate 310 of a SiO2 layer 516 and a layer 512 intended to form the waveguide, for example, a Si layer. Then the Si waveguide is etched (step 702) and the structure covered with a SiO2 layer (step 703) to form the set of layers 320. Next, a trench is etched (step 704), thus forming two regions 330 and 340, with the set of layers 320 located in region 330. Electrical contact pads 313, for example, metallic solder pads, are deposited in region 340 on the substrate 310 (step 705) to allow for soldering and electrical contact with the active semiconductor component.
[0101] There Fig. 8 represents an assembly 800 conforming to this description in an example implementation, comprising an example active semiconductor component 200 and an example passive component 300.
[0102] More specifically, assembly 800 illustrates the multiplexing of 4 lasers at different wavelengths for modulation in silicon.
[0103] In this example, the active semiconductor component 200 includes an optical gain region 802 in which the first waveguides 212 are configured to guide optical modes generated by four InP lasers emitting at four different wavelengths. Since the assembly method does not require molecular bonding or growth on a Si substrate of the same type as that described in the prior art in relation to the figure 1B The active semiconductor component may include other active sections (e.g., amplifiers), modulators, and / or passive sections, e.g., back-facing InP photodiodes arranged in a region 801 of the assembly 800, as illustrated in the FIG. 8 References 811 and 812 indicate the metal contacts of the photodiodes and laser, respectively. The photodiodes comprise all the semiconductor stack located under metal 811 as well as metal 811 itself, and the laser comprises all the semiconductor stack located under metal 812 as well as metal 812. The passive component 300 in this example comprises a multiplexer 814 in a region 804 of the assembly connected to a modulator 815 in a region 805 of the assembly, for example, a Mach-Zehnder modulator.
[0104] Region 803 of the assembly corresponds to a coupling region as described in more detail by means of the FIGS 4A, 4B In particular, we observe the evanescent coupling between the second waveguides 222 of the active component 200 and the waveguides 312 of the passive component connected to the multiplexer 814.
[0105] Of course, the assembly example illustrated on the FIG. 8 This list is not exhaustive, and other examples could be given that use one or more assemblies connected one after the other. For example: the emission of 2 orthogonally polarized lasers in the active component and combination of the two polarizations in silicon and transfer to an optical fiber using a vertical coupler; the coupling of an external multi-wavelength laser emitting a frequency comb, demultiplexing in a passive component, modulation of the different wavelengths in an array of electro-absorption modulators in an active component, multiplexing in a second passive component; gain section in an active component, coupled with a tunable selective filter based on rings or Mach-Zehnder in the passive component to create a wavelength-tunable laser cavity; emission in a first active component of 4 lasers with different wavelengths, multiplexing and Mach-Zehnder modulation, demultiplexing in a passive component and photodetection in a second active component.
[0106] Although described through a number of embodiment examples, the assembly and coupling method described in this description include various variants, modifications and improvements which will be obvious to those skilled in the art, it being understood that these various variants, modifications and improvements form part of the scope of the invention as defined by the following claims. Références
[0107] Réf. 1. K. Janiak et al., "1.55 µm BH-DFB laser with integrated spot-size converter for flip-chip applications" 16th IPRM. 2004 International Conference on Indium Phosphide and Related Materials, 2004., Kagoshima, 2004, pp. 476-479. Réf. 2. Yamada, Hirohito. "Analysis of Optical Coupling for SOI Waveguides." Piers Online, vol. 6, No. 2, 2010. Réf. 3. Y. Gao et al., "Hybrid Integration With Efficient Ball Lens-Based Optical Coupling for Compact WDM Transmitters" in IEEE Photonics Technology Letters, vol. 28, no. 22, pp. 2549-2552, 15 Nov.15, 2016. Réf. 4. G. Duan et al., "Hybrid III--V on Silicon Lasers for Photonic Integrated Circuits on Silicon" in IEEE Journal of Selected Topics in Quantum Electronics, vol. 20, no. 4, pp. 158-170, July-Aug. 2014, Art no. 6100213. Réf. 5. Demande de brevet publiée US2013 / 0195137A1. Réf. 6. Marco Lamponi, "Hybrid III-V on silicon lasers for telecommunication applications", PhD Thesis, Chapter 2, 2012. Réf. 7 : Demande de brevet publiée WO2016077499
Claims
1. An assembly (400) of an active semiconductor component (200) and of a silicon-based passive optical component (300), said assembly comprising: a carrier; said active semiconductor component (200) and said passive optical component (300) both arranged on said carrier, wherein said active semiconductor component (200) comprises: a first set (210) of semiconductor layers comprising at least one first waveguide (212) configured to guide, in a first section (1) of the assembly, at least one first optical mode, said first waveguide (212) being made of a semiconductor material selected from ternary or quaternary derivatives of GaAs, InP, or GaN.; a second set (220) of semiconductor layers, said set being superposed and contacting said first set of layers, and comprising at least one second waveguide (222) configured to guide at least one second optical mode, said second waveguide (222) being made of a semiconductor material selected from ternary or quaternary derivatives of GaAs, InP, or GaN; wherein at least some of said layers of the first set of layers and of the second set of layers are doped to form, in a first region (240) of the component, a PIN diode; said at least one first waveguide (212) and said at least one second waveguide (222) are configured to allow evanescent coupling therebetween, in a second section (2) of the assembly; said first set of layers is etched to form, in a second region (230) of the active semiconductor component (200), a first surface (231) flush with said second waveguide (222); wherein said passive component comprises: a substrate (310); a set (320) of silicon-compound layers comprising at least one waveguide (312) configured to guide at least one optical mode; and wherein: said at least one waveguide (312) lies flush with a first surface (321) of said set (320) of layers, which surface is opposite a second surface (322) of said set of layers contacting a surface (311) of the substrate; and wherein said first surface (321) of the passive optical component contacts said first surface (231) of said active semiconductor component (200) in order to allow evanescent coupling between said at least one waveguide (312) of the passive optical component and said at least one second waveguide (222) of said active semiconductor component.
2. The assembly as claimed in claim 1, wherein the carrier of the assembly is formed by said substrate (310) of the passive optical component.
3. The assembly as claimed in claim 2, wherein: the substrate (310) of said passive component comprises a first region (330) in which is arranged said set (320) of layers, and a second region (340) in which is arranged at least one electrical contact pad (313); said active semiconductor component comprises at least one electrical contact pad (213) contacting said first set (210) of semiconductor layers; said active semiconductor component is fastened to said substrate (310) of the passive optical component in such a way that said electrical contact pad (313) of the passive component contacts said electrical contact pad (213) of the active semiconductor component.
4. The assembly as claimed in any one of the preceding claims, wherein the geometries of said at least one first waveguide and at least one second waveguide of the active semiconductor component are configured such that an effective index of said at least one first optical mode is strictly higher than an effective index of said at least one second optical mode in the first section of the assembly and such that said effective index of said at least one first optical mode becomes strictly lower than the effective index of said at least one second optical mode in the second section of the assembly.
5. The assembly as claimed in any one of the preceding claims, wherein a distance between said at least one second waveguide and said first surface (231) of the active semiconductor component is comprised between 0 and about 1 µm.
6. The assembly as claimed in any one of the preceding claims, wherein a distance between said at least one waveguide and said first surface (321) of the passive optical component is comprised between 0 and about 1 µm.
7. The assembly as claimed in any one of the preceding claims, wherein the semiconductor layers of the first set of layers of the active semiconductor component are configured to perform one or more optical functions chosen from: optical amplification, laser emission, phase modulation, electro-absorption modulation, and photodetection.
8. The assembly as claimed in any one of the preceding claims, wherein the layers of the set of layers of the passive optical component are configured to perform one or more of the following functions: wavelength multiplexing, polarization multiplexing, filtering, modulation, photodetection, beam combining or splitting, and routing.
9. The assembly as claimed in any one of the preceding claims, wherein: the passive optical component comprises a plurality of N waveguides placed in parallel, where N is an integer higher than or equal to 2; the active semiconductor component comprises a plurality of N first waveguides and a plurality of N second waveguides, said first waveguides and said second waveguides being placed in parallel, and wherein: in operation, each first waveguide of the active semiconductor component is coupled, by evanescent coupling, to a second waveguide of the active semiconductor component, said second waveguide of the active semiconductor component being coupled, by evanescent coupling, to a waveguide of the passive optical component.
10. A method for coupling waveguides in an assembly as claimed in any one of the preceding claims, comprising: in said first section (1) of the assembly, propagating at least one optical mode through said at least one first waveguide (212) of the active component; in said second section (2) of the assembly, transferring, by evanescent coupling, said at least one optical mode to said at least one second waveguide (222) of the active component; in a third section (3) of the assembly, propagating said at least one optical mode through said at least one second waveguide (222) of the active component; in a fourth section (4) of the assembly, transferring, by evanescent coupling, said at least one optical mode to said at least one waveguide (312) of the passive component; in a fifth section (5) of the assembly, propagating said at least one optical mode through said at least one waveguide (312) of the passive component.
11. A process for fabricating an assembly as claimed in any one of claims 1 to 9, comprising: providing said active semiconductor component (200); providing said silicon-based passive component (300); assembling said active semiconductor component (200) and said passive component (300) by placing said first surface (321) of said set (320) of layers of the passive component (300) in contact with said first surface (231) of the active semiconductor component (200).
12. The process for fabricating an assembly as claimed in claim 11, wherein: the substrate (310) of said passive component comprises a first region (330) in which is arranged said set (320) of layers and a second region (340) in which is arranged at least one electrical contact pad (313); said active semiconductor component comprises at least one electrical contact pad (213) contacting said first set (210) of semiconductor layers; and wherein: assembling said active semiconductor component (200) and said passive component (300) comprises placing said electrical contact pad (313) of the passive optical component in contact with said electrical contact pad (213) of the active semiconductor component.
13. The process for fabricating an assembly as claimed in claim 12, wherein said electrical contact pads (313, 213) of the passive optical component and of the active semiconductor component are metal pads and assembling comprises fastening said active semiconductor component (200) and said passive component (300) by soldering or bonding said pads.