Photonic circuit unit with on-chip measurement structures for optical amplification

Abstract

A photonic circuit unit (100) for measuring optical gain, comprising: - a substrate (10) with a photonic circuit, the latter having two waveguides (71, 72) defining two waveguide sections aligned in the same direction and opposite each other, - an active amplification section (62 to 66) on top of the substrate (10) and coupled in the photonic circuit unit (100) to generate light by electrical pumping or optical pumping, - at least two optical couplers (75, 76) arranged such that at least part of the active gain section (62-66) lies between the optical couplers (75, 76) and configured such that light is coupled between the active gain section and the waveguide sections, - a partial reflector (90) arranged in one of the waveguide sections such that it reflects light propagating along the same direction back to a center of the amplification section (62-66), and wherein the photonic circuit unit (100) has no further reflector opposite the partial reflector (90) with respect to the active amplification section (62-66) and is configured to reflect light back to the center of the amplification section.

Description

FIELD OF THE INVENTION

[0001] The invention relates generally to the field of photonic circuit devices and, more particularly, to the measurement of optical gain. BACKGROUND OF THE INVENTION

[0002] Silicon photonics refers to photonic systems in which silicon is used as a means of light propagation due to the material's low losses. Silicon photonics utilizes long-established silicon fabrication processes used in complementary metal-oxide-semiconductor (CMOS) electronics. The elements are typically patterned into microphotonic components with sub-micrometer precision (to operate in the infrared range). Silicon-on-insulator (SOI) is commonly used as the preferred material. The fabrication of silicon photonic devices can further incorporate known semiconductor fabrication techniques; since silicon is already used as a preferred substrate for most integrated circuits, it is possible to fabricate hybrid devices in which the optical and electronic components are integrated on a single chip.

[0003] To meet the requirements of future computer systems, high-speed and energy-efficient alternatives for on-chip electrical interconnects are needed. Integrated optics, especially silicon photonics, fulfills these requirements. Integrated optical interconnects with suitable light sources are required for the mass production of high-performance, low-cost CMOS-based chips. Due to the indirect band gap of silicon, there is no Si-based light source available. Efficient light sources typically use III-V semiconductors integrated heterogeneously or hybridly on a Si photonics platform.

[0004] To date, the most promising approaches have focused on bonding a stack of fully epitaxial III-V-based gain layers, or a thin seed layer that can undergo subsequent regrowth, on top of the silicon optical waveguides. In any case, in order to measure, characterize, or evaluate laser structures or light sources, special test structures are required to gain insight into key device parameters. One of the most important parameters is optical gain. Optical gain calculations are not very precise, and therefore, optical gain measurements are necessary to model gain layers and thus optimize active optical devices, such as lasers or optical amplifiers. However, gain measurement is very difficult for on-chip applications such as silicon photonics.The reason is that existing techniques and methods use cleavage facets or require numerous devices, which is unattractive for implementation in on-chip applications because of the time-consuming or destructive nature of the process and / or the large area required. Furthermore, such techniques may require high-resolution spectrometers to resolve the oscillations of the output power spectrum. In addition, the majority of existing concepts can only exploit the optical gain below or around the threshold, which is typical for operating modes for a laser device. Thus, it is impossible for known standard techniques to characterize the gain properties at the operating conditions of the gain material under realistic pumping conditions used in light-emitting devices of interest (i.e., at pumping current densities corresponding to the operating conditions of the respective laser).

[0005] Documents have already been published in this context. Document US 2006 / 0 013 273 A1 describes a monolithic asymmetric laser with a stabilized wavelength. This laser uses a waveguide splitter having at least two branches. In addition, two non-identical resonators with different wavelengths are used. The resonators are coupled to each branch of the splitter, with a photodiode communicatively coupled to receive output signals from each of the resonators. Document WO 2011 / 069 225 A1 describes a detector associated with a waveguide that receives an optically pre-amplified signal, which is subject to bandpass wavelength filtering. Furthermore, document EP 2 439 822 A1 describes a monolithically integrated structure with a "buried" semiconductor optical amplifier having a heterostructure and a photodetector.In addition, document US 2007 / 0 110 357 A1 describes another integrated photonic amplifier and detector based on an asymmetric twin waveguide structure, whereby the structure is manufactured together with the integrated amplifier and detector in a single active waveguide layer. Finally, document US 2008 / 0 273 567 A1 describes hybrid waveguide systems and corresponding methods. A III-V semiconductor waveguide is coupled with a Si waveguide in the form of a hybrid structure.

[0006] Nevertheless, there is currently no suitable device that allows to measure the gain of an on-chip optical device in a simple manner, i.e., without manufacturing a large number of devices, cleaving the device or polishing its facets, attaching multiple contacts to the gain measurement structure, or measuring the gain precisely above the threshold under realistic pumping / operating conditions. SUMMARY OF THE INVENTION

[0007] This stated object is achieved by the subject matter of the independent patent claims. Further embodiments are provided by the respective dependent patent claims. According to a first aspect, the present invention is provided as a photonic circuit unit for measuring optical amplification, comprising: - a substrate with a photonic circuit, the latter comprising one or more waveguides defining two waveguide sections aligned in a same direction; - an active gain section on top of the substrate and coupled in the device to generate light by electrical pumping or optical pumping; - at least two light couplers arranged such that at least a part of the active gain section lies between the light couplers, and configured such that light is coupled between the active gain section and the waveguide sections; and - a partial reflector arranged to reflect light propagating along the same direction back to a center of the gain section, and wherein the unit has no further reflector opposite the partial reflector with respect to the active gain section and is configured to reflect light back to the center of the gain section.

[0008] In some embodiments, the light couplers extend longitudinally along the same direction.

[0009] Preferably, the light couplers each have at least one tapered section, wherein the tapered section terminates a waveguide section or is connected to the active gain section, and wherein preferably the tapered sections are each substantially parabolic.

[0010] In preferred embodiments, the light couplers each comprise two tapered sections oriented opposite one another and overlapping, wherein a first of the two tapered sections terminates an associated waveguide section and a second of the two tapered sections is connected to the active gain section.

[0011] Preferably, each of the waveguide sections comprises an additional light coupler, the partial reflector being located between one of the light couplers and such additional light coupler, each additional light coupler preferably being located at one end of each of the waveguide sections opposite the other end of each of the waveguide sections closest to the light couplers.

[0012] In some embodiments, the additional light couplers are grating couplers configured to enable vertical measurement of the light emitted at the additional light couplers.

[0013] Preferably, the photonic circuit is a silicon photonic circuit, and more preferably, the substrate further comprises an electrical circuit in addition to the photonic circuit.

[0014] In preferred embodiments, the waveguide sections each extend directly on a dielectric layer of the device, wherein the dielectric layer preferably lies on top of the substrate and has a thickness of more than 1 micrometer.

[0015] Preferably, each of the one or more waveguides is in contact with a bonding layer and is preferably partially embedded in the bonding layer, and more preferably has a surface that is planar with a surface of the bonding layer, wherein the bonding layer is preferably one of: a polymer; SiO2; or Al2O3 or any combination thereof.

[0016] In some embodiments, the gain section comprises one or more of: a III-V semiconductor material, a II-VI semiconductor material, a semiconductor such as germanium, semiconductor alloys such as silicon germanium, a polymer, a material having embedded quantum dots and / or quantum wells, or a quantum well material.

[0017] Preferably, the reinforcement section is a reinforcement section made of III-V semiconductor material, comprising: - a lower contact layer having a first metal contact; and - having an upper part on top of the lower contact layer and a second metal contact.

[0018] In preferred embodiments, the reinforcement section comprises a stack of epitaxial layers, the latter comprising an n-doped semiconductor, at least a part of which forms at least a part of the lower contact, and wherein the stack of epitaxial layers further comprises an upper stack forming at least a part of the upper part of the reinforcement section of III-V semiconductor material, and which in turn comprises: - a first intrinsic semiconductor on top of the n-doped semiconductor; - a multiple quantum well section on top of the first intrinsic semiconductor; - a second intrinsic semiconductor on top of the multiple quantum well section; and - a p-doped semiconductor on top of the second intrinsic semiconductor.

[0019] In particularly preferred embodiments, the photonic circuit comprises two waveguides, each defining one of the waveguide sections; and the light couplers are configured to enable light coupling between, on the one hand, the active gain section and, on the other hand, each of the two waveguides, wherein each of the two light couplers has two tapered sections oriented opposite each other and overlapping, and wherein: - a first of the two tapered sections forms one end of one of the waveguide sections; and - a second of the two tapered sections is connected to the active reinforcement section and preferably forms part thereof.

[0020] In some variants, the photonic circuit comprises only one waveguide defining the two waveguide sections, wherein the waveguide preferably has a different cross-section, particularly preferably having a central section having a reduced width compared to outer sections of the waveguide, wherein the outer sections define the two waveguide sections.

[0021] According to a further aspect, the invention may be embodied as a method for measuring optical gain, comprising: Providing a unit according to any one of the above-mentioned embodiments, wherein the two waveguide sections comprise a first waveguide section and a second waveguide section; Exciting the gain section to generate light therein by electrical pumping or optical pumping; Transmitting the generated light to each of the two waveguide sections with the aid of the light couplers, whereby the light is partially reflected at the partial reflector; and Sampling both the optical power of the light emitted by the first waveguide section and the optical power of the light emitted by the second waveguide section to determine the optical gain of the active gain section, wherein the sampling is carried out such that: the light emitted from the first waveguide section comprises both: Light that is generated and transmitted directly to the first waveguide section; as well as Light that is generated and reflected at the partial reflector and then transmitted to the first waveguide section, and wherein the light emitted from the second waveguide section comprises light that has not been reflected by the partial reflector.

[0022] In some embodiments, each of the waveguide sections of the provided unit comprises an additional light coupler, the partial reflector being arranged between one of the light couplers and such additional light coupler, the latter preferably being located at one end of each of the waveguide sections opposite the other end of each of the waveguide sections closest to the light couplers; and the optical power of light emitted at each of the additional light couplers is sampled to determine the optical gain of the active gain section.

[0023] Preferably, the optical power sampling is performed via optical fibers or photodetectors located near each of the additional light couplers.

[0024] Units, devices, and methods embodying the present invention will now be described by way of non-limiting example and with reference to the accompanying drawings. Technical features shown in the drawings are not to scale. BRIEF DESCRIPTION OF SEVERAL VIEWS OF THE DRAWINGS Fig. 1 to 6, 7 and 8A show 3D views of simplified representations of photonic circuit devices at various stages of their fabrication and according to the embodiments. Fig. 1 to 5 show units in intermediate stages of production; Fig. 6B and Fig. 8B show schematically a plan view of the units of the Fig. 6A or 8A; Fig. 9 shows the same top view as Fig. 8B and further shows how light is generated, transmitted, propagated and reflected therein, according to embodiments of methods according to the invention; Fig. Figure 10 is a flowchart showing high-level steps of methods for measuring optical gain in units such as those described in the Fig. 1 to 9 and 11 to 15, according to some embodiments; Fig. Figure 11 is a sectional view of a simplified representation of a unit such as that shown in Fig. 8 is shown; Fig. 12 is a sectional view of a simplified representation of a variant of the unit of Fig. 8; Fig. 13 is a sectional view of a simplified representation of another variant of the unit of Fig. 8; Fig. Figure 14 is a (longitudinal) side view of a simplified representation of a unit such as that shown in Fig. 8 is shown; and Fig. 15 is a cross-sectional view of a stack of epitaxial layers forming (part of) a gain section of III-V semiconductor material, as implemented in some embodiments. DETAILED DESCRIPTION OF THE INVENTION

[0025] The following description is structured as follows. First, general embodiments and higher-level variants are described (Section 1). The following section refers to more specific embodiments and details of the technical implementation (Section 2). 1. General embodiments and higher-level variants

[0026] With reference to the Fig. 1 to 9 and 11 to 15, an aspect of the invention will first be described, which concerns a photonic circuit unit 100 for measuring optical gain. This unit is provided with on-chip gain measurement structures, allowing the optical gain of an active gain section to be easily measured.

[0027] First, the unit 100 comprises a substrate 10 with a photonic circuit. The substrate is preferably a silicon wafer, but can also be made of gallium arsenide (GaAs) or indium phosphide (InP). In particular, the substrate can comprise: a silicon photonic circuit; a passive InP photonic circuit; or a passive GaAs photonic circuit. It will be referred to as a "wafer" for simplicity. Nevertheless, a typical product may comprise a single chip, as is common in the art.

[0028] The photonic circuit comprises one or more waveguides 71, 72. In most embodiments described here, it comprises two waveguides, with an exception to the embodiments with reference to Fig. 3. In any case, the one or more waveguides 71, 72 define (at least) two waveguide sections aligned in a same direction, as in the Fig. 1 to 9. "Waveguide section" refers to a core section of a waveguide, as is common in integrated photonics or silicon photonics. Surrounding components, layers, or materials, etc., serve as cladding. Other components of the photonic circuit are not shown in the accompanying drawings for the sake of brevity.

[0029] The device 100 further comprises an active gain section 62 to 66 on top of the wafer and connected within the device to generate light by electrical pumping or optical pumping. The active gain section 62 to 66 can, for example, be bonded (directly or indirectly) to the top of the wafer. In some variants, there can be an interface therebetween. The active gain section can be directly bonded by molecular bonding or a layer of polymer or SiO2 or even a bilayer of Al2O3 and SiO2 or a combination thereof. However, the use of a bilayer of Al2O3 and SiO2 is preferred because SiO2 is a standard material in CMOS processes and Al2O Sthe binding energy is increased. Note that "on top of" means above (or below) and not necessarily "in direct contact with," as is consistently assumed in the prior art. Thus, if a layer a of material A is "on top of" a layer b of material B, there is at least a partial overlap between layers a and b.

[0030] At least two light couplers 75, 76 are further provided, which are arranged so that at least a part of the active amplification section lies between the light couplers, as can be better seen in the Fig. 6 to 8. These couplers are each configured to couple light between the active gain section and the waveguide sections. Therefore, they allow light generated in the gain section to be transmitted to the waveguide sections 71, 72. As described in detail below, the light couplers 75, 76 may, for example, be provided in the waveguide sections 71, 72 and / or in the active gain section. That is, the couplers may be created by patterning: (i) the waveguide sections of the photonic circuit; (ii) the gain section (e.g., the III-V materials); or (iii) both the waveguide section and the gain section. In particular, the light couplers may be arranged on top of a surface of a layer continuous with the wafer or a surface of the wafer itself.Preferably, the couplers 75, 76 have a shape and arrangement such that they enable adiabatic coupling of the light between the active gain section and the waveguide section (adiabatic means that there is substantially no loss and substantially no reflection).

[0031] Finally, the unit 100 comprises a reflector 90 arranged to reflect light propagating in the same direction back to a center of the gain section. It is important that the unit 100 does not have any further reflectors on the other side of the gain section, i.e., there is no further reflector opposite the partial reflector 90 with respect to the active gain section that could reflect the light back to the center of the gain section. In other words, the partial reflector 90 is arranged (and configured) asymmetrically with respect to the center of the gain section, e.g., only on one side of the gain section. Nevertheless, the partial reflector 90 could actually be part of the active gain section, since its function is to reflect part of the generated light exclusively from one side back to the center of the gain section.

[0032] As is known to a person skilled in the art, a reflector never reflects 100% of the light (hence, the "partial reflector" will be referred to simply as "reflector"). It is precisely this property that is used to measure the optical gain, as will be described later. The present invention can, in principle, be implemented with any reflector that only partially reflects light. Preferably, the reflector 90 should reflect between 10 and 90% of the light that strikes it. A particularly preferred range is between 40 and 60% (e.g., a reflection of 50%), which in practice allows more suitable asymmetries of the light intensity in the two waveguides 71, 72 to be achieved for measurement purposes. Note that more than one reflector 90 may be present on either side of the center. Furthermore, in this invention, "light" must be understood as "electromagnetic radiation," i.e.It is not necessarily limited to visible light, despite some of the applications mentioned here. For example, it is common practice in this technical field to refer to it as "infrared light."

[0033] It is very convenient that the reflector 90 is arranged in the waveguide section 72, as shown in the Fig. 1 to 4. It could, in some variations, be located at one end of the gain section near coupler 76 to cross the optical path and reflect light back to the center of the gain section. Nevertheless, reflector 90 is preferably the only reflector that interacts with the light generated by the gain section in the optical path defined by the direction in which waveguide sections 71, 72 are aligned (apart from further couplers 81, 82 described below).

[0034] Such a device avoids the disadvantages of the prior art described in the Background section; it is well suited to representing the gain spectrum of on-chip devices such as laser devices or optical amplifiers. The description of related gain measurement methods makes it clear that the gain of an on-chip optical device can be easily measured directly on the wafer, without requiring the fabrication of multiple devices, the cleaving or polishing of facets, or further the need to attach multiple contacts to the gain measurement structure, or the accurate measurement of the gain under realistic bias conditions, i.e., above the threshold of a laser device. In contrast, the present methods essentially require sampling the optical intensity of the light emitted by the waveguide sections, from which the gain can then be determined.These procedures are described in detail below.

[0035] Various methods are known for coupling light. The light couplers 75, 76 are preferably longitudinal couplers, ie they extend longitudinally in the direction of an extension of the waveguide sections 71, 72, see, for example, the Fig. 6 to 8. This direction corresponds to the main direction of light propagation in the couplers and waveguide sections (light can propagate in either direction in this direction). The use of longitudinal optical couplers 75, 76 is preferred because they promote adiabatic coupling between the waveguide sections and the gain section.

[0036] How to get from the Fig. 1, 2, 4, and 6 to 8, the light couplers 75, 76 may each comprise, for example, at least one tapered section 752, 762, wherein a tapered section terminates a waveguide section 71, 72 or is connected to the active gain section. The tapered sections may be substantially parabolic, i.e., the lateral edges of the tapered sections are parabolic, and more generally, may be nonlinear. Appropriate design of the tapered sections allows for a smoother optical mode change, ensuring minimal scattering to the unwanted modes and a shorter length of the tapered section. Further investigation of this subject has shown that a parabolic shape is indeed not the optimal geometry. However, it can be considered an approximation to the optimal geometry and is at least a better approximation than linear tapers.

[0037] Nonlinear tapers can be achieved using a single nonlinear tapered section or multiple tapered sections, e.g., a linear section followed by a nonlinear, e.g., parabolic section, which in turn is followed by a linear section, etc. Preferred taper designs are based on the target coupling efficiency, the geometry and refractive indices of the waveguides, and the size constraints. Depending on the available fabrication techniques, it may be more practical to approximate a nonlinear tapered section by several consecutive linear subsections.

[0038] The conditions for adiabatic light coupling between tapers have been described in several publications. Analytical formulas describing tapers with optimal designs are cited in the literature. Optimal taper parameters (suitable for the present context) can be obtained from finite-difference time-domain (FDTD) simulations that solve Maxwell's equations in the time domain. This numerical method allows for the calculation of very accurate solutions for the taper parameters.

[0039] Preferably, the reflector 90 is arranged at one end of the tapered section of one of the waveguide sections, e.g., at the end of the tapered section 762 (in this case, it is integrated into the waveguide section 72), as shown in the accompanying drawings. In some variants, the reflector 90 may be located at one end of the tapered section 761 (in this case, it is integrated into the gain section at its edge). The closer the reflector is to the edge of the gain section, the fewer losses there are and the more accurate the measurements. However, from a manufacturing perspective, it is easier to manufacture the reflector in one of the waveguide sections because waveguide manufacturing processes are more sophisticated and can resolve smaller elements. Silicon, for example, is preferred. Therefore, providing the reflector at the end of the tapered section 762 provides a satisfactory compromise.

[0040] More generally, the reflector may be located at the beginning of a coupler, e.g. in the active amplification section, or at the end of a coupler, e.g. coupler 76 in the figures, regardless of the actual embodiment of the latter.

[0041] If we now refer to the Fig. 6 to 8, in some embodiments, the two light couplers 75, 76 may comprise two tapered sections 751, 752; 761, 762. Considering the light coupler 75, it may comprise two tapered sections 751, 752 that are aligned opposite each other and at least partially overlap. One of the tapered sections 752 terminates the waveguide section 71, while the other tapered section 751 is part of (or at least connected to) the active gain section 62 to 66 to efficiently optically couple the gain section to the waveguide sections.

[0042] As in the Fig. 1 to 8, the unit 100 may have additional light couplers 81, 82 directly integrated therein to enable gain measurement. In particular, an additional light coupler 81, 82 may be provided in each of the two waveguide sections 71, 72, so that the reflector may be located between one of the light couplers 76 and one of these additional couplers 82. Both additional light couplers 81, 82 are preferably located at one end of a waveguide section, i.e., opposite the end closest to the light couplers 75, 76. The additional light couplers 81, 82 may, for example, be grating couplers, which makes it possible to sample light vertically, i.e., to sample light emitted at an additional light coupler perpendicular to the surface comprising the waveguide sections. In some variants, the additional light couplers may be configured for lateral measurement of the emitted light.In other variants, the additional light couplers can be part of a scanning unit (external couplers that are not part of unit 100). The reader is referred to the publication "Flexible metal grating-based optical fiber probe for photonic integrated circuits" by Stijn Scheerlinck et al., Appl. Phys. Lett. 92, 031104 (2008), in which light is coupled between the probe and the waveguide without the need for integrated coupling structures.

[0043] As already mentioned, the photonic circuit is preferably a silicon photonic circuit. As described in the Fig. 2, Fig. As shown in Figure 3, the wafer may additionally include an electrical circuit 40 in addition to the photonic circuit. The electrical circuit 40 may, for example, be a complementary metal oxide semiconductor (CMOS) front end. More generally, the wafer may further include electronics.

[0044] How to be better in the Fig. 1, the two waveguide sections 71, 72 can extend directly on a dielectric layer 20. The dielectric layer can be provided on top of the wafer. This dielectric layer 20 can be referred to as a buried oxide, e.g., SiO2. It preferably has a thickness of more than 1 micrometer. The actual minimum thickness depends on the wavelength of the generated light: the light wavelength that is preferably used is, for example, 1.3 to 1.55 micrometers. The dielectric layer provides a lower cladding for the waveguide sections while providing a thermal and mechanical interface to the wafer. The dielectric layer 20 can thus be advantageously used to control the mechanical and thermal properties of the device.

[0045] If we now refer to the Fig. 4, 6 to 8, the waveguides 70 to 72 may be in contact with a bonding layer 50, which is usually made of a polymer, SiO2 or Al2O S (or any combination thereof). Again, a double layer of Al2O S and SiO2 serve as a boundary layer. The waveguides 70, 71, 72 can be partially buried in the bonding layer 50, as in Fig. 11, which makes it possible to determine the properties of the couplers 75, 76 by adjusting the thickness of the bonding layer 50. They can still have a surface that is planar with a surface of the bonding layer 50, as in Fig. 12, which reduces variations in the coupling properties of the couplers 75, 76 caused by variations in the thickness of the bonding layer 50. When the bonding layer 50 is flush with the surface of the photonic circuit, the latter determines the height or thickness of the bonding layer 50 very precisely. In other variants, the unit can be designed such that the waveguides 70, 71, 72 are partially (laterally) surrounded by air, as in Fig. 13. This allows for easier bonding of the reinforcement material to the wafer, as light can be passed through the gaps, reducing the risk of air bubbles and delamination of the bonded reinforcement material.

[0046] The gain material may be bonded on top of the bonding layer 50. In some variations, the active gain section may be disposed (directly or indirectly) on top of the waveguide sections by molecular bonding.

[0047] If we now refer to the Fig. 6 to 8 and 11 to 15, the gain section 62 to 66 should preferably comprise a III-V semiconductor material. In some variants, it may comprise a II-VI semiconductor material, germanium, semiconductor alloys such as silicon-germanium, or a polymer. The semiconductors used may be doped if necessary and preferably used with strain, e.g., to make the band gap direct rather than indirect (e.g., when germanium is used). Combinations of such materials may be further considered. It may further include a material having embedded quantum dots or quantum well material. The gain section 62 to 66 may further comprise, for example, a III-V semiconductor material and, in particular, comprises a lower contact layer 64 (with a first metal contact 65) and an upper part 62 on top of the lower contact layer 64 (with a second metal contact 66). The reasons for this will become clear later.

[0048] As in Fig. 15, the reinforcement section 62 to 66 may comprise a stack of epitaxial layers 67. The latter may comprise an n-doped semiconductor 642, a portion of which may (at least partially) form the lower contact 64. In particular, the layer 64 may be substantially formed by layer 642, by a residual portion of the layer 642, particularly if the latter is formed by patterning the stack of epitaxial layers 67 (e.g., by lithography and etching). The stack of epitaxial layers 67 further comprises an upper stack, which (at least partially) forms the upper part 62 of the reinforcement section made of III-V semiconductor material. Fig. 15 the upper stack shows: - a first intrinsic semiconductor 622 arranged on top of the n-doped semiconductor 642; - a multiple quantum well section 624 on top of the first intrinsic semiconductor; - a second intrinsic semiconductor 626 on top of the multiple quantum well section; and - a p-doped semiconductor 628 on top of the second intrinsic semiconductor.

[0049] The aforementioned layer stack can be easily grown, either by molecular beam epitaxy (MBE) or by metal-organic chemical vapor deposition (MOCVD). In particular, the gain stack has the advantage that the n-doped section is located close to the waveguide. This is particularly advantageous since the p-doped section typically has ten times higher optical loss for the same doping level or proportion of dopants contained in contact layers 628 and 642, respectively.

[0050] In some variants, the gain stack may also include a tunnel junction, which allows the device to be terminated with an n-type contact on both sides, so that only one type of contact metal, e.g. gold, tungsten, titanium, etc., needs to be used. It should be noted that normally p- and n-doped regions use different types of metal to match the Fermi levels and reduce the contact resistance.

[0051] A preferred example of implementation is the following (in the Fig. 6 to 9 and 14). First, a photonic circuit comprises two separate waveguides 71, 72 (each defining a waveguide section, as described above). Second, the light couplers 75, 76 couple light between the active gain section and the two waveguides, see Fig. 9. For this purpose, each of the two light couplers 75, 76 has two tapered sections that are aligned opposite each other and overlap, as described above. In particular, one of the tapered sections 752, 762 forms one end of a waveguide section, while the other tapered section 751, 761 is connected to the active gain section 62 to 66 (preferably forming part of it). The tapered sections 751, 761 widen toward the center of the active gain section, while the tapered sections 752, 762 (the tapers of the waveguides) narrow toward the center of the active gain section.

[0052] As described above, such a configuration has been shown to improve the adiabaticity of the coupling. Adiabaticity is achieved when the optical distribution is defined by the same eigenmode (i.e., the supermode of the coupled waveguide system, e.g., the fundamental even supermode, the fundamental odd supermode) within the taper with minimal scattering to other supermodes or radiation modes. Nevertheless, the loss is never exactly zero. Adiabaticity is a relative term, as is known to those skilled in the art; a coupler is considered adiabatic if the loss is below a predetermined value, e.g., below 15%, but usually below 10%.

[0053] The embodiments of the Fig. 6 to 9 include (a total of) four tapered sections. In some variants, only two tapered sections (a total) may be provided, which, for example, both form parts of the active amplification section. In other variants, only two tapered sections are provided, both of which are located in a respective waveguide section.

[0054] The waveguide sections are not necessarily defined by corresponding well-defined waveguides. Fig. 3, for example, only one waveguide 70 is provided, which defines the two waveguide sections 71, 72 (note that Fig. 3 and more generally the Fig. 1 to 5 show units in intermediate stages of manufacture). The waveguide 70 may have a different cross-section (not shown). The latter may, for example, have a central section with a reduced width compared to the outer sections 71, 72, which defines the two waveguide sections 71, 72.

[0055] Currently, methods for measuring optical gain are being discussed using units such as those described above. With particular reference to the Fig. 9, Fig. 10 These procedures generally include the following steps: - first, at S10, a unit 100 such as that described here is provided; - Second, at S20, the amplifying section 62 to 66 is excited to generate light therein at S30. According to the description of the unit described above, the amplifying section can be excited by electrical pumping or optical pumping. ◯ The gain section can be excited electrically (electrical pumping) by using electrical probes, wire bonding, flip-chip bonding or soldering, etc. ◯ The amplification section can also be excited optically (optical pumping) by a light pulse or by CW light. ◯ As in Fig. 9, an electron-hole pair (e.g. in the stack of epitaxial layers 67, in particular in the quantum well 624 of Fig. 15) recombine in the amplifying section to produce a photon; - then, at S40, the generated light is transmitted to each of the two waveguide sections 71, 72 through the light couplers 75, 76; the light is (at least partially) reflected at S45 by the reflector 90, resulting in an asymmetry between the optical power in the first waveguide section 71 compared to the second waveguide section 72; and - finally, at S60, the optical power of the light emitted by each of the two waveguide sections is sampled in order to determine at S70 the optical gain of the active amplification section 62 to 66. ◯ In particular, both the optical power of the light emitted from the first waveguide section 71 and the optical power of the light emitted from the second waveguide section 72 are sampled. ◯ The optical power is sampled so that: ▪ Light emitted from the first waveguide section includes light that has been transmitted both directly and indirectly thereto, in particular: • Light generated at S40 and then transmitted directly to the first waveguide section 71 at S40; and • Light generated at S40 and then reflected at S45 by the reflector 90 and then transmitted to the first waveguide section at S40, whereas ▪ the optical power of the light emitted by the second waveguide section 72 is sampled "after" the reflector 90, meaning that the light emitted by the second waveguide section 72 includes the light not reflected by the reflector 90, i.e., the portion of the light that has passed through the reflector. In other words, the optical power of the light sampled at the second section 72 includes only the light not reflected by the reflector 90 (taking into account negligible losses and the effects of stray light from the environment of the second section 72).

[0056] In this regard, it is understood that residual losses may occur, particularly when the light is transmitted to and propagated within the waveguide sections, is reflected at the reflector, and / or is emitted from the additional couplers.

[0057] Because of the asymmetric reflector 90, the light intensity in the left waveguide section 71 is Fig. 9 is greater than the luminous intensity in the right section 72, as indicated by the different widths of the curved arrows. The measured gain essentially reflects an asymmetry between these luminous intensities. The gain is not measured directly but calculated. Typically, the optical powers are sampled to measure the luminous intensities (and thus the emission rates) to ultimately obtain the optical gain. More details are given in the next section.

[0058] In preferred embodiments, the two waveguide sections 71, 72 comprise an additional light coupler 81, 82 (e.g., located at opposite ends of the waveguide sections). The optical power of the emitted light can thus be sampled at step S60 at the additional light couplers 81, 82 (and in particular behind the reflector 90 for the second section 72). Step S60 can be carried out, in particular, via optical waveguides 201, 202 or photodetectors arranged near the additional light couplers 81, 82. The unit 100 can comprise such sampling means, i.e., optical waveguides 201, 202 or photodetectors.

[0059] The above embodiments have been briefly described with reference to the accompanying drawings and may include numerous variations. Several combinations of the above elements may be considered. Examples are given in the next section. 2. Specific embodiments / technical details of implementation

[0060] Specific embodiments and their variants will now be described. These embodiments relate to a unit similar to a test structure for on-chip optical gain measurement, the main advantages of which are that the gain can be measured as follows: - using electric pumps; - by arranging only two contacts at the same time; - by using vertical scanning with fibers (which enables wafer-level testing) - without polishing the facets; - without manufacturing and measuring a large number of test structures (one unit is sufficient); - without having to use many (>2) contacts; and - without the need for high spectral resolution.

[0061] An integration plan is disclosed wherein an optically active gain section made of germanium, GaAs, InP, InGaAs, InAlAs, InGaAsP, NAsP, GaSb, any of their alloys, or any other suitable compound semiconductor is evanescently coupled to a waveguide or forms a hybrid mode with a waveguide, preferably made of silicon. This is achieved by bonding the active optical layer (i.e., the gain section) directly to the silicon waveguide or on top of the bonding layer 50 made of a low-refractive-index material, preferably silicon dioxide, silicon nitride, aluminum oxide, or a combination thereof. Layer 50 may further be made of a bonding material such as a polymer, or any other suitable solid-state material, or a combination thereof.

[0062] The preferred integration plan is based on molecular bonding. A III-V-based material is grown on a suitable substrate (III-V semiconductor, Si, Ge, etc.) and is optionally covered with a dielectric by molecular beam epitaxy, molecular vapor phase epitaxy, metal-organic chemical vapor deposition, atomic layer epitaxy, atomic layer deposition, sputtering, or any suitable thin-film deposition technique. This layer is then bonded on top of the electronics wafer containing the front-end electronics and optics. Either a full III-V layer stack is bonded, serving as the gain material, or a seed layer is bonded for subsequent regrowth, or a III-V layer stack is bonded, comprising both gain material and a seed layer with suitable etch stops.

[0063] Bonding is preferably performed on top of a dielectric layer 20 located on top of the CMOS wafer 10. In prior art CMOS processes, this layer is a silicon dioxide layer polished by chemical mechanical polishing (CMP) to provide a flat surface having low surface roughness.

[0064] Because it uses either wafer or wafer-level bonded III-V-based layers, the integration plan is suitable for mass production and easy integration with modern back-end manufacturing processes.

[0065] The gain test structure in the III-V layer is electrically connected during these back-end manufacturing plans. This allows the gain test structure to be fabricated for wafer-level testing.

[0066] Furthermore, this test structure can be controlled by the electronics and coupled to the optics arranged in a common layer, i.e., the upper silicon layer. The optical signal is coupled out by grating couplers 81, 82, which allow vertical scanning and thus testing at the wafer level.

[0067] An example of such a unit is in Fig. 8 shown.

[0068] Variants of this unit and their manufacturing processes are now described.

[0069] The structures may be designed such that an active material (e.g., a III-V based material) 62 to 66 is bonded on top of a wafer 10 having a photonic silicon circuit, as shown in Fig. 1 (which represents an early intermediate stage in the fabrication of one embodiment of the present invention). The silicon photonic circuit includes grating couplers 81, 82 on either side of the waveguide sections 71, 72, a reflector 90 of reflectivity R, and tapered sections 751, 761. Further, the silicon photonic circuit is supported on an oxide layer 20 (a buried oxide), which in turn is supported on the silicon wafer 10, thereby forming a silicon-on-insulator (SOI) structure.

[0070] As noted in the previous section, embodiments of the invention comprise a substrate comprising a silicon photonic circuit as above and additionally a CMOS or bipolar (Bi-)CMOS front-end-of-line (FEOL) 20 as in Fig. 2. This figure shows the stack of the preferred embodiment, which includes the electronics 40, passive photonics, the wafer 10, and the buried oxide layer 20; it shows a host substrate containing electronics 40 and photonics components, including additional coupling means 81, 82 (e.g., grating couplers), tapered waveguide sections 71, 72, and the reflector 90. The transistors and also the photonics are fully processed, ie, including silicide, implantation, and annealing steps.

[0071] In the variant of Fig. 3, the substrate is essentially the same as the embodiment in Fig. 2, except that the waveguide 70 between the grating couplers 81, 82 is unstructured, without the two adiabatic coupling sections 752, 762.

[0072] After fabricating the silicon photonics 70 to 72 and the front-end-of-line 40, the unit is further processed. The next manufacturing step is the deposition of a layer 50 suitable for bonding. The layer can be made of a polymer (for adhesive bonding) or, in preferred embodiments, a silicon dioxide layer, as shown in Fig. 4. Nevertheless, layer 50 may comprise aluminum oxide, hafnium dioxide, tantalum pentoxide, barium titanate, or strontium titanate, or even silicon nitride or silicon oxynitride. This layer also serves as an additional (lower) cladding layer.

[0073] In a preferred embodiment, this layer 50 resembles the first intermediate layer (the so-called ILD1) between the electronic FEOL and the back-end-of-line (BEOL). Furthermore, the silicon dioxide layer 50 preferably has a thickness between 10 and 2,000 nm and a root-mean-square surface roughness of less than 0.5 nm. This surface roughness can be achieved using a proprietary deposition process.

[0074] Preferably, the silicon dioxide layer 50 is deposited by plasma enhanced chemical vapor deposition (PECVD) and is subjected to subsequent chemical mechanical polishing (CMP).

[0075] After the oxide is deposited and planarized, the III-V layer 60 is bonded to the top of the wafer, where layer 60 is further processed. The III-V wafer 60 includes the active III-V quantum well (or QW) layer or the seed layer, or both. After bonding the III-V layer, the III-V substrate is removed by either grinding, wet-chemical etching, or smart-cutting. In a preferred embodiment, the III-V material is removed using a wet-chemical etchant containing HCl and water. Fig. Figure 5 shows this manufacturing stage with the bonded III-V layer 60 on top. As mentioned in a previous section, the III-V layer can be a complete layer stack comprising a QW 624 ( Fig. 15) and electrical contacts that serve as laser material. In some variants, the III-V layer is a seed layer for the subsequent growth of III-V material on top of the wafer. Note that the temperature range (450 to 650 °C) is sufficient to perform the re-growth on a wafer containing electronics.

[0076] Additionally, bonding can be performed on top of the ILD1 dielectric, which has tungsten contacts (not shown). Since the tungsten contacts are small compared to the total wafer area, bondability should not be compromised. Tungsten is the preferred material for the contacts because they can then serve as alignment markers for an electron beam (e-beam) process, as they have a high ion mass and therefore provide good contrast when using an electron beam imaging process.

[0077] Bonding of the III-V layer is carried out at elevated temperatures in a vacuum furnace or in a suitable gas environment to form a molecular bond. The temperature is preferably between 200 and 600 °C. However, better results can generally be achieved when the temperature is between 350 and 450 °C. A suitable pressure is usually below 80 kPa.

[0078] Note that the stack of epitaxial layers 67 ( Fig. 15) for emission is preferably fabricated to be suitable for electrical pumping to achieve the desired optical emission. In some variants, it is fabricated to be suitable for optical pumping by directing a high-intensity optical pulse onto the active gain section 62 by free-space propagation. The stack of epitaxial layers 67 (or simple variants thereof) can be used in the units described in the Fig. 6 to 9. In this respect, the Fig. 8 to 9 schematically show units designed mainly for electric pumping, whereas the unit used in Fig. 7 is used for optical pumping. Nevertheless, the stack of Fig. 15 can be used for electrical or optical pumping, although optical pumping does not necessarily require the doped regions 628 and 642 (which are only needed for electrical pumping). Thus, it is understandable that the units of the Fig. 8, Fig. 9 can also be used for optical pumping, although they are not optimal for this purpose (the Fig. 8 and Fig. 9 have metal contacts that shield the pump light from the top). Finally, the unit can be Fig. 7 as an intermediate stage for the unity of Fig. 8, Fig. 9 (electrical pumping) or as a finished structure for optical pumping.

[0079] In the next manufacturing step, the III-V layer 60 (which comprises the complete stack of epitaxial layers 67, which is formed by bonding or bonding and subsequent regrowth) is patterned. In particular, the III-V layer 60 can be patterned by either wet chemical or dry chemical treatment. In preferred embodiments, the III-V layer is formed using an inductively coupled plasma (ICP) reactive ion etching (RIE) system. The fabrication of the III-V gain section depends on the desired alignment for light propagation; the gain section is aligned with the silicon waveguides or the adiabatic tapered sections. This can be achieved thanks to alignment marks. The alignment marks can, for example, be made of silicon in the upper silicon layer during patterning of the FEOL.In some variants, the alignment marks are made of tungsten and are created during the processing of the transistor vias.

[0080] In some embodiments, the structure 100 can be fabricated to be suitable for optical pumping. An example of such a device is shown in the Fig. 6A to 6B. The III-V gain section extends over the layer 50 in which the waveguide sections are buried; it still has adiabatic coupling sections 75, 76 at both ends. Thus, it is suitable for optical pumping. It can be designed to have low reflectivity and coupling loss to the waveguide sections 71, 72 (preferably silicon waveguides). Of course, the unit used in the Fig. 6A to 6B is not necessarily "final" in that additional components (e.g., dielectric layers and contacts) may need to be arranged on the same wafer chip.

[0081] In the embodiments of the Fig. 7 to 9, the III-V section 62 to 66 is structured (as above) with respect to the silicon waveguides 71, 72 or the adiabatic coupling sections 75, 76 and additionally comprises a lower contact layer 64. The latter is preferably made of a heavily n-doped indium phosphide (InP), since this material has an optical loss an order of magnitude lower than that of p-doped InP (e.g., 10 times lower). However, it may also comprise indium gallium arsenide (InGaAs) or indium aluminum arsenide (InAlAs).

[0082] As mentioned above, Fig. 7 can be viewed as an intermediate step in the fabrication of a device suitable for electrical pumping, in which the III-V gain layer is patterned. In the figure shown, the III-V gain section 62 to 64 has adiabatic coupling sections 75, 76 at both ends. Such a device forms a basis for a device then suitable for electrical pumping; the lower contact section 64 will be metallized later. Again, the shown design of the gain section has low reflectivity and coupling loss to the silicon waveguides 71, 72.

[0083] After the patterning of the III-V gain layers is completed, metal contacts are arranged as shown in Fig. 8. The purpose of contacts 65, 66 is to enable electrical pumping of the gain section. The electrical contacts 65, 66 can be made of tungsten, titanium, titanium nitride, cobalt silicide, nickel silicide, polysilicon, gold, titanium, nickel, platinum, aluminum, copper, or a combination thereof. The contacts are preferably made of tungsten for the reasons stated above.

[0084] Additional variants are now described.

[0085] Instead of bonding a III-V material on top of the electronics front-end 40, germanium can also be considered, as it also exhibits optical gain (or is transparent in the mid-infrared range). A germanium-on-insulator (GOI) or germanium-on-silicon wafer (via a SiGe buffer) can be bonded on top of the ILD1, analogous to the fabrication process described above. Furthermore, analogous to the removal of the substrate in Fig. 2 The wafer can be removed either by chemical-mechanical polishing, grinding, dry etching, wet chemical treatment, or smart cutting. The silicon dioxide layer (which essentially consists of the buried oxide (BOX) of the wafer) can be removed either dry chemically or wet chemically. Alternatively, the silicon dioxide layer can remain on the germanium, serving as a dielectric or hard mask for subsequent metallization steps.

[0086] Unlike in the drawings shown, the grating couplers can also be arranged on one side of the waveguides so that they are in a row and have a specific spacing (e.g., 250 micrometers). In this way, commercially available fiber arrays can be used for measurements instead of individual fibers. Furthermore, only a single fiber alignment is required instead of two, which significantly speeds up the measurement process. Then, a total of two electrical contacts (which can be set up with a single electrical probe, see step S20, Fig. 9) and a fiber array is required to measure the gain of, for example, a laser structure on the chip 100.

[0087] The wavelength-dependent gain in the present invention can be easily calculated by a single measurement on a single unit with a given length L and a reflector with reflectance R from the equation given in the publication by Ma et al. in Optics Express Vol. 21 No. 8, pp. 10335 to 10341 (2013). G(λ)=1Lln(1−R)P1(λ)−P2(λ)R P2(λ) where P1 and P2 are the optical power levels measured at the additional couplers 81 and 82, respectively, and λ is the actual wavelength.

[0088] The methods and devices described above can be used in the manufacture of photonic chips. The photonic chips thus manufactured can be distributed by the manufacturer in unprocessed wafer form (i.e., as a single wafer containing multiple unpackaged chips), as a bare chip, or in packaged form. In the latter case, the chip is mounted in a single-chip package (such as a plastic carrier with leads attached to a printed circuit board or other superordinate carrier) or in a multi-chip package (such as a ceramic carrier containing surface-mounted and / or buried leads). In either case, the chip can then be integrated with other chips, discrete circuit elements, and / or other signal processing units as part of either (a) an intermediate product such as a printed circuit board or (b) a final product. The final product can be any product that incorporates photonic chips, e.g.Transceivers, integrated lasers, optical amplifiers, transmitters, receivers, active optical cables, etc.

[0089] While the present invention has been described with reference to a limited number of embodiments, variations, and the accompanying drawings, it will be apparent to those skilled in the art that various changes may be made and equivalents may be substituted without departing from the scope of the present invention. In particular, an element (in the form of a unit or method) specified or shown in a given embodiment or variation in a drawing may be combined with or replaced by another element in another embodiment, variation, or drawing without departing from the scope of the present invention. Various combinations of the elements described with reference to all of the above embodiments or variations are thus contemplated and remain within the scope of the appended claims.In addition, many minor modifications may be made to adapt particular circumstances or materials to the teachings of the present invention without departing from its scope. Therefore, the present invention is not intended to be limited to the particular embodiments disclosed, but the present invention is intended to include all embodiments falling within the scope of the appended claims. In addition, many other variations than those expressly described above are contemplated. Materials other than those expressly described (Si, SiO2, Al2O3, polymers, etc.) may be contemplated. Furthermore, other designs for the active gain section, the shape of the waveguides, the shape of the tapers, etc. may be contemplated. REFERENCE LIST 10 wafers 20 Dielectric layer (buried oxide) 40 Electrical circuit (CMOS front end) 50 bond layer 62 - 66 Active amplification section 62 Upper part of the reinforcement section 64 Lower contact layer 65 First metal contact 66 Second metal contact 67 stacks of epitaxial layers 70 Single waveguide 70, 71, 72 Photonic circuit (section) 71, 72 Waveguides (waveguide sections) 75, 76 Light couplers (opposite tapers) 81, 82 Additional (grating) light couplers 90 reflector 100 units for on-chip measurement of optical gain 201, 202 Scanning means (optical fibers or photodetectors) 622 First intrinsic semiconductor 624 Multiple quantum well section 626 Second intrinsic semiconductor 628 P-doped semiconductor 640 N-doped semiconductor 642 First intrinsic semiconductor 642 N-doped semiconductor 644 Multiple quantum well section 646 Second intrinsic semiconductor 648 P-doped semiconductor 650 stacks of epitaxial layers 751, 752 Tapered sections of the first light coupler 75 761, 762 Tapered sections of the second light coupler 76

Claims

[1] A photonic circuit unit (100) for measuring an optical gain, comprising: - a substrate (10) with a photonic circuit, the latter comprising two waveguides (71, 72) defining two waveguide sections aligned in the same direction and facing each other, - an active amplification section (62 to 66) on top of the substrate (10) and coupled in the photonic circuit unit (100) to generate light by electrical pumping or optical pumping, - at least two light couplers (75, 76) arranged such that at least a part of the active gain section (62-66) lies between the light couplers (75, 76) and configured such that light is coupled between the active gain section and the waveguide sections, - a partial reflector (90) arranged in one of the waveguide sections so as to reflect light propagating along the same direction back to a center of the gain section (62-66), and wherein the photonic circuit unit (100) has no further reflector opposite the partial reflector (90) with respect to the active gain section (62-66) and is configured so as to reflect light back to the center of the gain section. [2] The photonic circuit unit (100) according to claim 1, wherein the light couplers (75, 76) extend longitudinally along the same direction. [3] The photonic circuit unit (100) according to claim 1 or 2, wherein the light couplers (75, 76) each have at least one tapered section (752; 762), wherein the tapered section (752; 762) terminates a waveguide section or is connected to the active gain section (62-6, 60), and wherein preferably the tapered sections (752; 762) are each substantially parabolic. [4] The photonic circuit unit (100) according to claim 3, wherein the light couplers (75, 76) each have two tapered sections (751, 752; 761, 762) oriented opposite to each other and overlapping, wherein a first of the two tapered sections (751, 752; 761, 762) terminates an associated waveguide section and a second of the two tapered sections (751, 752; 761, 762) is connected to the active gain section (62-66). [5] The photonic circuit unit (100) according to any one of the preceding claims, wherein each of the waveguide sections has an additional light coupler (81, 82), wherein the partial reflector (90) is located between one of the light couplers (75, 76) and such additional light coupler (82), wherein each additional light coupler (81, 82) is preferably arranged at one end of each of the waveguide sections that is opposite the other end of each of the waveguide sections that is closest to the light couplers (75, 76). [6] The photonic circuit unit (100) of claim 5, wherein the additional light couplers (81, 82) are grating couplers configured to enable vertical measurement of the light emitted at the additional light couplers (81, 82). [7] The photonic circuit unit (100) according to any one of the preceding claims, wherein the photonic circuit is a silicon photonic circuit, and preferably wherein the substrate further comprises an electrical circuit (40) in addition to the photonic circuit. [8] The photonic circuit unit (100) according to any one of the preceding claims, wherein the waveguide sections each extend directly on a dielectric layer (20) of the photonic circuit unit (100), wherein preferably the dielectric layer (20) lies on top of the substrate (10) and has a thickness of more than 1 micrometer. [9] The photonic circuit unit (100) of claim 8, wherein each of the two waveguides (71, 72) is in contact with a bonding layer (50) and preferably is partially embedded in the bonding layer (50) and more preferably has a surface that is planar with a surface of the bonding layer (50), wherein the bonding layer (50) is preferably one of: a polymer; SiO2; or Al2O3 or any combination thereof. [10] The photonic circuit unit (100) according to any one of the preceding claims, wherein the gain section (62 to 66) comprises one or more of: a III-V semiconductor material, a II-VI semiconductor material, a semiconductor such as germanium, semiconductor alloys such as silicon-germanium, a polymer, a material having embedded quantum dots and / or quantum wells or a quantum well material, and wherein preferably the gain section (62 to 66) is a gain section made of III-V semiconductor material, comprising: - a lower contact layer (64) having a first metal contact (65); and - having an upper part (62) on top of the lower contact layer (64) and a second metal contact (66). [11] The photonic circuit unit (100) of claim 10, wherein the gain section (62 to 66) comprises a stack of epitaxial layers (67), the latter comprising an n-doped semiconductor (642), at least a portion of which forms at least a portion of the lower contact (64), and wherein the stack of epitaxial layers (67) further comprises an upper stack forming at least a portion of the upper portion (62) of the gain section (62-66) of III-V semiconductor material, and which in turn comprises: - a first intrinsic semiconductor (622) on top of the n-doped semiconductor (642); - a multiple quantum well section (624) on top of the first intrinsic semiconductor (622); - a second intrinsic semiconductor (626) on top of the multiple quantum well section (624); and - a p-doped semiconductor (628) on top of the second intrinsic semiconductor (626). [12] The photonic circuit unit (100) according to any one of the preceding claims, wherein: - the light couplers (75, 76) are configured to enable light coupling between, on the one hand, the active amplification section (62-66) and, on the other hand, each of the two waveguides (71, 72), each of the two light couplers (75, 76) having two tapered sections (751, 752; 761, 762) oriented opposite one another and overlapping, - a first (752, 762) of the two tapered sections (751, 752; 761, 762) forms an end of one of the waveguide sections; and - a second (751, 761) of the two tapered sections (751, 752; 761, 762) is connected to the active reinforcement section (62 to 66) and preferably forms part thereof. [13] A method for measuring optical gain, comprising: Providing (S10) a photonic circuit unit (100) according to one of claims 1 to 12, wherein the two waveguide sections represent a first waveguide section and a second waveguide section, wherein the partial reflector (90) is arranged in the second waveguide section; Exciting (S20) the amplifying section (62 to 66) to generate light therein by electrical pumping or optical pumping (S30); transmitting the generated light (S40) to each of the two waveguide sections with the aid of the light couplers (75, 76), wherein the light is partially reflected at the partial reflector (90) (S45); and Sampling (S60) both the optical power of the light emitted by the first waveguide section and the optical power of the light emitted by the second waveguide section (72) to determine (S70) the optical gain of the active gain section (62 to 66), wherein the sampling (S60) is carried out such that: the light emitted by the first waveguide section has both Light that is generated (S40) and transmitted directly to the first waveguide section (S40), as well as Light that is generated (S40) and reflected (S45) at the partial reflector (90) and then transmitted to the first waveguide section (S40), and wherein the light emitted from the second waveguide portion (72) comprises light that is not reflected by the partial reflector (90) (S45). [14] The method for measuring optical gain according to claim 13, wherein: - each of the waveguide sections of the provided photonic circuit unit (100) has an additional light coupler (81, 82), wherein the partial reflector (90) is arranged between one of the light couplers (76) and such an additional light coupler (82), wherein each of the additional light couplers (81, 82) is preferably arranged at one end of each of the waveguide sections that is opposite the other end of each of the waveguide sections that is closest to the light couplers (75, 76); and - the optical power of emitted light is sampled (S60) at each of the additional light couplers (81, 82) to determine (S70) the optical gain of the active amplification section (62 to 66). The sampling (S60) of the optical power is preferably carried out via optical fibers (201, 202) or photodetectors located near each of the additional light couplers (81, 82).

Images