Loss measurement system in an integrated photonic circuit and associated method

Abstract

Loss measurement system in an integrated photonic circuit and associated method. The invention relates to a system (100) for measuring the absolute insertion loss ILcp ​​of a photonic circuit (20) comprising input (21) and output (22) interfaces, the system comprising a first optical fiber (1) configured to transmit the beam emitted by a light source (10) to a first connector, an injection optical fiber (2) connected to the first fiber using the first connector and configured to transmit the beam to the input interface, a collection optical fiber (3) configured to collect the beam from the output interface, a second optical fiber (4) connected to the collection fiber using a second connector and configured to transmit the beam to a detector for measuring a characterization power Pcp, the system comprising a calibration element (110) having an insertion loss ILmir,The system is configured to measure ILcp ​​insertion losses after calibration via the calibration element. Figure for the abbreviation: Fig. 1A,

Description

Title of the invention: Loss measurement system in an integrated photonic circuit and associated method technical field

[0001] The field of the invention is that of integrated on-chip photonics. The invention relates more particularly to the optical coupling of a photonic chip with an external device, for example with an optical fiber or a set of optical fibers. STATE OF THE ART

[0002] Integrated optical components, particularly silicon photonics components, can combine many functions on a single chip, also called a PIC, from the English acronym for "Photonic Integrated Circuit". In these components, light is conducted in small optical guides, typically less than a micrometer wide, between densely distributed functional blocks within the PIC.

[0003] In silicon photonics, light transported in the plane of the PIC can be directed out of the plane by dedicated functional elements, such as diffraction grating optical couplers. These couplers can be used as inputs to couple light from an optical fiber to a waveguide of the photonic circuit, or as outputs to couple light from a waveguide to an optical fiber. The diffractive grating optical coupler is integrated, along with the waveguides, on the same substrate.

[0004] For the development and deployment of these optical couplers, it is important to accurately characterize the optical insertion losses associated with these couplers. Currently, the insertion losses of diffraction grating couplers can approach a value of approximately 1 dB for the most efficient components. Approximately 80% of the light is therefore transferred between the optical fiber and the waveguide of the integrated photonic circuit. To accurately characterize the components, it is important that the uncertainty in the measurement of the insertion losses be much lower than the measured value. However, a measurement accuracy of less than 0.2 dB is currently difficult to achieve. This is because the insertion loss measurement systems of an integrated photonic circuit include optical fibers that are generally connected to each other by fiber connectors.These allow for easy connection and disconnection of the fibers, but introduce uncertainty in the measured values.

[0005] One object of the present invention is therefore to improve the measurement accuracy of insertion losses in an integrated photonic circuit. In particular, the invention provides a solution for measuring absolute insertion losses in an integrated photonic circuit independently of the losses produced by fiber connectors.

[0006] The other objects, features and advantages of the present invention will become apparent from an examination of the following description and accompanying drawings. It is understood that other advantages may be incorporated. SUMMARY

[0007] To achieve this objective, according to a first aspect of the invention, a system for measuring the absolute insertion loss (ILcp) of a photonic circuit integrated on a substrate is provided, said photonic circuit comprising at least one waveguide, an input interface and an output interface coupled to the at least one waveguide, the input interface being configured to receive a light beam and direct it towards the at least one waveguide and the output interface being configured to receive the light beam transmitted by the at least one waveguide and direct it out of the photonic circuit, the system comprising: • a light source emitting the light beam, • a first optical fiber coupled to the light source, configured to transmit the light beam to a first fiber connector, • an injection optical fiber connected to the first optical fiber using the first fiber connector, said injection optical fiber being configured to transmit the light beam to the input interface of the photonic circuit, the first fiber connector having a first insertion loss ILCi2, • a collection optical fiber configured to collect the light beam from the output interface of the photonic circuit, • a second optical fiber connected to the collection optical fiber using a second fiber connector, the second fiber being configured to transmit the light beam collected by the collection optical fiber to a detector, the second fiber connector having a second insertion loss ILC34, • the detector configured to measure the optical power (Pcp) of the light beam transmitted by the second optical fiber, the system further comprising a calibration element having a fixed and known insertion loss ILmir, said calibration element allowing the first and second insertion losses ILC12, ILC34 to be eliminated, so that the system is configured to measure the absolute insertion losses ILcp ​​of the photonic circuit, after calibration of the system via the calibration element.

[0008] The system described above allows for the precise measurement of the absolute insertion losses ILcp ​​of the integrated photonic circuit, without resorting to approximations or comparisons with external standards. Measuring absolute losses is particularly advantageous when the input and output interfaces of the photonic circuit are grating couplers commonly used in integrated photonic circuits, and whose optimization requires precise insertion loss measurements with low uncertainties. The calibration element, whose insertion loss ILmir is advantageously fixed and known with good accuracy, allows for reliable calibration of the measurement system. This makes it possible to measure the absolute insertion losses ILcp ​​of the photonic circuit and grating couplers, eliminating the ILcn and ILC34 losses associated with fiber connectors, which can be a source of uncertainty.

[0009] According to a second aspect of the invention, a method for measuring the absolute insertion losses ILcp ​​of the photonic circuit is provided using the system according to the invention, said method comprising, during a calibration step: • transmission via the first optical fiber, of the light beam emitted by the light source, to the injection optical fiber via the first fiber connector, • transmission via the optical fiber from the injection of the light beam to the calibration element, • a collection via optical fiber of the light beam from the calibration element, to the second optical fiber, passing through the second fiber connector, • transmission, via the second optical fiber, of the light beam collected by the primary optical fiber to the detector. • a first measurement by the calibration optical power detector Prm of the light beam transmitted by the second optical fiber, said first measurement including the first and second insertion losses ILcn, ILC34, the process further includes, during a measurement step: • transmission via the first optical fiber, of the light beam emitted by the light source, to the injection optical fiber via the first fiber connector, • transmission via optical fiber from the injection of the light beam to the input interface of the photonic circuit, • a collection via optical fiber of the light beam originating from the output interface of the photonic circuit, to the second optical fiber, passing through the second fiber connector, • transmission, via the second optical fiber, of the light beam collected by the primary optical fiber to the detector. • a second measurement by the optical power detector characterizing Pcp of the light beam transmitted by the second optical fiber, the second measurement allowing to determine the absolute insertion losses ILcp ​​of the photonic circuit, by comparison with the first measurement.

[0010] This method for measuring the absolute insertion losses (ILcp) of the photonic circuit eliminates the insertion losses associated with fiber connectors. The measurement circuit, essentially comprising the optical fibers, remains unchanged from the system calibration stage using the calibration element to the optical power (Pcp) characterization measurement stage. This allows for the precise and absolute measurement of the insertion losses of the photonic circuit without the need to connect / disconnect the optical fibers from the fiber connectors. BRIEF DESCRIPTION OF THE FIGURES

[0011] The aims, objects, features and advantages of the invention will be apparent better than a detailed description of a method of implementation of the latter, which is illustrated by the following accompanying drawings in which:

[0012] [Fig.1A] [Fig.1B] Figures IA and IB schematically illustrate a system for measuring the absolute insertion losses of a photonic circuit according to different examples of embodiments of the present invention.

[0013] [Fig.2] Fig.2 schematically illustrates a system for measuring the insertion losses of a photonic circuit according to a conventional approach.

[0014] [Fig.3] Fig.3 schematically illustrates a system for measuring the absolute insertion losses of a photonic circuit in a calibration configuration according to an example embodiment.

[0015] [Fig.4] Fig.4 illustrates a graph representing the relative positions of the different components of the absolute insertion loss measurement system of a photonic circuit according to an example embodiment.

[0016] [Fig.5A] [Fig.5B] Figures 5A and 5B schematically illustrate a system for measuring the absolute insertion losses of a photonic circuit according to a first embodiment.

[0017] [Fig.6A] [Fig.6B] Figures 6A and 5B schematically illustrate a system for measuring the absolute insertion losses of a photonic circuit according to a second embodiment.

[0018] [Fig.7A] [Fig.7B] Figures 7A and 7B schematically illustrate a system for measuring the absolute insertion losses of a photonic circuit according to a third embodiment.

[0019] The drawings are given by way of example and are not limiting of the invention. They constitute schematic representations of principle intended to facilitate understanding of the invention and are not necessarily to scale with practical applications. DETAILED DESCRIPTION

[0020] Before proceeding with a detailed review of embodiments of the invention, optional features that may be used in combination or alternatively are listed below:

[0021] According to one example, the system is such that: • The calibration element is configured to receive the light beam transmitted by the injection optical fiber, and reflect the light beam back to the collection optical fiber. • The detector is configured to measure a calibration optical power Prm of the light beam received and reflected by the calibration element.

[0022] The calibration element allows, during system calibration, the transmission by reflection of the light beam from the injection optical fiber to the detection optical fiber, without the need to disconnect the fibers from the connectors. This transmission is achieved by one or more reflections at the calibration element.

[0023] The optical axis is defined as the path of the central ray of the light beam from the injection fiber and its evolution after each optical surface encountered, up to the detection optical fiber.

[0024] The normal axis of a mirror is called the axis perpendicular to the surface of the mirror at the point of reflection of the optical axis.

[0025] In one example, the calibration element comprises at least one mirror having an ellipsoidal curved surface. In another example, a direction of propagation of the light beam and a normal axis of at least one mirror lie in the same plane, which is not parallel to a principal extension plane of the substrate.

[0026] The curved ellipsoidal surface of the mirror allows the light beam from the injection optical fiber to be reflected towards the collection optical fiber without significant loss. Indeed, the ellipsoidal shape of the mirror allows it to focus and direct the light beam accurately during beam transmission between the two fibers while minimizing losses.

[0027] According to one example, the root mean square deviation Aô between the curved ellipsoidal surface and an ideal ellipsoidal surface is less than or equal to 30 nm, preferably less than or equal to 20 nm.

[0028] The deviation from the ideal ellipsoidal shape limits the transmission rate of the light beam in the optical collection fiber after reflection on the curved surface. Current manufacturing technologies make it possible to obtain a root mean square deviation Aô from the ideal ellipsoidal shape of less than 30 nm, resulting in an insertion loss ILsurf = 0.36 dB, i.e., an optical loss rate of less than 8%.

[0029] According to one example, the curved surface of at least one mirror is covered with an aluminum-based metallic layer.

[0030] According to one example, the insertion loss ILmir of the calibration element is less than or equal to 0.85 dB.

[0031] These ILmir insertion loss values ​​of the calibration element are fixed and are determined with an accuracy of the order of ± 0.1 dB, which is less than the reproducibility of the connection between two optical fiber connectors.

[0032] According to one example, the calibration element comprises a single mirror configured so that the light beam transmitted by the injection optical fiber is directly reflected on said mirror towards the collection optical fiber.

[0033] This first embodiment of the measurement system is adapted to a common cross configuration of the injection and collection optical fibers used in photonic circuit insertion loss measurement systems. The use of a mirror with an ellipsoidal surface in this cross configuration allows the transmission of the light beam from the injection optical fiber to the detection optical fiber in a single reflection of the light beam on the curved surface of the mirror.

[0034] According to one example, the calibration element comprises a pair of mirrors, each having a curved surface, the two mirrors being arranged so that their normal axes intersect, the calibration element being configured so that the light beam transmitted by the injection optical fiber is reflected by one of the two mirrors towards the other and then reflected by the other of the two mirrors towards the collection optical fiber.

[0035] This second embodiment of the measurement system is adapted to a parallel configuration of the injection and collection optical fibers, common in insertion loss measurement systems for a photonic circuit. The use of two parabolic surface mirrors in this parallel configuration allows the transmission of the light beam from the injection optical fiber to the collection optical fiber. detection by two successive reflections of the light beam on the parabolic surfaces of the mirrors.

[0036] According to one example, the normal axes of the two mirrors of the calibration element form an angle of 90 degrees between them.

[0037] According to one example, the calibration element comprises a mirror having a curved surface and a prism surmounting the mirror, said prism having two upper faces inclined relative to each other, the calibration element being configured so that the light beam transmitted by the injection optical fiber is refracted by one of the upper faces towards the center of the curved surface of the mirror, then reflected off the mirror towards the other of the upper faces, and then refracted towards the collection optical fiber.

[0038] This third embodiment of the measurement system is also suitable for the parallel configuration of the injection and collection optical fibers. This third embodiment is advantageous in the case of a photonic circuit integrated into a thin substrate, where manufacturing two mirrors with normal axes forming a 90-degree angle between them is complicated. The addition of a prism in the calibration element makes it possible to transmit the light beam from the injection optical fiber to the collection optical fiber using a single mirror whose normal axis is perpendicular to the principal extension plane of the underlying substrate. The prism redirects the light beam toward the center of the curved surface of the mirror.

[0039] According to one example, the calibration element is directly integrated on the same substrate as the photonic circuit.

[0040] According to one example, the calibration element is integrated on a calibration substrate that is different from and separate from the substrate of the photonic circuit.

[0041] The calibration substrate, dedicated solely to system calibration, allows for flexibility in terms of the manufacturing of the calibration element, its dimensions, and the inclination of the normal axes of the mirrors. The constraints imposed by integrating the photonic circuit and the calibration element onto the same substrate are thus circumvented.

[0042] According to one example, the absolute insertion losses ILcp ​​of the photonic circuit are determined from the characterization optical power Pcp, the calibration optical power Prm, and the insertion loss ILmir of the calibration element, such that: TT - TT + P - P'cp — 'mtr x rm x cp*

[0043] The characterization optical power Pcp and the calibration optical power Prm are typically expressed in decibel-milliwatts (dBm). The characterization optical power Pcp is typically measured after the photonic circuit. The calibration optical power Prm is typically measured after the calibration device.

[0044] In the context of the present invention, the propagation of a light beam between a first optical element and a second optical element means that the beam can be routed from the first optical element to the second optical element, or conversely, that the beam can be routed from the second optical element to the first optical element. The light beam typically propagates along a path, a direction of propagation, or an optical path.

[0045] In the context of the present invention, an object or material "transparent" means that the object or material allows at least 90% of the light intensity of the light beam to pass through.

[0046] For the purposes of this application, "insertion losses" means the decrease in luminous power of a light beam propagating through an optical system, which occurs when an optical component such as a diffraction grating coupler, a mirror, or a fiber connector is inserted into the path of the light beam. These losses may be caused by the absorption, scattering, reflection, or dispersion of light as it passes through the optical component.

[0047] The term "reflection" or its equivalents refers to the phenomenon of re-emission from a surface of an incident light beam, along one or more directions opposite to the incident direction. In this description, a surface is considered reflective when it re-emits at least 85% of the intensity of an incident light beam. Reflection can be specular (one direction of reflection) or diffuse (several directions of reflection).

[0048] A substrate, layer, or device "based" on a material M means a substrate, layer, or device comprising only that material M or that material M and possibly other materials, for example alloying elements, impurities, or dopant elements.

[0049] A frame of reference, preferably orthonormal, comprising the x, y, z axes directed along the first, second and third directions is shown in the attached figures.

[0050] In this patent application, the terms thickness for a layer and height for a structure or device will be preferred. The thickness is measured in a direction normal to the principal extension plane of the layer, and the height is measured perpendicular to the basal xy plane of the substrate. Thus, a layer typically has a thickness along z, and a waveguide core has a height along z. The relative terms "on," "above," "under," and "below" refer to positions measured along the z direction.

[0051] Dimensional values ​​are understood to be within manufacturing and measurement tolerances.

[0052] The terms "approximately", "about", "in the order of" mean, when referring to a value, "within 10%" of that value or, when referring to an angular orientation, "within 10°" of that orientation. Thus, a direction substantially normal to a plane means a direction presenting an angle of 90+10° with respect to the plane.

[0053] Furthermore, the term "step" does not necessarily mean that the actions carried out during a step are simultaneous or immediately successive. Certain actions of a first step may, in particular, be followed by actions related to a different step, and other actions from the first step may be repeated later. Thus, the term "step" does not necessarily imply unitary actions that are inseparable in time and in the sequence of phases of the process.

[0054] The system 100 for measuring insertion losses in a photonic circuit 20 is now described with reference to figures IA and IB.

[0055] As illustrated in [Fig.1A], the system 100 is configured to measure the insertion losses of a photonic circuit 20 integrated on a substrate S. The substrate S extends in an xy plane defined by an x ​​direction and a y direction perpendicular to the x direction.

[0056] The photonic circuit 20 comprises at least one waveguide 23 and has an input interface 21 and an output interface 22 coupled to the waveguide 23. The input interface 21 is configured to receive a light beam F and direct it towards the waveguide 23 of the photonic circuit 20. The light beam F is guided by the waveguide 23 towards the output interface 22. The output interface 22 is configured to receive the light beam F and direct it out of the photonic circuit 20, towards another optical component such as a fiber for example.

[0057] In order to characterize this photonic circuit 20, it is important to measure the absolute insertion losses ILcp ​​of the photonic circuit 20. The system 100 allows these absolute insertion losses ILcp ​​to be measured using a circuit based on optical fibers connected together by fiber connectors. The system 100 comprises a light source 10 emitting the light beam F, a detector 30, an injection optical fiber 2, and a collection optical fiber 3. The injection optical fiber 2 is configured to transmit the light beam F to the input interface 21 of the photonic circuit 20, and the collection optical fiber 3 is configured to collect the light beam F from the output interface 22 of the photonic circuit 20.The injection 2 and collection 3 optical fibers are preferably aligned relative to the input 21 and output 23 interfaces, so as to transmit the light beam F to the input interface 21 and collect the light beam F from the output interface 22 while minimizing optical transmission losses. The injection 2 and collection 3 optical fibers may be aligned in the same plane not parallel to the xy plane, such that their axes intersect at a non-zero angle, or such that their axes are parallel.

[0058] The system 100 further includes a first optical fiber 1 connected to the light source 10 at one end and to the injection fiber 2 at the other end by means of a first connector C12. This first optical fiber 1 is configured to transmit the light beam F emitted by the light source 10 to the first fiber connector C12. The first connector C12 then transmits the light beam to the injection fiber 2. The system 100 also includes a second optical fiber 4 connected to the detector 30 and to the collection fiber 3 by means of a second connector C34. The second optical fiber 4 is configured to transmit, via the second fiber connector C34, the light beam F collected by the collection fiber 3 to the detector 30. The detector 30 is configured to measure the optical power of the light beam F transmitted by the second optical fiber 4.This optical power can be expressed in decibel-milliwatts (dBm) to facilitate calculations for extracting absolute insertion losses ILcp. The first and second fiber connectors C12 and C34 exhibit a first insertion loss ILcn and a second insertion loss ILC34, respectively. Since fiber connectors are passive elements, the associated insertion losses are positive. These insertion losses ILcn and ILC34 are due to the fiber splicing and can vary depending on factors such as the cleanliness of the contact surfaces or imperfect alignment of the spliced ​​fibers.

[0059] The system 100 further comprises a calibration element 110 as illustrated in [Fig. 1A]. The calibration element 110 has a fixed and precisely known insertion loss ILmir. The system 100 is configured to measure the absolute insertion losses ILcp ​​of the photonic circuit 20 after calibration of the system 100 via the calibration element 110.

[0060] The calibration element 110 can be integrated on the same substrate S as the photonic circuit 20, as illustrated in [Fig.1A], or can be integrated on a different substrate called the calibration substrate S', dedicated to the calibration of the system 100, as illustrated in [Fig.1B].

[0061] In order to characterize the absolute insertion losses ILcp ​​of the photonic circuit 20, a characterization optical power Pcp of the light beam F transmitted by the second optical fiber 4 is measured by the detector 30 in an absolute insertion loss characterization configuration ILcp ​​described previously with reference to Figures IA, IB. The light beam F emitted by the light source 10 propagates in the first optical fiber 1 with an initial intensity denoted II. The initial optical power PI of the light beam F propagating in the first optical fiber 1 can be expressed in dBm as a function of the initial intensity II in the following form:

[0062] [Math.l] mm 1 / H(mW) PI — 10 x log yj J

[0063] The optical power characterization Pcp of the light beam F transmitted by the second optical fiber 4 in the absolute insertion loss characterization configuration ILcp ​​of the photonic circuit 20 can be expressed in the following form:

[0064] [Math.2] Pcp = PI - ILci2 "ILcp" ILc34

[0065] The initial optical power PI of the light beam F propagating in the first optical fiber 1 cannot be measured directly in the characterization setup described above. For this reason, in conventional insertion loss measurement systems, a reference power Prf is measured in a reference setup illustrated in [Fig. 2]. In this reference setup, the first optical fiber 1 is disconnected from the first fiber connector C12, the second optical fiber 4 is also disconnected from the second fiber connector C34, and the first and second optical fibers 1 and 4 are connected to each other using a third fiber connector C14 having a third insertion loss ILCi4. This reference optical power can be expressed as follows:

[0066] [Math.3] Prf = Pl-ILC|4

[0067] Approximate insertion losses ILrf of the photonic circuit 20 can thus be extracted by calculating the difference between the measured reference power Prf and the measured characterization power Pcp, such that:

[0068] [Math.4] ILrf = Pu - Pcp

[0069] We deduce from the preceding equations the relationship between the approximate insertion losses ILrf and the absolute insertion losses ILcp, expressed in the following form:

[0070] [Math.5] ILrf = ILcp ​​- ILCi4 + ILC12 + ILc34

[0071] The absolute insertion loss value ILcp ​​cannot be obtained reliably and precisely in this way. Indeed, following this conventional approach, one arrives at approximate insertion losses ILrf that depend on the insertion losses ILCi4, ILCi2, ILC34 of the fiber connectors C14, C12, C34, which constitute a source of uncertainty in the measurement of insertion losses.

[0072] The insertion losses ILCÿ of a fiber connector Cij vary from one connection to another. The insertion losses ILCÿ are typically less than 0.5 dB, and in average equal to 0.2 dB. The standard deviation of the measured values ​​of the insertion losses ILCÿ of a Cij fiber connector is on the order of 0.2 dB. We deduce that the approximate insertion losses ILrf are on average greater than the absolute insertion losses ILcp, which can be symbolically expressed as follows: ILcp ​​ILrf - 0.2 dB ± 0.2 dB.

[0073] Introducing the calibration element 110 into the ILcp ​​absolute insertion loss measurement system 100 eliminates the ILCi4, ILCi2, and ILc34 insertion losses of the C14, C12, and C34 fiber connectors from the calculations. The ILmir insertion loss of the calibration element 110 is fixed and can be measured precisely, thus enabling reliable calibration of the measurement system 100.

[0074] Figure 3 illustrates the system 100 in a calibration configuration using the calibration element 110. In this calibration configuration, the injection optical fiber 2 is configured to transmit the light beam F to the calibration element 110. The calibration element 110 is configured to receive the light beam F and reflect the light beam F back to the collection optical fiber 3, which collects the light beam F reflected by the calibration element 110. This allows the optical calibration power Prm of the light beam F received and reflected by the calibration element 110 to be measured using the detector 30. This optical calibration power Prm can be expressed as follows:

[0075] [Math.6] Prm = Pl-ILcl2-ILniir-ILC34

[0076] A difference in AIL losses can be deduced by calculating the difference between the measured calibration power Prm and the measured characterization power Pcp, such that:

[0077] [Math.7] A IL = Prm - Pcp

[0078] The absolute insertion losses ILcp ​​of the photonic circuit 20 can be extracted from the preceding equations, and are expressed in the following form:

[0079] [Math. 8] ILcp = ILmjr + A IL

[0080] The insertion losses ILCu, ILCi2, ILC34 of the fiber connectors C14, C12, C34 are thus eliminated from the calculation because the fiber circuit remains unchanged when switching from the characterization configuration to the calibration configuration via the calibration element 110. In other words, the first and second optical fibers 1, 4 are not plugged / unplugged to switch from one configuration to the other.

[0081] As illustrated in [Fig. 3], the calibration element comprises at least one mirror 111 having an ellipsoidal curved surface. This ellipsoidal curved surface This device reflects the light beam F from the injection optical fiber 2 towards the collection optical fiber 3, focusing and directing the light beam F towards the collection optical fiber 3. During this reflection of the light beam F on the curved surface of the mirror 111, an insertion loss ILmir occurs. This insertion loss ILmir, associated with the calibration element 110, depends on the dimensions, shape, and nature of the curved surface of the mirror 111.

[0082] The dimensions of the mirror 111 are preferably chosen with respect to the position and type of the injection and collection optical fibers 2, 3, in order to optimize the transmission of the light beam F from one fiber to the other. The propagation direction of the light beam F incident on the surface of the mirror 111 and reflected on the mirror 111, as well as the normal axis of the mirror 111, lie in the same plane, which is not parallel to the xy plane. As illustrated in [Fig. 4], the position of the end of the injection optical fiber 2 is denoted A, the position of the end of the collection optical fiber 3 is denoted B, the position of the bottom of the ellipsoidal surface of the mirror 111 is denoted D, and the normal axis of the mirror 111 is the line (DO), O being the midpoint of the segment connecting points A and B. The light beam F propagates along the direction AD at the exit of the injection optical fiber 2.It is then reflected off the surface of mirror 111 and follows the DB direction after reflection to be subsequently collected by the optical collection fiber 3.

[0083] The dimensions of the ellipsoidal surface of mirror 111 can be calculated by considering an ellipsoidal surface of mirror 111. A more precise shape can be determined using optical simulation, for example, using the Zemax software with the POP (Physical Optics Propagation) module. However, for the reasoning below, we consider an ideal curved ellipsoidal surface. The parameters for determining the dimensions of the curved surface of mirror 111 include: • the working distance AB between the collection optical fiber 2 and the injection optical fiber 3 taken between points A and B; this working distance AB may preferably be equal to or less than 1 mm, • the angle α of the light beam F with respect to a direction z perpendicular to the x and y directions, with a typical value α = 11.6° for testing integrated photonic components, • the working height H of the injection and collection optical fibers 2, 3, relative to the bottom of the ellipsoidal surface, taken between points O and D, such that H = OD = AB / (2*tan(a)), • the distance L between the end of an optical fiber 2, 3, and the bottom of the ellipsoidal surface, taken between points A or B and D, such that L = AD = BD = AB / (2*sin(a)), • the radius of curvature Rx of the ellipsoidal surface at point D, taken along the x direction (parallel to AB), such that Rx = L*(l - (AB / (2L))2)1 / 2, • the radius of curvature Ry of the ellipsoidal surface at point D, taken along the y direction (perpendicular to plane ABD), such that Ry = H, • the wavelength X of the light beam F, • the diameter of the MFD propagation mode of the light beam F in the optical fiber 2, 3, diameter evaluated at 1 / e2 in intensity (MFD or "Mode Field Diameter" in English), • the Rayleigh length zR of the light beam F in air, such that zR = ir*(MFD / 2)2 / X, • the total divergence 0 at 1 / e2 of the light beam F at the output of the optical fiber 2, 3, such that 0 = MFD / zR, • the diameter W of the propagation mode of the light beam F at 1 / e2 on the curved surface of the mirror 111, such that W = H*0.

[0084] The injection and collection optical fibers 2, 3 can be standard fibers widely used in the telecommunications field, such as Corning SMF-28 fibers, for example. The X wavelength of the light beam F can be chosen from the near-infrared wavelength range, which is commonly used in telecommunications networks. The X wavelength of the light beam F can be, for example, around 1.310 pm.

[0085] The table below summarizes the parameters used to calculate the dimensions of the mirror 111. This table also presents examples of values ​​chosen according to an example embodiment of the system 100, for fibers compatible with telecommunications networks. [Tables 1] Parameter Value Working distance AB between the injection and collection optical fibers AB = 0.5 mm Angle α of the light beam F relative to the direction za = 11.6° Working height H of the injection and collection optical fibers H = 1.22 mm Distance L between the aperture of an optical fiber and the bottom of the ellipsoid L = 1.25 mm Radius of curvature Rx at point D along the x direction Rx = 1.27 mm Radius of curvature Ry at point D along the y direction Ry = 1.22 mm Refractive index n of silica n = 1.447 Wavelength X of the light beam X = 1.310 pm Diameter of the propagation mode MDF of the light beam F in the optical fiber MDF = 9.2 pm Rayleigh length zR of the light beam zR = 50.7 pm Total divergence θ at 1 / e² of the light beam θ = 10.4° Diameter W of the light beam at 1 / e² on the curved surface of the mirror W = 0.22 mm

[0086] The diameter Dmir of the mirror 111 is chosen so that it is sufficiently large to reflect the entire light beam F, for example, such that Dmir = 1.6*W. The diameter Dmir of the mirror 111 can thus be equal to 0.35 mm. The depth hmir of the ellipsoidal surface of the mirror 111, measured along the normal axis of the mirror, is such that hmir = D2 / (8*Ry). The depth hmir of the ellipsoidal surface of the mirror 111 can thus be equal to 13 pm.

[0087] The insertion loss ILmir of mirror 111 can be calculated from the following equation:

[0088] [Math.9] ILmir — 2 x ILintert- + ILK)rd + ILmet + ILsurf

[0089] In this equation, ILinterf represents the loss due to partial reflection of the light beam F on the air / glass interface of the optical fiber 2, 3, ILbord represents the loss due to the limitation of the light beam F at the edge of the mirror 111, ILmet represents the loss due to the reflection coefficient of the metal of the ellipsoidal surface which may be slightly less than 1, and ILsurf represents the loss due to imperfections of the ellipsoidal surface of the mirror 111.

[0090] The reflection coefficient R at the air / glass interface of the injection optical fiber 2 or the collection optical fiber 3 is such that R = ((nl) / (n+l))2 and can be equal to 3.3% for a refractive index of silica (SiO2) n = 1.447. The transmission Tinterf for each air / glass interface is such that Tinterf = 1 - R and can be equal to 96.7%, which corresponds to a loss ILinterf of 0.15 dB.

[0091] The transmission rate Tbord of the light beam F limited by the edge of the mirror 111 depends on the diameter Dmir of the mirror 111, such that Tbord = 1 - exp(-2 (Dmir / W)2). The transmission rate Tbord can, for example, be equal to 99.4% for the previous examples of Dmir and W values, which corresponds to a loss ILbord of the order of 0.03 dB.

[0092] The ellipsoidal surface of the mirror 111 can be covered with a metallic layer, such as an aluminum layer for example. For a layer thickness of aluminium around 200 nm, the reflection coefficient Rmet of the ellipsoidal surface of mirror 111 is around 96.6%, which corresponds to an ILmet loss of the order of 0.14 dB.

[0093] During the manufacturing of the mirror 111, several factors can lead to deviations from an ideal ellipsoidal shape. Manufacturing tolerances and mechanical deformations are among the factors that can contribute to imperfections in the shape of the curved surface of the mirror 111. The transmission rate Tsurf in the optical fiber of collection 3 due to these imperfections in the curved surface is such that Tsurf = exp(-(4ir*Aô / X)2), where Aô is the root mean square deviation between the curved ellipsoidal surface and an ideal ellipsoidal surface. The root mean square deviation Aô with respect to the ideal ellipsoidal shape can be precisely measured using a profilometer, for example, an optical profilometer that allows for precise measurements on the nanometer scale (< 5 nm). From this measurement of the root mean square deviation Aô, the ILsurf losses can be determined.

[0094] Current manufacturing technologies make it possible to obtain a root mean square deviation Aô from the ideal ellipsoidal shape of less than 30 nm. The ellipsoidal shape can be fabricated, for example, by grayscale photolithography and typically exhibits a root mean square deviation Aô of less than 20 nm. The following table shows some examples of ILsurf values, as a function of the root mean square deviation Aô. [Tables 2] Aô Tsurf TT c -1 ^surf 30 nm 92.1% 0.36 dB 20 nm 96.4% 0.16 dB 10 nm 99.1% 0.04 dB

[0095] The insertion loss ILmir of the mirror 111 can thus be calculated as a function of the parameters described above. For a root mean square deviation Aô = 20 nm, ILmir can, for example, be on the order of 0.62 dB ± 0.1 dB, where an uncertainty of 5 nm has been considered on the measurement of Aô.

[0096] As illustrated in Figures 5A and 5B, according to a first embodiment of the system 100, the calibration element 110 may comprise a single mirror 111 with a normal axis parallel to the z-direction. This calibration element 110 with a single mirror 111 can be used to calibrate a system 100 in which the axis of the injection optical fiber 2 and that of the collection optical fiber 3 are crossed as illustrated in [Fig. 5B] according to one embodiment. This crossed configuration of the injection 2 and collection optical fibers 3 can be adopted in the case where the The input interfaces 21 and output interfaces 22 of the photonic circuit 20 are arranged opposite each other as illustrated in [Fig. 5A]. According to this first embodiment, the light beam F transmitted by the injection optical fiber 2 is directly reflected off the mirror 111 towards the collection optical fiber 3.

[0097] As illustrated in Figures 6A and 6B, according to a second embodiment of the system 100, the calibration element 110 may comprise a pair of mirrors 111, each having a parabolic curved surface. According to this second embodiment, the two mirrors 111 are inclined with respect to the surface of the substrate S or the calibration substrate S', so that the light beam F transmitted by the injection optical fiber 2 is reflected by one of the two mirrors 111 towards the other and then reflected by the other of the two mirrors 111 towards the collection optical fiber 3, as illustrated in [Fig. 6B].

[0098] This calibration element 110 with double mirrors 111 can be used to calibrate a system 100 in which the axis of the injection optical fiber 2 and that of the collection optical fiber 3 are parallel as illustrated in [Fig. 0B] according to an exemplary embodiment. This parallel configuration of the injection optical fibers 2 and collection optical fibers 3 can be adopted in the case where the input interfaces 21 and output interfaces 22 of the photonic circuit 20 are arranged parallel to each other as illustrated in [Fig. 0A].

[0099] To enable efficient transmission of the light beam F from one fiber to the other, the normal axes of the two mirrors 111 can advantageously be oriented so that the two axes form a 90° angle with each other, according to one embodiment. In other words, the two mirrors 111 can be inclined at 45° with respect to the surface of the substrate S or the calibration substrate S'. According to this embodiment, the calibration substrate S', dedicated solely to calibrating the system 100, allows flexibility in terms of the fabrication of the calibration element 110, the dimensions of the calibration element 110, and the inclination of the normal axes of the mirrors 111. Indeed, the photonic circuits 20 are generally integrated on substrates S with thin thicknesses, which might not be optimal for the fabrication of mirrors 111 inclined with respect to the substrate surface.The constraints imposed by integrating the photonic circuit 20 and the calibration element on the same substrate S can be circumvented by integrating the mirrors 111 on a calibration substrate S', separately from the photonic circuit 20.

[0100] As illustrated in Figures 7A and 7B, according to a third embodiment of the system 100, the calibration element 110 may comprise a single mirror 111 and a prism 112 mounted above the mirror 111. The prism 112 has two upper faces 112a, 112b inclined relative to each other. According to this third embodiment, the calibration element 110 is configured such that the transmitted light beam F by the injection optical fiber 2 is refracted by the upper face 112a towards the center of the curved surface of the mirror 111, then reflected on the mirror 111 towards the upper face 112b, to be subsequently refracted towards the collection optical fiber 3 by the upper face 112b. This calibration element 110 comprising the prism 112 can be used to calibrate a system 100 in which the injection optical fibers 2 and collection 3 are arranged in a parallel "close" configuration, where the optical fibers 2, 3 are in the immediate vicinity of each other.

[0101] This third embodiment of the measurement system is therefore particularly suited to the close parallel configuration. It is also advantageous in the case of a photonic circuit 20 integrated into a thin substrate S, in which the fabrication of two mirrors 111 having normal axes forming a 90° angle between them is complicated. The introduction of a prism 112 in the calibration element 110, above the mirror 111, makes it possible to transmit the light beam F from the injection optical fiber 2 to the collection optical fiber 3 using a single mirror 111 whose normal axis is perpendicular to the underlying substrate. The light beam F is advantageously redirected towards the center of the curved surface of the mirror 111 by the prism 112.

[0102] According to a first example illustrated in [Fig. 7A], the injection 2 and collection 3 optical fibers can be arranged parallel to each other on a strip 50 serving as a support for the fibers. The strip 50 can be configured so that the injection 2 and collection 3 optical fibers are inclined with respect to the z-direction at an angle of approximately 8°. The lower face 50b of the strip 50 can be parallel to the surface of the substrate S or the calibration substrate S', which in this example is parallel to the xy plane. According to this first example, the tangent plane of the mirror 111 is parallel to the surface of the substrate S or the calibration substrate S'. According to this example, the upper faces 112a, 112b of the prism 112 can be tilted, so that the apex edge of the prism 112 is inclined with respect to the surface of the substrate S or that of the calibration substrate S'.The angle formed between the two upper faces 112a, 112b of the prism 112 can preferably be chosen so that the directions of propagation of the light beam F inside the prism 112, before and after reflection on the mirror 111, are included in a vertical plane parallel to the plane (xz) containing the axis of the mirror 111.

[0103] According to a second example illustrated in [Fig. 7B], the injection 2 and collection 3 optical fibers can also be arranged parallel to each other on the strip 50. The strip 50 can be configured so that the injection 2 and collection 3 optical fibers are inclined at an angle of 8° to the z-direction. The lower face 50b of the strip 50 can be parallel to the xy plane. According to this second example, the normal axis of the mirror 111 is perpendicular to the surface of the substrate S or of the calibration substrate S', the apex edge of the prism 112 is parallel to the surface of the substrate S or that of the calibration substrate S'.

[0104] In order for the normal axis of the mirror 111 to lie in the plane formed by the propagation directions of the light beam F propagating inside the prism 112, before and after reflection on the mirror 111, a shim 60 having a surface inclined with respect to the xy plane can be inserted below the substrate S or the calibration substrate S'. This shim 60 allows the surface S or S' to be inclined so that it is perpendicular to the plane containing the axis of the beam from fiber 2 and the axis of the beam received by fiber 3. In this example, the surface of the shim 60 is inclined at 11.6° with respect to the xy plane. The surface of the shim 60 can advantageously be oriented with respect to the xy plane with a precision finer than 1 arcmin, i.e., 0.017°.

[0105] In the two embodiments described above, the prism 112 is joined to the mirror 111 using an adhesive that polymerizes under ultraviolet light. The faces of the prism 112 can advantageously be oriented relative to each other with a precision finer than 0.017°. The prism 112 and the mirror 111 are preferably joined with a lateral precision finer than 0.5 pm relative to the center of the curved surface of the mirror 111. The prism 112 can be made of silica glass (SiO2), for example.

[0106] The method for measuring the absolute insertion losses ILcp ​​of the photonic circuit 20 using the system 100 is now described.

[0107] The method comprises a calibration step and a measurement step. The calibration step includes transmission, via the first optical fiber 1 coupled to the light source 10, of the light beam F emitted by the light source 10 to the injection optical fiber 2, passing through the first fiber connector C12. The light beam F is then transmitted via the injection optical fiber 2 to the calibration element 110. The light beam F is reflected by the calibration element 110 to the collection optical fiber 3. The method further comprises collection, via the collection optical fiber 3, of the light beam F from the calibration element 110. The light beam F is then routed to the second optical fiber 4, passing through the second fiber connector C34. The light beam F is transmitted via the second optical fiber 4 to the detector 30.

[0108] In this calibration step, a first measurement of the calibration optical power Prm of the light beam F transmitted by the second optical fiber 4 is carried out using the detector 30. This first measurement makes it possible to determine or take into account the first and second insertion losses ILcn, ILC34 of the first and second connectors Cl3, C34.

[0109] The method further comprises, during a measurement step, transmission via the first optical fiber 1 of the light beam F emitted by the light source 10 to the injection optical fiber 2, passing through the first fiber connector C12. The light beam F is then transmitted via the injection optical fiber 2 to the input interface 21 of the photonic circuit 20. The collection optical fiber 3 then collects the light beam F from the output interface 22 of the photonic circuit 20 and transmits it to the second connector C34. The light beam F passes through connector C34 into the second optical fiber 4. The light beam F is then transmitted via the second optical fiber 4 to the detector 30. During the measurement step, a second measurement by the detector 30 of the optical power characterization Pcp of the light beam F is performed.This second measurement allows us to determine the absolute insertion losses ILcp ​​of the photonic circuit 20 by following the calculation described previously.

[0110] According to this calculation, the absolute insertion losses ILcp ​​of the photonic circuit 20 can be determined from the characterization optical power Pcp measured during the measurement step, the calibration optical power Prm measured during the calibration step, and the insertion loss ILmir of the calibration element 110, by: ILcp ​​= ILmir + P - PT / (1 / rm x cp)

[0111] The method may further include a step of aligning the injection 2 and collection 3 optical fibers according to one of the crossed or parallel fiber configurations described above. The method may further include, during the calibration step, one or more reflections and / or refractions of the light beam F by the calibration element 110.

[0112] According to the first embodiment of the system 100, in which the calibration element comprises a single mirror 111, the method typically includes a single reflection of the beam F on the curved surface of the mirror 111 during the calibration step.

[0113] According to the second embodiment of the system 100, in which the calibration element comprises a pair of mirrors 111, the method typically includes two reflections of the beam F on the curved surfaces of the two mirrors 111 during the calibration step.

[0114] According to the third embodiment of the system 100, in which the calibration element comprises a single mirror 111 surmounted by a prism, the method may include a refraction of the light beam F by one of the upper faces 112a, 112b of the prism 112, a reflection of the light beam F on the curved surface of the mirror F and a second refraction of the light beam F by the other of the upper faces 112a, 112b of the prism 112.

[0115] At least one mirror 111 of the calibration element can be integrated onto the same substrate S as the photonic circuit 20, for example, by grayscale lithography or nanoprinting. It can also be fabricated on the calibration substrate S' using possibly different manufacturing techniques, such as diamond-tipped turning.

[0116] The photonic circuit 20 may include several waveguides 23 and other integrated elements such as directional couplers, transducers, interferometers, or Bragg mirrors. The substrate S may be based on glass, silicon, or other materials. The calibration substrate S' may be based on a different material than that of the substrate S.

[0117] The invention is not limited to the embodiments described above. Various specific examples of the 100 insertion loss measurement system have been described. Other embodiments are possible, for example, by combining features described above, without departing from the principle of the present invention. Furthermore, the features described with respect to one aspect of the invention can be combined with another aspect of the invention.

Claims

1. Demands System (100) for measuring the absolute insertion loss ILcp ​​of a photonic circuit (20) integrated on a substrate (S), said photonic circuit (20) comprising at least one waveguide (23), an input interface (21) and an output interface (22) coupled to the at least one waveguide (23), the input interface (21) being configured to receive a light beam (F) and direct it towards the at least one waveguide (23) and the output interface (22) being configured to receive the light beam (F) transmitted by the at least one waveguide (23) and direct it away from the photonic circuit (20), said system (100) comprising: • a light source (10) emitting the light beam (F), • a first optical fiber (1) coupled to the light source (10), configured to transmit the light beam (F) to a first fiber connector (Cl2), • an injection optical fiber (2) connected to the first optical fiber (1) using the first fiber connector (Cl2), said injection optical fiber (2) being configured to transmit the light beam (F) to the input interface (21) of the photonic circuit (20), the first fiber connector (C12) having a first insertion loss ILCi2, • a collection optical fiber (3) configured to collect the light beam (F) from the output interface (22) of the photonic circuit (20), • a second optical fiber (4) connected to the collection optical fiber (3) using a second fiber connector (C34), the second fiber being configured to transmit the light beam (F) collected by the collection optical fiber (3) to a detector (30), the second fiber connector (C34) having a second insertion loss ILC34, • the detector (30) configured to measure an optical power characterization Pcp of the light beam (F) transmitted by the second optical fiber (4), said system (100) being characterized in that it further comprises a calibration element (110) having a fixed and known insertion loss ILmir, said calibration element (110) making it possible to overcome the first and second insertion losses ILcn, ILC34, so that the system (100) is configured to measure the absolute insertion losses ILcp ​​of the photonic circuit (20), after calibration of the system (100) via the calibration element (110).

2. System (100) according to the preceding claim, wherein: • the calibration element (110) is configured to receive the light beam (F) transmitted by the injection optical fiber (2), and reflect the light beam (F) back to the collection optical fiber (3), • the detector (30) is configured to measure a calibration optical power Prm of the light beam (F) received and reflected by the calibration element (110).

3. System (100) according to any one of the preceding claims, wherein the calibration element (110) comprises at least one mirror (111) having an ellipsoidal curved surface and wherein the injection optical fiber (2), the collection optical fiber (3) and a normal axis of the at least one mirror (111) are contained in the same plane not parallel to a principal extension plane of the substrate (S).

4. System (100) according to the preceding claim, wherein the root mean square deviation Aô between the curved ellipsoidal surface and an ideal ellipsoidal surface is less than or equal to 30 nm, preferably less than or equal to 20 nm.

5. System (100) according to any one of claims 3 and 4, wherein the curved surface of at least one mirror (111) is coated with an aluminum-based metallic layer.

6. System (100) according to any one of claims 3 to 5, wherein the insertion loss ILmir of the calibration element (110) is less than or equal to 0.85 dB ± 0.1 dB.

7. System (100) according to any one of the preceding claims, wherein the calibration element (110) comprises a single mirror (111) configured so that the light beam (F) transmitted by the injection optical fiber (2) is directly reflected off the mirror (111) towards the collection optical fiber (3).

8. System (100) according to any one of claims 1 to 6, wherein the calibration element (110) comprises a pair of mirrors (111) each having a curved surface, the two mirrors (111) being arranged so that their normal axes intersect, the calibration element (110) being configured so that the light beam (F) transmitted by the injection optical fiber (2) is reflected by one of the two mirrors (111) towards the other and then reflected by the other of the two mirrors (111) towards the collection optical fiber (3).

9. System (100) according to the preceding claim, wherein the normal axes of the two mirrors (111) of the calibration element (110) form an angle of 90 degrees between them.

10. System (100) according to any one of claims 1 to 6, wherein the calibration element (110) comprises a mirror (111) having a curved surface and a prism (112) surmounting the mirror (111), said prism (112) having two upper faces (112a, 112b) inclined relative to each other, the calibration element (110) being configured such that the light beam (F) transmitted by the injection optical fiber (2) is refracted by one of the upper faces (112a, 112b) towards the center of the curved surface of the mirror (111), then reflected off the mirror (111) towards the other of the upper faces (112a, 112b), and then refracted towards the collection optical fiber (3).

11. System (100) according to any one of the preceding claims, wherein the calibration element (110) is directly integrated on the same substrate (S) as the photonic circuit (20).

12. System (100) according to any one of claims 1 to 10, wherein the calibration element (110) is integrated on a calibration substrate (S') different from and separate from the substrate (S).

13. Method for measuring absolute insertion losses ILcp ​​of the photonic circuit (20) using the system (100) according to any one of the preceding claims, said method comprising, during a calibration step: • transmission through the first optical fiber (1), of the light beam (F) emitted by the light source (10), to the injection optical fiber (2) via the first fiber connector (Cl2), • transmission via the injection optical fiber (2) of the light beam (F) to the calibration element (110), • a collection by the collection optical fiber (3) of the light beam (F) from the calibration element (110), to the second optical fiber (4), passing through the second fiber connector (C34), • a transmission via the second optical fiber (4), of the light beam (F) collected by the collection optical fiber (3), towards the detector (30), • a first measurement by the detector (30) of the calibration optical power Prm of the light beam (F) transmitted by the second optical fiber (4), said first measurement including the first and second insertion losses ILcn, ILC34, said process further includes, during a measurement step: • a transmission via the first optical fiber (1), of the light beam (F) emitted by the light source (10), to the injection optical fiber (2) via the first fiber connector (Cl2), • transmission via the injection optical fiber (2) of the light beam (F) to the input interface (21) of the photonic circuit (20), • a collection by the optical collection fiber (3) of the light beam (F) coming from the output interface (22) of the photonic circuit (20), to the second optical fiber (4), passing through the second fiber connector (C34), • a transmission via the second optical fiber (4), of the light beam (F) collected by the collection optical fiber (3), towards the detector (30), • a second measurement by the detector (30) of the optical power characterization Pcp of the light beam (F) transmitted by the second optical fiber (4), said second measurement allowing the determination of the absolute insertion losses ILcp ​​of the photonic circuit (20), by comparison with the first measurement.

14. Method for measuring absolute insertion losses ILcp ​​of the photonic circuit (20) according to the preceding claim, wherein the absolute insertion losses ILcp ​​of the photonic circuit (20) are determined from the characterization optical power Pep, the calibration optical power Prm and the insertion loss ILmir of the calibration element (110), such that: ILcp ​​= ILmir + Prm - Pcp.