WO2004104661A1

WO2004104661A1 - Integrated photonic circuit fitted with means for interconnection with add-on optoelectronic components

Info

Publication number: WO2004104661A1
Application number: PCT/FR2004/050204
Authority: WO
Inventors: Serge Valette; Christophe Martinez
Original assignee: Teem Photonics
Priority date: 2003-05-19
Filing date: 2004-05-19
Publication date: 2004-12-02
Also published as: EP1625435A1; FR2855274A1; FR2855274B1

Abstract

The invention relates to an integrated photonic circuit comprising a substrate (1) comprising at least one optical circuit (4) and means (2,3) for interconnecting the optical circuit with at least one optoelectronic component (6, 8, 9, 11, 12) which is added onto the substrate. The interconnecting means comprise at least one substrate area wherein the index of refraction is modified in order to ensure said interconnection. The area of the substrate comprises at least one lens with an index gradient.

Description

PHOTONIC INTEGRATED CIRCUIT HAVING INTERCONNECTION MEANS WITH OPTOELECTRONIC COMPONENTS

REPORTED

DESCRIPTION

TECHNICAL AREA

The present invention relates to a photonic integrated circuit equipped with interconnection means with attached optoelectronic components.

STATE OF THE PRIOR ART

Photonic integrated circuits do not use a unique technological approach known as monolithic unlike microelectronic integrated circuits. For example, in the field of microelectronic integrated circuits for MOS technology on silicon substrate, the key component constituted by the MOS transistor is duplicated a large number of times to build the desired integrated circuit. Photonic integrated circuits, on the contrary, involve very different individual components in terms of functionality or geometry and which, more often than not, call upon various technologies produced on different substrates. Thus, laser diodes or certain types of optical modulators or amplifiers are produced on InP substrates, very high frequency electro-optical modulators are produced on lithium niobate substrates, passive components (power dividers, wavelength multiplexers, etc.) are produced on glass substrates or silicon. The technologies involved in the production of these components are very different from each other: epitaxy of thin semiconductor layers on InP substrates, diffusion of titanium or proton exchange on lithium niobate substrates, ion exchange on substrates in glass or CVD deposition of silica layers on silicon substrates.

In addition, some components are not currently available in an integrated version. They are carried out either on optical fiber (spectral filters for example), or in free space micro-optics (light isolators, certain interference filters, etc.). In addition, the transmission of light information uses single-mode or even multimode optical fibers, the optical and geometrical characteristics of which are most often different from those of the mentioned integrated circuits.

This strongly hybrid character of optoelectronics implies the use of sophisticated connection techniques, difficult to master (use of micro-lenses and need for extremely precise alignment, etc.) These techniques are very disadvantageous from an economic point of view. These problems of connection and adaptation of optical impedance between the elements of optoelectronic devices are very generic problems which arise in all fields of application of optoelectronics from instrumentation (industrial, medical, scientific, etc. ) up to telecommunications optics. These problems are all the more critical as production cost constraints become high, which is more and more the case in optical communications as optical transmissions approach the subscriber and as the number of components or optical modules involved increases enormously in number of parts.

STATEMENT OF THE INVENTION

The invention proposes a photonic integrated circuit making it possible to remedy the above drawbacks.

It therefore relates to a photonic integrated circuit comprising a substrate comprising at least one optical circuit and means for interconnecting the optical circuit with at least one optoelectronic component attached to the substrate, the interconnection means consisting of at least one area of said substrate whose refractive index is modified to ensure said interconnection, said area of the substrate comprising at least one lens with an index gradient.

An optoelectronic component can thus be added to the periphery of the substrate, on one face of the substrate or in a cavity of the substrate. The interconnection between the optical circuit of the photonic integrated circuit and the attached optoelectronic component is then made by optical routing in free space. The invention therefore makes it possible to associate, on the same integrated circuit, optical elements in free space (3D) and planar optics (2D). There are numerous publications mentioning the production of planar optics components allowing a modification, in general partial, of the mode profiles of a planar guide via the use of adiabatic structure whose role can be assimilated to that of a " funnel "for light. Most often and for reasons related to the use of masks to define the desired patterns in guided optics, this type of adaptation is brought into play in a single dimension defined by the direction perpendicular to the plane of the propagation of the wave. and located in the plane of the substrates used.

Solutions for producing adiabatic structures allowing a modification of mode in the two directions perpendicular to the direction of propagation have been proposed but have generally encountered a very significant increase in technological stages, which are too penalizing in terms of manufacturing yields to be really used. In any case, this type of structure based on clever stacks of layers of different thicknesses retains in all cases the guided nature of the light wave and has never used the passage from guided optics to free space optics ( or vice versa) as proposed in the invention.

The optical circuit can in particular be produced by an ion exchange technique. The same applies to the substrate area constituting the interconnection means. The realization, during a manufacturing process, of elements in planar optics and in 3D optics is original. Indeed, although based on a physical approach, similar to modifying the optical parameters of glass (the exchange between glass ions and external ions provided by suitable salt baths), these two types of elements bring into play very different technological approaches. The production of the planar components involves masking steps which define the desired patterns in the surface plane of the substrates (or of the wafer by a collective production). The production of 3D components, and more particularly of gradient index lenses, is based on a symmetrical diffusion approach of the ionic species from not a planar substrate but from bars immersed in suitable salt baths, without involving masks. According to the prior art, these two types of elements use substrates of completely different geometries and which are not suitable for implementing the invention. The production of these elements can be obtained by chemical or ionic etching techniques or by irradiation using a laser beam. The ion exchange technique gives greater freedom of configuration and seems well suited for the industrial implementation of the concept of one invention.

The optical circuit may include an optical guide, one end of which is situated opposite said zone of the substrate in order to make a connection with said zone of the substrate. Said area of the substrate can be located in the substrate so as to be opposite an input or an output of the optoelectronic component in order to make a connection with said optoelectronic component. It can include at least two gradient index lenses whose axes are offset from each other.

The optoelectronic component can be attached to the periphery of the substrate or to the surface of the substrate. In the latter case, the substrate can be provided with at least one cavity allowing the housing of at least one optoelectronic component and / or at least one optical component.

The invention also relates to a process for the collective production of photonic integrated circuits as defined above, characterized in that it comprises the following steps: supply of a wafer intended to supply as many substrates as there are photonic integrated circuits to be produced,

- production of interconnection means on the wafer for each photonic integrated circuit,

- production on the wafer of the optical circuit of each photonic integrated circuit, possibly, production on the wafer of housing means of at least one optoelectronic and / or optical component for each photonic integrated circuit, - cutting of the wafer to obtain the circuits integrated photonics. The method can also comprise a step of transferring or adding at least one optoelectronic component and / or at least one optical component to complete the photonic integrated circuits. The interconnection means can be produced by means of a mask having a main opening of generally rectangular shape and adjacent sides connected by a fillet. The opening may have two opposite sides of convex and / or concave shape. It can have two opposite sides of convex shape and the other two opposite sides of concave shape. All sides of the opening can be convex in shape. They can all be concave. The mask may also have at least one secondary opening located near at least one side of the main opening.

BRIEF DESCRIPTION OF THE DRAWINGS

The invention will be better understood and other advantages and features will appear on reading the description which follows, given by way of nonlimiting example, accompanied by the appended drawings among which:

- Figures 1A-1D are top views of ^one substrate subjected to the manufacturing steps to provide an integrated photonic circuit according to the present invention,

FIGS. 2A to 2C illustrate certain steps of a process for the collective production of the photonic integrated circuit of FIGS. 1A to 1D. Figures 3 and 4 are partial views respectively from above and side of a photonic integrated circuit according to the invention, Figures 5 and 6 are partial views respectively from above and side of another photonic integrated circuit according to 1 invention, Figure 7 is a side and partial view of another photonic integrated circuit according to the invention, - Figure 8 is a side and partial view of yet another photonic integrated circuit according to the invention, FIG. 9 is a side and partial view of yet another photonic integrated circuit according to the invention,

- Figures 10 and 11 give possible embodiments of interconnection zones for photonic integrated circuits according to the invention, Figures 12 to 14 give other possible embodiments of interconnection zones for photonic integrated circuits according to the invention, FIG. 15 represents a cylindrical optical element capable of providing a lens with circular symmetry, FIG. 16 represents a form of profile of index of refraction for producing an interconnection zone according to the present invention, FIGS. 17 to 24 represent different configurations of masks possible for implementing the present invention. DETAILED PRESENTATION OF PARTICULAR EMBODIMENTS

Figures 1A to 1D illustrate the concept of the invention in general. They show the main technological steps leading to the realization of a photonic integrated circuit according to

1 invention.

FIG. 1A shows a glass substrate 1 on which zones 2 and 3 have been produced intended to ensure the interconnection between an optical circuit to be produced on the substrate 1 and optoelectronic components to be attached. In general, zones 2 and 3 intended for interconnection are produced before the optical circuit or circuits in order to minimize the interactions between the manufacturing processes. This is shown in Figure 1A.

FIG. 1A shows interconnection zones 2 located at the periphery of the substrate 1 for connection with electrooptical components which will be connected at the periphery of the substrate. It also shows interconnection zones 3 located inside the surface of the substrate 1 for the connection of an optoelectronic component to be located within the substrate 1 itself.

FIG. 1B shows the substrate 1 after the production of an optical circuit 4 in relation to interconnection zones 2 and 3 pre-integrated in the previous step.

FIG. 1C shows that a cavity 5 has been produced on the substrate 1, between the interconnection zones 3. The cavity 5 has dimensions chosen to allow the positioning of a optoelectronic component to be connected in the optical circuit 4.

FIG. 1D represents the photonic integrated circuit obtained when the optoelectronic components provided have been transferred to the substrate 1. As an example, a semiconductor laser diode 6, a single-mode optical input fiber 7 has been transferred to the periphery of the substrate 1 , an integrated optical circuit 8 involving another technology and / or another substrate than the substrate 1, a multimode optical fiber 9, a spectral filter 10 associated with a photodetector 11. Between the zones 3 and in the cavity 5, a An insulator 12 has been housed. The interconnection zones 2 and 3 are obtained by local modification Δni (x, y) of the refractive index n _v of the glass substrate, for example by ion exchange techniques.

The value of the Δni (x, y) profiles as well as their positioning within or at the periphery of the substrate 1 are adapted to the optical characteristics of the components or elements to be connected on the one hand and of the optical circuit or circuits on the other hand. Strictly speaking, each Δni (x, y) profile may be different from the others but, for reasons of technical simplicity, it is of course preferable to limit the number of different profiles necessary and, as far as possible, to play on the respective positions of the components or elements to optimize the various connections required. The concept of one invention is compatible with a collective manufacturing method of integrated circuits as illustrated by FIGS. 2A to 2C. Figure 2A shows a glass plate

100 intended for the manufacture of a plurality of photonic integrated circuits of the type described in FIGS. 1A to 1D. To facilitate understanding, the delimitation of the substrates 1 has been shown. FIG. 1A shows that the interconnection zones 2 and 3 have been made collectively.

FIG. 2B shows the collective embodiment of the optical circuits 4.

FIG. 2C shows the collective production of the cavities 5.

At the end of the collective manufacturing treatment, the wafer is cut to provide a plurality of photonic integrated circuits such as that shown in FIG. 1C. The construction of the interconnection zones will now be described in more detail.

Figures 3 and 4 are respectively top and side views of a photonic integrated circuit according to the invention. This integrated circuit is formed on a glass substrate 21. This circuit has an association between a single-mode or multimode optical guide 24 and an integrated interconnection zone 22 arranged opposite one end of the optical guide 24. The zone 22 consists of a lens with index gradient of pre-calculated parameters so as to obtain at the output of the integrated circuit, and in response to a light beam conveyed by the optical guide 24, a collimated optical beam 20 of given diameter and of low angular divergence. Figures 5 and 6 are respectively top and side views of another photonic integrated circuit according to the invention.

This integrated circuit is formed on a glass substrate 31. This circuit has an association between a single-mode or multimode optical guide 34 and an integrated interconnection zone 32 disposed opposite one end of the optical guide 34. The zone 32 consists of a lens with a gradient of index of pre-calculated parameters so as to obtain, at the output of the integrated circuit, and in response to a light beam conveyed by the optical guide 34, an optical beam 30 refocused at a distance and with pre-defined focal point dimensions so as to form an image of the output of the light guide of given position and magnification, this image being able to be real or virtual according to the type of coupling to be optimized.

The interconnection zones 22 or 32 are defined by conventional lithographic masking methods as will be described later. In the two previous examples, the light guide is buried below the surface of the substrate, at a depth X _0G (see Figures 4 and 6) and the axis of the index gradient lens at a depth X _oL . In Figures 4 and 6, the depth X _oG and X _oL were chosen to be equal and the axes of the optical beams generated by the gradient index lenses remain parallel to the axis of the guide.

It is quite possible to choose different depths for X ₀ G and X ₀ ι ,. This is shown in FIG. 7. The integrated circuit is formed on a glass substrate 41 and has an association between an optical guide 44 and an integrated interconnection zone 42 arranged opposite one end of the optical guide 44. The zone 42 consists of a lens with a gradient of index of pre-calculated parameters so as to obtain, at the output of the integrated circuit, and in response to a light beam conveyed by the optical guide 44, an optical beam collimated 40.

In the case shown, the depths _0G X and X _oL are different. These depths are offset by a value ΔX = X _oG - X _OL corresponding to an angle inclination Δθ determined and directly linked to the offset ΔX.

Similarly, when viewed from above (in the surface plane of the integrated circuit), these same optical beam axes can be offset by a value ΔY which leads in this plane to an angular offset Δφ.

From this simple configuration where the interconnection zone consists of a simple index gradient lens, it is easy to imagine more complex configurations of integrated optical elements which will make it possible to overcome certain intrinsic limitations. One of these limitations can be reached if the particular constraints of optical circuit coupling require for example the production of a collimated beam of large diameter, corresponding to low desired angular divergence values. Indeed, the depth X _ûG of the axis of the light guides is most often of a few μm, or even a few tens of μm, in particular in the guides obtained by ion exchange in the glasses which, in practice, will be favorably used for implement the invention. The classic depth is around 15 μm although it is possible to exceed this value. This limitation of the value of X _0G results in a limitation of the diameters accessible in the XOZ cutting plane when using a single lens because the maximum value of the diameter cannot exceed 2X _oG due to the presence of the surface of the integrated circuit. This problem can be overcome in several ways.

According to a first solution, the planar guide can be buried locally in order to increase the depth of the object point whose image will be formed by the lens with index gradient produced. FIG. 8 illustrates this solution in the case of a collimated image beam.

In this figure, the integrated circuit is formed on a glass substrate 51 and has an association between an optical guide 54 and an integrated interconnection zone 52 arranged opposite one end of the optical guide 54. The guide optic 54 is buried at a depth X _oG over part of its length, then at a depth X ' _OG on another part, closer to zone 52. Thus, it is possible to obtain, by means of the lens forming the zone 52, a collimated beam 50 of relatively large diameter.

This solution can have drawbacks of ^'burying the fact that the light guides (eg in electric field as it is known to do with the ion exchange technique in glasses) comes with profile changes refractive index of the guided mode or modes and therefore of the dimensions of the object point to be considered. This effect can be taken into account, for example via the optical parameters of the lens. This variant can be applied to the case illustrated in FIG. 7.

According to the second solution, two lenses with axes offset in the XOZ plane can be used. FIG. 9 illustrates this solution in the case of a collimated beam.

In this figure, the integrated circuit is formed on a glass substrate 61 and has an association between an optical guide 64 and an integrated interconnection zone 62 comprising a first lens 162 facing one end of the optical guide 64 and a second lens 262 facing the outlet of the first lens 162. The optical guide 64 is buried at a depth X _0Gr the lens 162 at a depth X _oL ι and the lens 262 at a depth X _oL2 so that X ₀ L2>DI> Xoβ-

In the case where the first lens works with a magnification G _L ι equal to -1, we obtain the relation:

XoL2 ⁼ XoLl + (XoLl ~ OG) ⁼ 2X ₀ L1 ^- XoG - To do this, it is sufficient to form the image of the exit of the light guide on the side opposite its position relative to the axis of the first lens 162. In this case, only the lower part of the lens 162 is involved in the phenomenon. deflection of light rays and colli ation. The collimated beam can therefore no longer see the surface of the optical guide and the use of a lens 262 that is not symmetrical with respect to its axis (defined by the position of the maximum refractive index for the index gradient produced) and of sufficient depth to avoid the desired diameter limitation for the collimated output beam 60.

The magnification G _L ι can be chosen different from -1. It can be equal to -0.5 or -0.25 to name a few interesting examples. The use of a magnification less than 1 makes it possible for example, for a given diameter value, to shorten the focal length of the lens 262 and therefore the length of this lens working with an infinite magnification due to the increase in the digital aperture image of the optical beam diffracted by the light guide.

FIG. 10 gives a first exemplary embodiment of an interconnection zone for a photonic integrated circuit according to the invention. In this example, the substrate is seen in section along the XOZ plane. The dashed line 75 represents the surface of the substrate. The optical guide 74 is also shown, one end of which is opposite the zone. interconnection 72 consisting of a first lens 172 and a second lens 272.

An OZ (representing the propagation distance of a light wave) and OX (representing the burial depth) axis diagram and superimposed on the cross-section of the substrate. The axis OZ originates from the end of the optical guide 74. The axis OX originates from the central part of the lens 172. The beam passing through the interconnection zone 72 has the same configuration as for FIG. 9.

The image point given by the first lens 172 coincides with the input face of the second lens 272. This configuration is only one illustration among many other possibilities which makes it possible to obtain, from an object point corresponding to the light distribution at the output of the waveguide, a collimated optical beam of given size.

FIG. 10 gives an example of a possible embodiment with parameters of index gradient lenses achievable by the technique of ion exchange in planar structure which will be described later.

The characteristics of the interconnection zone are as follows for a wavelength of the transported light beam of 1.55 μm:

- burial depth of the inlet guide, X ₀ G = 15 μm;

- radius of the "waist" of the guided mode 5 μ; - refractive index of the glass substrate maximum index variation of the first lens 0.08;

- burial depth of the first lens 20 μm; - radius of first lens 20 μ;

- length of the first lens 90 μm; maximum index variation of the second lens 0.03;

- burial depth of the second lens 25 μm;

- radius of the second lens 100 μ;

- length of the second lens 800 μm. It follows from these characteristics that the burial depth X _F c of the collimated beam (from the axis of the beam relative to the surface of the substrate) is 66 μm and that the width of the collimated beam ΔX _FC is 63.9 μ.

FIG. 11 gives a second embodiment of an interconnection zone for a photonic integrated circuit according to the invention. As for the first exemplary embodiment, the substrate is seen in section along the XOZ plane. Line 85 represents the surface of the substrate. The optical guide 84 is also shown, one end of which is opposite the interconnection zone 82 consisting of a first lens 182 and a second lens 282.

As in the previous figure, an OZ and OX axis diagram is superimposed on the cross-section of the substrate. The origins of the axes are the same as before. The image point given by the first lens 182 coincides with the entry face of the second lens 282.

FIG. 11 gives an example of a possible embodiment with parameters of index gradient lenses achievable by the ion exchange technique in planar structure.

- burial depth of the entry guide,

- radius of the "aist" of the guided mode 5 μm;

- refractive index of the glass substrate

1.52 maximum index variation of the first lens 0.08; burial depth of the first lens 20 μm; - radius of the first lens 20 μm;

- length of the first lens 90 μm; maximum index variation of the second lens 0.03; burial depth of the second lens 25 μm

- radius of the second lens 150 μm,

- length of the second lens 1200 μm. It follows from these characteristics that the burial depth X _FC of the collimated beam (of the axis of the beam relative to the surface of the substrate) is 86.5 μm and the width of the collimated beam ΔX _FC is 95.8 μm.

From these examples, it appears that collimated beams of diameter greater than 90 μm are easily achievable for lengths of integration zone of the coupling elements of the order of 1500 μm, entirely compatible with the size requirements of optoelectronic devices. . Significantly larger beam diameters can be envisaged if necessary via various modifications to the parameters of the optical system.

Figures 12, 13 and 14 give other possible examples of interconnection zones which no longer provide collimation of the object point (image at infinity) but a finite distance image of the object.

FIG. 12 therefore gives a third exemplary embodiment of an interconnection zone. As for the two previous embodiments, the substrate is seen in section along the XOZ plane. The dotted line 95 represents the surface of the substrate. There is also shown the optical guide 94, one end of which is opposite the interconnection zone 92 consisting of a first lens 192 and a second lens 292.

As before, an OZ and OX axis diagram is superimposed on the section of the substrate. The origins of the axes are the same as before.

The characteristics of the interconnection zone are as follows for a wavelength of the transported light beam of 1.55 μm: - burial depth of the entry guide,

- radius of "aisf'du guided mode 5 μm;

- glass substrate refractive index 1.52; maximum index variation of the first lens 0.08; burial depth of the first lens 20 μm; - radius of the first lens 20 μm;

- length of the first lens 100 μ; maximum index variation of the second lens 0.08; burial depth of the second lens 30 μm;

- radius of the second lens 20 μm,

- length of the second lens 100 μm.

FIG. 13 gives a fourth exemplary embodiment of an interconnection zone. As for the three previous exemplary embodiments, the substrate is seen in section along the XOZ plane. The dashed line 105 represents the surface of the substrate. There is also shown the optical guide 104, one end of which is opposite the interconnection zone 102 consisting of a first lens 202 and a second lens 302.

As before, an OZ and OX axis diagram is superimposed on the section of the substrate. The origins of the axes are the same as before. The characteristics of the interconnection zone are as follows for a wavelength of the transported light beam of 1.55 μm:

- burial depth of the entry guide,

-ray of the "aist" of the guided mode 5 μm;

- refractive index of the glass substrate

1.52 maximum index variation of the first lens 0.08; burial depth of the first lens 20 μm;

- radius of the first lens 20 μm;

- length of the first lens 100 μm; - maximum index variation of the second lens 0.08; burial depth of the second lens 30 μ;

- radius of the second lens 20 μm, - length of the second lens 100 μm.

FIG. 14 therefore gives a fifth exemplary embodiment of an interconnection zone. As for the four previous exemplary embodiments, the substrate is seen in section along the XOZ plane. The dashed line 115 represents the surface of the substrate. There is also shown the optical guide 114, one end of which is opposite the interconnection zone 112 consisting of a first lens 212 and a second lens 312. As before, an OZ and OX axis diagram is superimposed on the section of the substrate. The origins of the axes are the same as before.

- burial depth of the entry guide,

- radius of the "waist" of the guided mode 5 μm; - refractive index of the glass substrate

1.52; maximum index variation of the first lens 0.08; burial depth of the first lens 20 μm;

- radius of the first lens 20 μm;

- length of the first lens 100 μm; maximum index variation of the second lens 0.08; - burial depth of the second lens 30 μm;

- radius of the second lens 50 μm,

- length of the second lens 200 μm. It follows from these characteristics that the burial depth X _F c of the collimated beam (of the beam axis relative to the surface of the substrate) is 86.5 μm and that the width of the collimated beam ΔX _FC is 95 , 8 μm.

The simulations shown in Figures 12 and 13 show special cases interesting in which the axis of angular symmetry of the image beam remains parallel to the surface plane. This configuration is obtained by judiciously choosing the parameters of the optical system formed by the two lenses constituting the interconnection zone and in particular the position of the axes of the lenses relative to the axis of the waveguide. It can also be seen, by comparing FIGS. 12, 13 and 14, that the angular opening of the image optical beam can be easily adapted as a function of the parameters of the two lenses involved and in particular of the lengths and the distances between the lenses. It is therefore possible, by a judicious choice of parameters, to achieve the optimum angular opening in order to obtain an optimum energy coupling with a component located outside the integrated optical chip (optical fiber, other light guide, etc. .).

All these different examples are only illustrative of the possibilities offered by the invention. Thanks to the approach proposed by the invention, it is possible to achieve all the configurations given by known optical systems comprising one, two or more lenses with index gradients and suitably chosen optical and geometric parameters.

Figures 8 to 14 show the sectional views of the lenses and therefore the optical imaging figures in the XOZ section plane.

In top view, as already illustrated by FIGS. 3 and 5, the lenses must also have the necessary imaging properties. The optical parameters following these planes parallel to the surface plane YOZ will be deduced from the corresponding refractive index profiles.

We must therefore clearly see all the optical structures presented as symmetrical or non-symmetrical two-dimensional structures.

It is also possible to play with these asymmetries to modify the geometric shape of the light beams between the object and image planes and obtain light distributions in the image plane better suited to optimal energy coupling between two components in guided optics or between a component in guided optics and an optical component in free space. This type of property can for example be used to optimize the coupling between a laser diode which emits a very astigmatic optical beam and a symmetrical light guide whose guided mode is for example very close to that of an optical fiber.

An extreme case is also provided by the cylindrical lenses for which the index profiles will be invariant along the planes parallel to the YO plane.

Control of the refractive index profiles Δni (x, y) or possibly Δ i (x, y, z), for example in the case of lenses with convex entry or exit surfaces, making it possible to obtain the properties One of the points of the invention to take into consideration is the desired imaging methods for each index gradient lens constituting the optical system. We will now describe in detail how to make the interconnection zones.

Traditionally, index gradient lenses sold commercially have parabolic refractive index profiles with circular symmetry because the ion exchange between the exchange bath and the glass cylinder used takes place under perfect conditions. of symmetry. This is of course no longer possible in the case of the invention for which the profits in refractive index must imperatively be made from the surface of a substrate (or wafer). In this case, the index profiles to be produced, which must be approximately parabolic in shape as a function of the distance ri to the axis in order to obtain the required imaging optical properties, require a series of original steps which constitute another aspect of the invention.

FIG. 15 represents a cylindrical element 120 capable of providing a lens with circular symmetry. For this type of element, the index profiles to be produced must be, at each cutting plane Pi, perpendicular to the axis of the lens and of the shape: Δm (Xi, y _± ) = Δnx (π ) = n _{0i *} sech (g_ri) #n _{oi *} (l-gi ² rι ^2/2 ) with π = (i ² + yχ ² ) ^{1 2} , g _± [8 (Δni (O) -Δι-i ( r _oi )) / r _oi ² ] ⁰ ' ⁵ and r ₀ ι is the radius of the lens.

In this formula the coordinate axes are Oi i, Oty and O-jZi (OiXj perpendicular to the surface plane of the wafer Oi i parallel to this plane of surface, the OiXY plane being perpendicular to the axis of the lens ÛZ).

The parameter gi depends on the variation in refractive index between the center and the edge of the lens and on the maximum radius r _oi of the lens. The value of g _± determines, among other things, the focal length f _A and the front face if (distance between the object focal point and the entry face of the lens) of the lens Li considered with: fi = l / n _{oi *} gi _* sin (gili)

l being the length of the supposed lens, to simplify the presentation and the mathematical formulas, to have plane input and output faces. It should however be remembered that the results of the invention apply to lenses whose input and output faces are for example convex.

Obtaining the desired Δni (ri) profiles at the level of the imaging elements is based on the following technological approach which comprises at least three stages.

A first step consists of an ion exchange through a mask comprising the appropriate geometric shapes as a function of the A ions and the B ions contained in the substrate. This ion exchange is for example obtained by tempering the glass plate provided with a mask of suitable geometric shape in a bath at temperature T i containing ions A and for a time ti by favorably applying an electric field between the surface of substrate on which is deposited the mask and the opposite face. This operation makes it possible to obtain a first refractive index profile over a depth di.

A second step consists in diffusing A ions back into the substrate by still favorably applying an electric field between the face of the substrate on which the mask is deposited and the opposite face. This ion exchange is for example obtained by dipping the wafer in the bath at temperature T ₂ containing B ions and for a time t ₂ by favorably applying an electric field between the face of the substrate on which the mask is deposited and the opposite face. This operation leads to the burial of the previous index profile over an average depth d ₂ . A third step consists of a thermal re-diffusion at a temperature T ₃ and for a time t ₃ without application of an electric field. According to the required specifications, this operation can be carried out in a bath containing B ions or simply in a controlled atmosphere.

FIG. 16 represents a form of refractive index profile close to the desirable profiles obtained with the following operating parameters:

- the ions A are silver ions A _g ⁺ , - the ions B are sodium ions N _a ⁺ ,

the temperature Ti is around 330 ° C and the time t _x is around 22 minutes,

- the temperature T ₂ is approximately 260 ° C and the time t ₂ is approximately 80 minutes, - the temperature T ₃ is approximately 330 ° C and the time t ₃ is approximately 1200 minutes. Δn represents the variation of normalized index.

The mask used can have a rectangular opening with a width of the order of & i = 15 μm and a length i corresponding to the desired functionality (collimation or finite distance imaging). The material used for local masking of ion exchanges can be of different types. Among the most common masks, there may be mentioned aluminum, aluminum oxide, silicon, titanium or nickel, but other materials could also be suitable.

Typically and taking into account the variation in maximum index Δni (, y) accessible by the manufacturing method described above comprised between 0.01 and 0.07 and the radii of achievable lenses comprised between 15 and 300 μm, the length l may vary from a few tens of micrometers to a few millimeters taking into account the focal lengths and the desired imaging architectures. The electric voltage V applied during the field exchanges can be approximately 325 volts.

Optionally, it may be necessary for certain configurations of optical elements to introduce an additional step between the first and the second step described above. This additional step consists of a thermal re-diffusion at a temperature T'i and for a time t'i. This re-diffusion is for example obtained by soaking the wafer in a bath at temperature T'i containing B ions and for a time t'i without application of an electric field.

To obtain a lens Li of length li and of radius r _oi such as that shown diagrammatically in FIG. 16, the mask used must have dimensions close to a rectangular opening of length 1 ′ and width ei.

The values of i and of course depend on the exchange and diffusion parameters chosen for the various manufacturing stages.

In the practical example given above we can see that for an opening ei of the order of

15 μm, the radius r ₀ i of the lens obtained is of the order of 110 μm. As for the length i, it can be deduced from the desired length li, taking into account the axial widening caused by ion exchange and thermal diffusions.

In practice, the i will be chosen less than l. It should however be borne in mind that if a rectangular shape of the mask is the simplest to be able to be implemented, slightly different geometries can have various advantages depending on the desired imaging performance. Figures 17 to 24 show different configurations of masks that can be used to implement the present invention.

Compensation for excessively marked edge effects can for example be obtained by geometric mask shapes such as that shown in FIG. 17. In this figure, the mask 300 is provided with a rectangular opening 301, two adjacent sides of the opening being connected by a fillet.

Likewise, the diffusion phenomena involved during the different stages will naturally lead to lenses having convex entry and exit faces. The associated curvature can then be increased or reduced, or even reversed by the use of masks of suitable geometric shape. Thus, FIG. 18 represents a mask 310 having a suitable opening 311 and FIG. 19 represents a mask 320 having a suitable opening 321.

Of course, intermediate shapes with inlet face and outlet face having different curvatures can be produced by combining the geometric shapes described. This is shown in Figure 20 which shows a mask 330 having an opening 331 of non-symmetrical shape.

Furthermore, the sides of the masks may not be straight as illustrated above in order to produce non-constant index profiles along the axis of the Z and to be able to modify the imaging properties of the components. By this method, it is in particular possible to modify the value of the aberrations of the optical component thus produced and take advantage of the flexibility offered by the simple modification of the geometric shape of the mask. Possible forms are illustrated by FIG. 21 where the mask 340 is provided with the opening 341 and by FIG. 22 where the mask 350 is provided with the opening 251. It is still possible to use segmented masks to obtain more brutal modifications of the index profiles and to take advantage of certain side effects. Figures 23 and 24 give possible examples of such geometries. FIG. 23 shows a mask 360 having a central main opening 361, lateral secondary openings 362 and 363 on the same side of the central opening and lateral secondary openings 364 and 365 on another side (opposite to the previous one) of the central opening. FIG. 24 shows a mask 370 having central openings 371 and 372, lateral secondary openings 373 and 374 on the same side with respect to the central openings and lateral secondary openings 375 and 376 on another side (opposite to the previous one) relative to the central openings.

Of course, in the case of an optical system comprising several lenses and in particular following the burial depths of each of the components, it may be necessary to separate the stages of production of the various elements of the system if the exchange parameters and are very different. In this case, the most extensive elements in depth will be produced first and in order of extension. The duration of exchanges and dissemination will then be calculated for each item produced, taking into account the previous and subsequent operations.

Claims

1. Photonic integrated circuit comprising a substrate (1) comprising at least one optical circuit (4) and means for interconnecting (2, 3) the optical circuit (4) with at least one optoelectronic component (6, 8, 9, 11, 12) attached to the substrate, the interconnection means consisting of at least one zone of said substrate, the refractive index of which is modified to ensure said interconnection, said zone of the substrate (2,3) comprising at least one gradient index lens.

2. Photonic integrated circuit according to claim 1, characterized in that said zone of the substrate (62) comprises at least two gradient index lenses (162, 262) whose axes are offset with respect to each other .

3. Photonic integrated circuit according to claim 1, characterized in that the optical circuit (4) is a circuit produced by an ion exchange technique.

4. Photonic integrated circuit according to any one of claims 1 to 3, characterized in that said substrate area (2, 3) is an area produced by an ion exchange technique.

5. Photonic integrated circuit according to claim 1, characterized in that the circuit optical (4) comprises an optical guide, one end of which is situated opposite said zone of the substrate in order to make a connection with said zone of the substrate.

6. Photonic integrated circuit according to claim 1, characterized in that said zone of the substrate is located in the substrate so as to be vis-à-vis an input or an output of the optoelectronic component in order to make a connection with said optoelectronic component.

7. Photonic integrated circuit according to claim 1, characterized in that the optoelectronic component (6, 8, 9, 11) is attached to the periphery of the substrate (1).

8. Photonic integrated circuit according to claim 1, characterized in that the optoelectronic component (12) is attached to the surface of the substrate (1).

9. Photonic integrated circuit according to claim 8, characterized in that the substrate (1) is provided with at least one cavity (5) allowing the housing of at least one optoelectronic component (12).

10. Photonic integrated circuit according to claim 8, characterized in that the substrate is provided with at least one cavity allowing the housing of at least one optical component.

11. A method of collective production of photonic integrated circuits as defined in claim 1, characterized in that it comprises the following steps:

- supply of a wafer (100) intended to supply as many substrates (1) as there are photonic integrated circuits to be produced,

- realization on the wafer (100) of the interconnection means (2, 3) for each photonic integrated circuit, realization on the wafer (100) of the optical circuit of each photonic integrated circuit, possibly, realization on the wafer of the housing means (5) at least one optoelectronic and / or optical component for each photonic integrated circuit,

- cutting of the wafer to obtain photonic integrated circuits.

12. The method of claim 11, characterized in that it further comprises a step of postponing the addition of at least one optoelectronic component and / or at least one optical component to complete the photonic integrated circuits.

13. Method according to claim 11, characterized in that the interconnection means are produced by means of a mask having a main opening of generally rectangular shape and adjacent sides connected by a fillet.

14. Method according to claim 13, characterized in that the opening has two opposite sides of convex shape and the other two opposite sides of concave shape.

15. The method of claim 13, characterized in that the opening has two opposite sides of convex and / or concave shape.

16. The method of claim 13, characterized in that the opening has all its sides of convex shape or concave shape.

17. Method according to any one of claims 13 to 16, characterized in that the mask also has at least one secondary opening located near at least one side of the main opening.