FR3160477A1 - Method for manufacturing a ferroelectric layer transferred onto a substrate and for polarization with improved homogeneity - Google Patents

Abstract

A method for manufacturing a structure comprising a negatively polarized ferroelectric layer (P1), comprising the steps of: providing a wafer of ferroelectric material (Ferrosub) having a polarization (-P1) oriented toward a face (Top); forming a weakening plane (Frgl) in the wafer; assembling the wafer to a support assembly (Sprt.Set); detaching a portion of the wafer so as to define a ferroelectric layer (Ferrolay) assembled to the support assembly (Sprt.Set) so as to obtain a structure (Struct), the ferroelectric layer having a negative polarization (P1); carrying out an additional full-field hydrogen implantation step, configured so as to correct or prevent the occurrence of polarization inversion in the volume of the ferroelectric layer and/or at its interface with the support assembly (Sprt.Ens); and applying at least a first heat treatment to the structure (Strct). Figure to be published with the abstract: Fig. 2

Description

Method for manufacturing a ferroelectric layer transferred onto a substrate and for polarization with improved homogeneity TECHNICAL FIELD OF THE INVENTION

The invention relates to a method of manufacturing a ferroelectric layer transferred onto a substrate and polarization with improved homogeneity in its thickness. Such a structure can be used to form, for example, radiofrequency (RF) components, in particular elastic wave components, in particular surface acoustic wave (SAW) or volume acoustic wave (BAW) components.

TECHNOLOGICAL BACKGROUND

There are several types of applications that take advantage of or are influenced by the polarization properties of ferroelectric materials and the presence of polarization domains opposite each other. These include electroacoustic devices such as surface acoustic wave devices (SAW devices) or bulk acoustic wave devices (BAW devices). The existence of these applications has motivated the development of methods for controlling the polarization domains of ferroelectric layers.

The article “Seeing Is Believing – In-Depth Analysis by Co-Imaging of Periodically-Poled X-Cut Lithium Niobate Thin Films” by Sven Reitzig et al., published in Crystals 2021, 11, 288, describes a lithium niobate layer embedded on a silicon substrate via a silica layer, and the control of the polarization of this layer parallel to the plane in which it extends, by applying a voltage between electrodes arranged periodically on the free surface of the layer. It should be noted that, in such a structure with electrodes only on the free face of the crystal, controlling a polarization that would be perpendicular to the substrate by means of an electric field, in the absence of a buried electrode, would require voltages that risk causing the breakdown of the silica layer.

Document EP 0 592 226 A1 describes an optical frequency conversion device obtained by the periodic juxtaposition of parallel bands of inverted polarization at one face of a ferroelectric substrate having a spontaneous polarization perpendicular to the plane of extension of the substrate, i.e. perpendicular to this face. The inversion of the polarization according to the bands is obtained by carrying out a proton exchange through a mask.

Document WO 2005/052682 A1 describes the localized reversal of the polarization of a ferroelectric crystal with spontaneous polarization perpendicular to one of its faces, by application of an electric field along juxtaposed periodic bands, by means of gel electrodes arranged on this face and the opposite face of the crystal.

The structures and methods presented above, if they effectively allow the control of the polarization domains of a ferroelectric layer on its surface, remain impractical and do not allow the control of the polarization in the thickness of this layer or its integration on a support.

In response to these shortcomings, the international patent application with publication number WO 2020/200986 A1 and the patent applications in France with filing numbers FR2301220 and FR2301221 propose techniques for transferring ferroelectric layers onto a substrate. However, these techniques do not solve the problem in a completely satisfactory manner, in that, at the end of manufacturing, the transferred ferroelectric layers may exhibit uncontrolled polarization inhomogeneities in the thickness of the layer, in particular with regard to the transfer of ferroelectric layers having a negative polarization onto the final support substrate.

An object of the invention is to provide a method for manufacturing a ferroelectric layer transferred onto a substrate, the method correcting or preventing any polarization inhomogeneities which could appear in their thickness, in particular for the situation where it is sought to obtain a ferroelectric layer of negative polarization, that is to say oriented towards the support of the ferroelectric layer.

To achieve these objects, one aspect of the invention is a method for manufacturing a structure comprising a negatively polarized ferroelectric layer, comprising the steps of: providing a plate of ferroelectric material having a first face and having a polarization oriented towards this face; forming a weakening plane in the plate of ferroelectric material by implanting hydrogen ions through the first face; assembling the plate of ferroelectric material having the weakening plane to a support assembly by bringing the first face into contact with a free surface of the support assembly; detaching a portion of the plate of ferroelectric material so as to define a ferroelectric layer detached from the plate of ferroelectric material and assembled to the support assembly so as to obtain a structure formed from the ferroelectric layer and the support assembly, the ferroelectric layer having a negative polarization; carrying out an additional full-field hydrogen implantation step, configured so as to correct or prevent the occurrence of polarization inversion in the volume of the ferroelectric layer and/or at its interface with the support assembly; and applying at least a first heat treatment to the structure.

An advantage of the method according to the invention is its ability to produce a structure including a ferroelectric layer transferred onto a substrate and with a negative polarization, oriented towards the interface between the ferroelectric layer and its support, presenting improved homogeneity over its thickness compared to known manufacturing methods. Such a structure facilitates, for example, the integration of volume acoustic functions while maintaining a low cost and allowing a high level of performance.

According to additional non-limiting characteristics of the invention, considered individually or in any technically feasible combination:

- the additional hydrogen implantation step can be carried out after the at least one first heat treatment, the method further comprising a step of applying a second heat treatment to the structure after the additional hydrogen implantation step;

- the method may further comprise a step of thinning the ferroelectric layer, preferably between the step of applying the first heat treatment and the step of applying the second heat treatment;

- the additional hydrogen implantation step can be carried out before at least one first heat treatment;

- the method may further comprise a step of thinning the ferroelectric layer, preferably after the step of applying the first heat treatment;

- the method may comprise a step of characterizing a reference structure, an adjustment of parameters of the additional hydrogen implantation step being able to be defined in response to this characterization;

- the step of additional implantation of hydrogen ions may comprise a first correction ion implantation in the ferroelectric layer, parameterized so as to generate a projected path of the implanted hydrogen ions such that a depth of its maximum reaches or exceeds a depth of interface between (i) a volume of stable polarization of the ferroelectric layer and (ii) a volume of reversed polarization due to the formation of the embrittlement plane or polarization which would reverse due to the formation of the embrittlement plane if a heat treatment were applied;

- the additional hydrogen ion implantation step may comprise a second correction ion implantation, parameterized so as to generate a projected path of the implanted hydrogen ions such that a depth of its maximum reaches or exceeds the depth of the interface between (i) the ferroelectric layer and (ii) the support assembly, preferably at a distance to this interface which is less than the thickness of a polarization volume which is limited by this interface and which is multidomain or inverted with respect to a polarization of a stable polarization volume of the ferroelectric layer or which would invert or become multidomain if a heat treatment were applied;

- the additional full-field hydrogen implantation step can be configured to obtain a hydrogen concentration of between 10 ¹⁹ and 10 ²² at/cm ³ in a volume of the structure;

- the ferroelectric layer can extend in an extension plane and the polarization makes an angle with included in an angular range from -20° to -160° relative to the extension plane;

- the ferroelectric layer can be a layer of lithium niobate or lithium tantalate.

BRIEF DESCRIPTION OF THE FIGURES

Other characteristics and advantages of the invention will emerge from the detailed description of the invention which follows with reference to the appended figures in which:

FIG. 1 There FIG. 1 illustrates crystal sections of single crystals;

FIG. 2 There FIG. 2 represents the manufacture of a structure comprising a ferroelectric layer transferred onto a support;

FIG. 3 There FIG. 3 illustrates the influence of the manufacture of the FIG. 2 on the polarization of the reported ferroelectric layer;

FIG. 4 There FIG. 4 illustrates a polarization reversal mechanism in a layer of ferroelectric material;

FIG. 5 There FIG. 5 illustrates a method of correcting the polarization of a ferroelectric material by means of a first ion implantation;

FIG. 6 There FIG. 6 illustrates a method of correcting the polarization of a ferroelectric material by means of a second ion implantation;

FIG. 7 There FIG. 7 illustrates a structure comprising a polarization-deferred ferroelectric layer with improved homogeneity;

FIG. 8 There FIG. 8 illustrates a first manufacturing method according to the invention;

FIG. 9 There FIG. 9 illustrates a negative orientation polarization; and

FIG. 10 There FIG. 10 illustrates a second manufacturing method according to the invention.

DETAILED DESCRIPTION OF THE INVENTION First embodiment of the invention

Depending on the piezoelectric device to be produced, it is necessary to choose an orientation for the polarization of the active layer of piezoelectric material, which is made up of a layer of monocrystalline ferroelectric material.

This ferroelectric material can be obtained by crystal growth using the so-called "Czochralski" method, which makes it possible to obtain a solid monocrystalline material in the form of an elongated cylinder called a "ball" or "ingot" by pulling a crystal seed of chosen orientation immersed in a molten material corresponding to the crystal to be formed, in a pulling direction. The monocrystal grows in this pulling direction. Platelets of the ferroelectric material are then obtained by cutting the ingot at a determined angle relative to its growth direction, then prepared and integrated as a ferroelectric layer in, for example, a piezoelectric device.

The orientation of the polarization of the layer depends on the crystalline orientation of the cut wafer, which does not correspond, in the general case, to the direction of growth of the single crystal. Of course, the skilled person knows how to choose the most appropriate seed close to the desired crystalline orientation for the drawing of the ingot in order to reduce material losses during cutting and machining into circular wafers.

Figure 1 illustrates in (A) and (B) the orientations of LiTaO ₃ layers designated by 42RY and -138RY, respectively. These names derive from the identification of the crystallographic orientation of these layers, corresponding to respective rotations of 42° and -138° of the y and z axes with respect to the counterclockwise direction around the x axis to give the x', y' and z' axes, the x, y and z axes corresponding to the crystallographic axes proper to the single crystal and the y' axis being aligned with the direction normal to the cut and machined surface. More specifically, the x, y and z axes correspond to the axes , , And , respectively.

After pulling in a chosen direction, it is known to those skilled in the art to carry out polarization by applying an electric field in the z direction when the temperature falls below the Curie temperature to determine the direction of polarization.

Conventionally, a so-called negative polarization is a polarization of orientation directed towards the interior of the material or structure concerned, or in other words the projection of the z axis onto the y' axis is negative.

Conventionally, for a ferroelectric layer fixed to a substrate, a layer whose polarization orientation is directed towards the free surface of the layer will be designated as a positive polarization layer.

Conversely, still for a ferroelectric layer fixed to a substrate, we will designate as a negative polarization layer a layer whose polarization orientation is directed inwards, that is to say towards the interface between the layer and its support.

Figure (2) illustrates in (D) a situation in which the polarization P1 of a ferroelectric layer Ferro _lay is negative: it is oriented towards the interface between the ferroelectric layer P1 and the assembly supporting it, the latter here consisting of an intermediate layer Int on a support layer Sprt. We can also say that the component of the polarization in the vertical direction Vert perpendicular to the ferroelectric layer and oriented from the support assembly towards the ferroelectric layer, is negative. This component is obtained by a normal projection of the polarization on the direction Vert.

To obtain a piezoelectric device equipped, for example, with a negatively polarized ferroelectric layer, it is necessary to choose in advance the characteristics of the ferroelectric material wafer used to manufacture the device, and to take into account the manufacturing process and its potential impacts on the polarization of the wafer.

There FIG. 2 illustrates the general principle of a method for manufacturing the structure Struct illustrated in (D), and which includes a step of transferring a ferroelectric layer Ferro _lay onto a support assembly Sprt.Set formed of the intermediate layer Int on the support layer Sprt, the intermediate layer being interposed between the ferroelectric layer Ferro _lay and the support layer. In a non-limiting manner, it is also possible to find a layer of another material, in particular a layer of silicon oxide, on the face of the ferroelectric layer Ferro _lay before arranging it on the intermediate layer Int. The structure Struct is for example designed to be integrated into an electroacoustic device.

The characteristics of the support assembly can affect the acoustic properties of the Ferro _lay layer, which is important in the case of a structure forming part of an acoustic device. The nature and thickness of these layers can therefore also be decisive in achieving the desired processing of an electrical signal, or at least influencing this processing. In the example shown in the FIG. 2 , the support assembly may comprise the support layer Sprt, and the intermediate layer Int of dielectric assembly, preferably comprising an oxide, directly in contact with the ferroelectric thin layer Ferro _lay .

For reasons of availability and cost, the Sprt support layer may be made of silicon. It may be a support consisting of a solid base substrate of monocrystalline silicon, but the invention is not limited to this support which may, more generally, be made of any material, for example silicon, even electrically insulating such as sapphire or glass. The Sprt support layer, when it is formed of a solid substrate, typically has a thickness of several hundred microns. This can limit the density of electrical charges, holes or electrons, which are likely to move in the support layer and which could affect the proper functioning of an RF component which would be formed on the basis of the Struct structure. The Sprt support layer may thus be made of a high-resistivity silicon substrate, i.e. having a resistivity greater than 1000 ohm-centimeter, and more preferably still greater than 3000 ohm-centimeter. To support the resistive nature of the Sprt support layer, it is possible to provide the Sprt support layer with a charge trapping layer on the side of the ferroelectric layer, for example formed from polycrystalline silicon. It is of course possible to provide this charge trapping layer by a technique other than that providing a layer formed from polycrystalline silicon. This layer may also comprise carbon or be made of or comprise silicon carbide or an alloy of silicon and carbon. Alternatively, it may involve producing the electric traps by ion bombardment of relatively heavy species (for example argon) in a surface part of the support layer in order to create crystalline defects capable of trapping electric charges. It is also possible to provide a charge trapping layer formed from a porous material, for example by porosification of a surface part of the Sprt support layer when the latter is made of silicon. But the invention is not limited to a support layer having such characteristics.

For example, the intermediate layer Int may be made of silicon oxide, silicon nitride, or be formed from a stack of layers composed of these materials.

Alternatively, the intermediate layer may be an electrically conductive metal layer interposed between the ferroelectric layer Ferro _lay and the support layer Sprt. In this example, the metal layer Int is in direct contact with each of the support Sprt and the ferroelectric layer Ferro _lay . The layer Int may in this case subsequently be used as a buried electrode intended to apply an electric field to the ferroelectric layer Ferro _lay .

The presence of an Int intermediate layer is only an option. Not all applications require the presence of such a layer.

Conventionally, the Struct structure may be in the form of a circular plate whose diameter may be 100, 200, 300 or even 450 mm, but the invention is in no way limited to these dimensions or this shape.

The ferroelectric layer Ferro _lay may be made of a single-crystal ferroelectric material, such as lithium tantalate LiTaO ₃ or lithium niobate LiNbO3, or materials such as BaTiO ₃ , PbZrTiO ₃ , KNbO ₃ , BaZrO ₃ , PbTiO ₃ or KTaO ₃ . These materials also have piezoelectric properties. Generally, the ferroelectric layer may have a thickness of between 10 nanometers and 10 microns, depending on the intended application of the structure Struct and the expected performance of the components, but the invention does not exclude the use of different thicknesses, always depending on the intended application. It is recalled that a ferroelectric material is a material that has an electric polarization in its natural state, a polarization that can be reversed by applying an external electric field greater than the coercive field of the material. As illustrated in this document, the ferroelectric layer preferably has a single-domain polarization, that is, all the dipole moments are aligned parallel to each other in a given direction. Here, the given direction is inclined relative to the plane of the ferroelectric layer, that is, inclined relative to the free face of this layer.

• une étape S00 de choix et de fourniture d’une plaquette de matériau ferroélectrique monocristallin pour sa composition chimique et son orientation cristalline par rapport à son plan d’extension, afin qu’elle serve en tant que substrat donneur ferroélectrique Ferro_sub, la plaquette choisie présente une polarisation positive monodomaine -P1, c’est à dire ayant une composante orientée dans la direction de sa face désignée par Top, c’est-à-dire que la plaquette a une orientation cristalline comprise entre 0RY et 180RY, comme illustré en (B) de laFIG. 2;
• une étape S10 de préparation d’un ensemble de support Sprt.Ens, constitué ici d’une couche de support Sprt munie d’une couche intermédiaire Int à sa surface comme illustrée en (A) de laFIG. 2;
• une étape S20 de préparation du substrat donneur ferroélectrique Ferro_subafin d’y former un plan de fragilisation Frgl en vue de la séparation d’une couche ferroélectrique Ferro_laydu substrat donneur Ferro_subà une étape ultérieure, comme illustré en (B), la profondeur du plan de fragilisation Fgrl dans le substrat donneur Ferro_subdéfinissant l’épaisseur de la couche ferroélectrique Ferro_lay;
• une étape S40 d’assemblage de l’ensemble de support Sprt.Ens avec le substrat donneur Ferrosub, par mise en contact de la face libre Fr.Fac de la couche intermédiaire Int avec la face Top du substrat donneur ferroélectrique Ferro_subafin de constituer une structure intermédiaire Struct_interillustrée en (C), avec la couche intermédiaire Int interposée entre la couche de support Sprt et la face Top du donneur du substrat ferroélectrique Ferro_sub, l’orientation de la polarité de la couche Ferrolay par rapport à la couche support est inversée, passant de -P1 à P1 du fait du retournement du substrat donneur Ferro_subpour l’assemblage ; et
• une étape de S60 de détachement d’une partie du substrat donneur Ferro_subde la structure intermédiaire au niveau du plan de fragilisation, laissant la couche ferroélectrique Ferro_layfixée sur le support Sprt et permettant d’obtenir la structure Struct illustrée en (D), la couche ferroélectrique Ferro_layayant comme déjà dit une polarisation négative par rapport à la normale Vert de la surface libre ayant servie comme interface de détachement.

With reference to figures 2 to 9, the structure Struct can be produced by a manufacturing method 100 by transferring a ferroelectric layer onto a support assembly, the method being summarized by the FIG. 8 and including:

• a step S00 of choosing and providing a wafer of monocrystalline ferroelectric material for its chemical composition and its crystalline orientation with respect to its extension plane, so that it serves as a ferroelectric donor substrate Ferro _sub , the chosen wafer has a monodomain positive polarization -P1, that is to say having a component oriented in the direction of its face designated by Top, that is to say that the wafer has a crystalline orientation between 0RY and 180RY, as illustrated in (B) of the FIG. 2 ;
• a step S10 of preparing a support assembly Sprt.Ens, consisting here of a support layer Sprt provided with an intermediate layer Int on its surface as illustrated in (A) of the FIG. 2 ;
• a step S20 of preparing the ferroelectric donor substrate Ferro _sub in order to form therein a weakening plane Frgl for the separation of a ferroelectric layer Ferro _lay from the donor substrate Ferro _sub in a subsequent step, as illustrated in (B), the depth of the weakening plane Fgrl in the donor substrate Ferro _sub defining the thickness of the ferroelectric layer Ferro _lay ;
• a step S40 of assembling the support assembly Sprt.Ens with the donor substrate Ferrosub, by bringing the free face Fr.Fac of the intermediate layer Int into contact with the Top face of the ferroelectric donor substrate Ferro _sub in order to constitute an intermediate structure Struct _inter illustrated in (C), with the intermediate layer Int interposed between the support layer Sprt and the Top face of the donor of the ferroelectric substrate Ferro _sub , the orientation of the polarity of the Ferrolay layer relative to the support layer is reversed, passing from -P1 to P1 due to the turning over of the donor substrate Ferro _sub for the assembly; and
• a step of S60 of detaching a part of the Ferro _sub donor substrate from the intermediate structure at the level of the weakening plane, leaving the Ferro _lay ferroelectric layer fixed on the Sprt support and making it possible to obtain the Struct structure illustrated in (D), the Ferro _lay ferroelectric layer having as already said a negative polarization with respect to the Green normal of the free surface having served as a detachment interface.

The Ferro _sub donor substrate illustrated in (B) is a substrate made of the ferroelectric material of the Ferro _lay ferroelectric layer. It could alternatively comprise a surface thickness of this material. Thus, the donor substrate may, for example, be formed of a bulk substrate of lithium tantalate or lithium niobate, or of a composite substrate formed of a first substrate on which rests a thickness (at least equal to that of the Ferro _lay layer) of lithium tantalate or lithium niobate. The use of a composite substrate comprising a support substrate and a layer of ferroelectric material proves necessary if the difference in the coefficient of thermal expansion of the ferroelectric material and the final substrate is too great to allow the application of the Smart Cut process. This approach is described in detail in documents WO2019002080 and WO2019186032 which are incorporated by reference.

The donor substrate therefore comprises at least one layer of ferroelectric material having a first positive polarization -P1 monodomain inclined relative to the extension plane of the ferroelectric layer Ferro _lay which depends on the chosen crystalline orientation of the wafer. This layer is intended to be fixed to the intermediate layer Int before being separated from the donor substrate. In a non-limiting manner, it is also possible to find a layer of another material, in particular a layer of silicon oxide, on the face of the ferroelectric layer Ferro _lay before arranging it on the intermediate layer Int.

There FIG. 2 illustrates a situation in which the ferroelectric layer comes from a LiTaO ₃ section of the 42RY type, with the orientation shown in (A) of the FIG. 1 . The polarization of the layer is oriented upwards in (B) of the FIG. 2 as in (A) of the FIG. 1 .

More generally, the method 100 applies to a crystalline section in the range extending from 0RY to 180RY, preferably 20RY to 160RY, for the Ferro _sub donor substrate. The polarization of the final structure is therefore negative, i.e. oriented from 0° to -180°, preferably in an angular range [Ang] going from -20° to -160°, relative to the plane of the ferroelectric layer Ferro _lay , as illustrated by the FIG. 9 .

The ferroelectric layer Ferro _lay can be transferred from the ferroelectric donor substrate Ferro _sub by implementing the Smart Cut ^TM technology, in which case the donor substrate must be prepared by introducing light species such as hydrogen and/or helium into this donor substrate. This introduction may correspond to a hydrogen implantation, i.e., an ion bombardment of hydrogen of the flat face Top of the donor substrate Ferro _sub . In a manner known per se, and as illustrated in (B), the implanted hydrogen ions H ⁺ are intended to form a weakening plane Frgl delimiting the ferroelectric layer Ferro _lay of ferroelectric material to be transferred which is located on the side of the face Top and another part Ferro _sep forming the rest of the substrate and which will be separated from the ferroelectric layer Ferro _lay at a later stage.

The nature, the dose of the implanted species and the implantation energy are chosen according to the thickness of the layer that we wish to transfer and the physicochemical properties of the Ferro _sub donor substrate. In the case of a LiTaO ₃ donor substrate, we can choose to implant a dose of hydrogen between 10 ¹⁶ and 5.10 ¹⁷ at/cm² with an energy between 30 and 300 keV to delimit a ferroelectric Ferro _lay layer of the order of 200 to 2000 nm thick.

In (C) of the FIG. 2 , the polarization P1 is oriented towards the interface between the ferroelectric layer Ferro _lay and the intermediate layer Int, so the polarization of the ferroelectric layer is considered negative.

The Sprt support substrate may have the same size and shape as the Ferro _sub donor substrate, but the invention is not limited to such a configuration and different sizes, shapes and configurations may be used. Prior to assembly, it may be envisaged to prepare the faces of the substrates to be assembled by a cleaning, brushing, drying, polishing, or plasma activation step, or before or after cleaning a deposit of a bonding layer.

The assembly may correspond to the intimate contact of the Ferro _sub donor substrate with the Sprt support by molecular adhesion and/or electrostatic bonding.

As is well known, during a molecular adhesion process, the exposed surfaces of the Sprt support and the Ferro _sub donor substrate, which are perfectly clean, flat and smooth, are brought into intimate contact to promote electrostatic bonding or the development of molecular bonds, for example of the van der Waals or covalent type. The assembly of the two bodies is then obtained without the use of an adhesive.

The assembly may include the application of a low temperature heat treatment (e.g., 50 to 300°C, typically 100°C) to cure crystalline defects in the ferroelectric layer and to sufficiently enhance the bonding energy to allow a possible subsequent thinning step.

In the present embodiment, the step of detaching a portion of the donor substrate is carried out by applying Smart Cut™ technology, according to which a layer intended to form the ferroelectric layer Ferro _lay is delimited by the embrittlement plane Frgl defined by implantation of hydrogen ions in the donor substrate, as illustrated in (B) of the FIG. 2 . After the assembly step, this layer is detached from the donor substrate by fracture at the level of the weakening plane Frgl and thus transferred to the support Sprt, as illustrated in (D) of the FIG. 2 .

This detachment step may thus comprise the application to the intermediate structure Struct _inter of a heat treatment in a temperature range of the order of 80°C to 300°C to enable the detachment of the part of the donor substrate from the ferroelectric layer Ferrol _ay and thus complete the transfer of the latter onto the support assembly. As a replacement or in addition to the heat treatment, this step may comprise the application of a blade or a jet of gaseous or liquid fluid, or any other force of a mechanical nature at the level of the embrittlement plane Frgl.

Following step S60 which led to obtaining the structure illustrated in (D) of the FIG. 2 by separating the Ferro _lay layer from the rest of the Ferr _sub donor substrate, a stabilizing heat treatment H.Treat-1 is applied to the Struct structure, at a step S80. The stabilizing heat treatment makes it possible to cure crystalline defects present in the ferroelectric layer and contributes to consolidating the bonding between this Ferro _lay ferroelectric layer and the intermediate Int layer. In the case of LiTaO3, this heat treatment is intended to bring the ferroelectric layer to a temperature between 300°C and the Curie temperature of the ferroelectric material (and preferably greater than or equal to 450°C, 500° or 550°, up to 600°) for a period of between 5 minutes and 10 hours. This heat treatment is preferably carried out by exposing the free face of the dielectric layer to an oxidizing or neutral gas atmosphere.

However, the inventors found that the H.Treat-1 heat treatment, combined with the previous manufacturing steps and the Struct structure itself, has several effects locally modifying the polarization of the Ferro _lay layer, as explained using the FIG. 3 .

One of these effects is particularly troublesome, modifying the polarization of the core of the Ferro _lay layer, that is to say a part of the layer located at a distance from its surface or its interface with the support assembly. Indeed, the presence of a hydrogen ion concentration gradient in a ferroelectric layer combined with a heat treatment at a temperature of the order of 300° to 600°C, such as the H.Treat-1 treatment, causes an inversion of the polarization of the ferroelectric material in the volume where the gradient is sufficiently strong. It is interpreted that the gradient of implanted H+ ions causes the appearance of an electric field E _H which, thanks to the thermal activation caused by the increase in temperature, can induce the inversion of the polarization in a volume of the material. More specifically, during heat treatment at sufficiently high temperature, the polarization of the ferroelectric material tends to take the same orientation as the E _H field, thus reversing the polarization when E _H is oriented in a direction opposite to that of the material before application of the heat treatment.

There FIG. 3 illustrates in four lines Lin.A to Lin.D the impact of the heat treatment H.Treat-1 of step S80 on the polarization in the volume of the Ferro _lay layer, in connection with certain manufacturing steps. Each line corresponds to a manufacturing step and is divided into two columns. A first column Col.1 illustrates the polarization of the Ferro _lay layer in the absence of the heat treatment and corresponds to the manufacturing steps of the FIG. 2 , while a second column Col.2 illustrates the effects caused by these different steps during the application of the heat treatment of step S80. Particular attention will be paid to the appearance and evolution of volumes of distinct polarizations appearing in the Ferro _lay layer.

Line Lin.A illustrates the ferroelectric donor substrate Ferro _sub at step S00. Its polarization is indicated by -P1 because it is opposite in direction to the polarization P1 of the final Struct structure to be obtained, illustrated in (D) of the FIG. 2 . This polarization is homogeneous in the whole volume V _B which is defined, at this moment, by the volume of the whole donor substrate Ferro _sub . The two columns Col.1 and Col.2 do not present any difference: in the absence of hydrogen ions implanted according to an inhomogeneous profile in the volume V _B of Ferro _sub , the thermal treatment does not lead to a reversal of polarization.

Line Lin.B illustrates the state of the Ferro _sub donor substrate after the formation of the Frgl embrittlement plane at step S20. Without the heat treatment (see Col.1), the polarization of the donor substrate is not modified. On the other hand, following the heat treatment, or if the heat treatment were applied immediately without proceeding to steps S40 and S60, the polarization would be reversed in a volume of material V _C from the material volume V _B , in the vicinity of the Frgl embrittlement plane. The polarization of the volume V _C then corresponds to that of a -138RY section of ferroelectric material LiTaO ₃ , as indicated by the number -138 in the figure. Outside the volume Vc, the polarization remains unchanged.

This mechanism is illustrated by the FIG. 4 , which shows the distribution of hydrogen ions implanted in the Ferro _lay layer in the form of a hydrogen concentration [H] according to the implantation depth Dpth from the surface of the Ferro _lay layer (representation also referred to as "projected path" in the field of ion implantation). Such a distribution, highly inhomogeneous according to the depth of the Ferro _lay layer, generates the electric field E _H , likely to cause the polarization inversion of the volume concerned when a heat treatment at a sufficiently high temperature is applied. Indeed, the polarization tends under these conditions to take the same orientation as the electric field E _H . Here, a part of the volume V _B , influenced by the concentration gradient of the hydrogen ions and the electric field that it generates, has seen its polarization reverse, passing P1 to -P1 in the volume V _C .

Line Lin.C illustrates the state of the donor substrate following its assembly to the support assembly at step S40. To do this, the donor substrate is turned over. By symmetry, we find ourselves in the situation that we would have if we had used a -138RY cut of ferroelectric material LiTaO3 as the _Ferrosub substrate without the turning over during assembly having been carried out. In practice, the polarizations of the two volumes V _B and V _C are found to be reversed, as indicated by the polarization arrows P1 and -P1 and by the indications 42 and -138 corresponding to the 42RY and -138RY cuts.

Line Lin.D illustrates, in addition to the presence of the two volumes V _B and V _C already described, the presence of two other volumes V _A and V _D of reversed polarizations, originating respectively from the volume V _B and V _C following the heat treatment of step S80. These two volumes appear on both faces of the Ferrolay layer. Volume V _D is defined by a part of volume V _C as defined in line Lin.C, the polarization of which is partially or totally reversed at the free surface of the Ferro _lay layer. Volume V _A is defined by a part of volume V _B as defined in line Lin.C, the polarization of which is partially or totally reversed at the interface of the Ferro _lay layer with the support assembly, and therefore with the Int layer of this example. These polarity inversions occur due to surface and interface physicochemical interactions related to the applied heat treatment temperature, and cause the transformation of surface portions of this ferroelectric layer, which passes from a single-domain ferroelectric structure of a first polarization to a ferroelectric structure of a second polarization opposite to the first polarization or to a multi-domain ferroelectric structure at its free surface and/or at its interface with the support assembly. These surface portions are said to be multi-domain when they have a ferroelectric structure having zones of various polarization orientations. These surface portions typically have thicknesses of the order of 150 nm or less, and can develop over the entire extent of the ferroelectric layer.

We see that the manufacturing process of the Struct structure must preferably take into account the formation of the volumes V _A , V _C and V _D from the volume V _B , to improve the homogeneity of the polarization of the ferroelectric layer Ferro _lay , inhomogeneous in depth, and therefore leading to poorly controlled characteristics for the devices integrating this layer. The part of the volume V _B not undergoing any polarization inversion is considered as a stable polarization volume of the ferroelectric layer Ferro _lay .

The V _D volume can be simply removed by surface polishing of the Ferro _lay layer.

In a step S100 following step S80 during which the heat treatment is applied, the ferroelectric layer is thinned. This thinning may correspond to the polishing of the free face of the ferroelectric layer Ferro _lay , for example by mechanical, chemical-mechanical and/or chemical etching thinning techniques. It makes it possible to prepare the free face so that it has a low roughness, for example less than 0.5nm RMS 5x5 µm by atomic force measurement (AFM) and to remove the volume V _D of the ferroelectric layer Ferro _lay . A removal of 50 to 300 nm of thickness is generally provided to achieve the target thickness of the ferroelectric layer Ferro _lay , and in all cases a thickness greater than that of the volume V _D . This improves the homogeneity of the polarization of the Ferro _lay layer.

In order to improve the homogeneity of the polarization of the Ferro _lay layer in depth, it is proposed to carry out, after step S100, a step S120 of implantation Corr.Imp of correction hydrogen ions in the thickness of the structure Struct followed by a second heat treatment H.Treat-2 during a step S140, so as to correct the polarization of the volumes V _A and V _C . Step S120 may optionally comprise several ion implantation phases, but comprises at least one. Of course, for other embodiments, and without this being limiting for the invention, it is also possible to carry out step S120 of implantation Corr.Imp of correction hydrogen ions before said thinning step S100.

ThereFIG. 5illustrates a first correction ion implantation leading to a projected path R_Pimplanted hydrogen ions such that depth D1(R_P) of its maximum reaches or exceeds the depth of the interface between the volumes V_Band V_Cand is located at this interface or close to the interface either in volume V_Ceither in volume V_B, preferably at a predetermined distance such that the hydrogen concentration in the volume V_Cexceeds a threshold of 10¹⁹hydrogen atoms per cubic centimeter, for example with a hydrogen concentration between 10¹⁹and 10²²atoms/cm³ at the level of a plane parallel to the free surface of the Ferro layer_layThis first corrective ion implantation will have an effect counteracting that of the ion implantation which formed the weakening plane and will therefore correct the polarization of volume V_Cby reversing its polarization again and reorienting it like that of volume V_B.

ThereFIG. 6illustrates a second corrective ion implantation leading to a projected path R_Pimplanted hydrogen ions such that depth D2(R_P) of its maximum reaches or exceeds the depth of the interface between the volumes V_HASand the Int layer, i.e. the interface between the ferroelectric layer Ferrolay and the support assembly Sprt.Ens, and is located at this interface or in the Int layer, preferably at a predetermined distance such that the hydrogen concentration in the volume V_HASexceeds a threshold of 10¹⁹hydrogen atoms per cubic centimeter, for example with a hydrogen concentration between 10¹⁹and 10²²atoms/cm³ at the level of a plane parallel to the free surface of the Ferro layer_layThis second corrective ion implantation, combined with the second thermal treatment H.Treat-2, will have an effect counteracting that of the physicochemical effects occurring at the interface between V_HASand Int, and will therefore correct the polarization of the volume V_HASby reversing its polarization again and reorienting it like that of volume V_B.

The first implantation and the second implantation are carried out "full field", that is to say on the entire surface of the Ferro _lay layer, without a mask intended to create implantation patterns in a useful area of the layer. A holding system can however be used to hold the structure during operation and mask certain parts of the Ferro _lay layer, in particular the periphery, these masked parts not being considered as part of the useful area of the Struct structure. Implantation energies between 30 keV and 300 keV, and doses between 1.10 ¹⁵ and 9.10 ¹⁶ atoms/cm ² can be considered.

Depending on the characteristics of the ferroelectric layer Ferro _lay , and in particular its thickness, the second implantation may be sufficient to correct the orientations of the polarizations of the two volumes V _A and V _C . Indeed, the electric field generated by the hydrogen concentration gradient can be sufficiently extensive and intense to include the volume V _C in its zone of influence and influence the polarization during the heat treatment H.Treat-2. It is then understood that a relatively thin ferroelectric layer will benefit more easily from this advantage than a relatively thick ferroelectric layer.

The first correction implantation and the second correction implantation may be performed successively during the correction step S120. Alternatively, the first implantation or the second implantation may be performed alone.

An adjustment of the implantation parameters of the correction step S120 is preferably carried out according to the geometric characteristics (thickness, location) of the volumes V _A and V _C of the Ferro _lay layer. For this purpose, a first preparatory step Prep-S00 consists of manufacturing a reference structure Ref according to the same parameters as the structure Struct that we wish to produce, therefore according to steps S00 to S100.

A second preparatory step Prep-S10 consists of carrying out a polarization analysis along a transverse plane of the Ferro _lay layer of the reference sample, in order to obtain the Char characteristics of the polarization of the Ref sample. To do this, a bevel cut of the Ferro _lay layer is carried out and then an analysis of this cut by piezoelectric force microscopy (or PFM in English terminology, for Piezoresponse Force Microscopy). Such an analysis makes it possible to determine the orientation of the polarization of the Ferro _lay layer over its entire thickness, and therefore to determine the absence or presence of volumes of reversed polarizations and, if necessary, to locate the volumes V _A , V _B , V _C and V _D , that is to say the thicknesses of each of these volumes and therefore the depths at which the interfaces between V _A and V _B , V _B and V _C , and V _C and V _D are located, measured from the free surface of the Ferro _lay layer.

Step S120 may be applied with implantation parameters adjusted in response to the analysis performed in step Prep-S10. Specifically, the implantation energy of the first ion implantation will be determined such that the depth D1(R _P ) of the maximum of the projected path R _P of the hydrogen ions reaches or is located close to the depth of the interface between volumes V _B and V _C or is located at this interface or even in volume V _c or in volume V _B , preferably at a predetermined distance already mentioned above to exceed the concentration threshold. The same principle applies for the depth D2(R _P ) of the second implantation, whose implantation energy will be determined in such a way that the depth D2(R _P ) of the maximum of the projected path R _P of the hydrogen ions reaches or is located close to the depth of the interface between the volume V _A and the layer Int, or is located at the level of this interface, in the volume V _A or even in the layer Int.

The second heat treatment H.Treat-2 of step S140, applied to the structure Struct after the correction implantation step S120 Corr.Imp, may be provided to bring the ferroelectric layer to a temperature between 300°C and the Curie temperature of the ferroelectric material (and preferably greater than or equal to 450°C, 500° or 550°, up to 600°) for a duration between 5 minutes and 10 hours. This heat treatment is preferably carried out by exposing the free face of the dielectric layer to an oxidizing or neutral gas atmosphere. H.Treat-2 leads to an inversion of the polarization of the volumes VC and/or VA and therefore to a homogenization of the polarization of the ferroelectric layer Ferrolay. Ideally, this layer is made, at the end of treatment, of a single monodomain of negative polarization P1, oriented towards the interface between the Ferrolay layer and the support assembly, here comprising the intermediate layer Int and the support layer Sprt, as illustrated by the FIG. 7 .

Second embodiment of the invention

According to the first embodiment of the invention, a hydrogen implantation step and an associated heat treatment are applied intended to correct the polarization of the ferroelectric layer in response to unwanted polarization inversions caused by a first heat treatment. This is a polarization repair method.

The second embodiment consists, as an alternative to the first embodiment, in preventing polarization inversions which, although undesired, can be expected.

There FIG. 10 illustrates a method 200 of manufacturing the structure Struct illustrated by the FIG. 2 . Steps S00 to S60 are the same as those of method 100 of the first embodiment.

In the second embodiment, it is following the detachment of the Ferro _lay layer from the Ferro _sub donor substrate, in step S60, that an ion implantation step S270 is implemented. This step is similar to step S120 of the first embodiment, except that it is necessary to take into account the fact that the Ferro _lay layer has not yet been thinned or has not yet been subjected to a heat treatment. The principle, however, remains the same, and can be applied after the steps of preparing and characterizing a reference sample Ref'.

The reference sample Ref' is a sample obtained in a Prep-S00 step by means of steps S00 to S60, further subjected to a step of applying a stabilizing heat treatment such as that of step S80 of the first embodiment. This heat treatment makes it possible to reveal the effects on the polarization of the Ferro _lay layer of the manufacturing steps S00 to S60, as explained above, in particular in the comments referring to the FIG. 3 , concerning the inverted polarization volumes. The analysis step Prep-S10 makes it possible to obtain the Char' characteristics of the volumes V _A , V _B , V _C and V _D on the basis of which the implantation parameters of step S270 are chosen, as in the first embodiment.

Following step S270, a step S80 of applying a heat treatment H.Treat' to the structure Struct is applied during a step S280. The treatment H.Treat' replaces the two treatments H.Treat-1 and H.Treat-2 of the first embodiment, with here an effect of stabilizing the structure Struct like the treatment H.Treat-1. Step S270 results in the formation of hydrogen ion concentration gradients generating electric fields opposing polarization inversions in the volume of the ferroelectric layer. Thus, the volumes V _A and V _C of polarization inversion of the volume V _B in the first embodiment are not formed. Step S270 makes it possible to prevent polarization inversions likely to occur in the volume of the ferroelectric layer, at the buried interface of the ferroelectric layer with the support assembly, or at these two regions.

However, the heat treatment H.Treat' may cause the appearance of a multi-domain volume on the surface of the Ferro _lay layer, as previously explained. It is then necessary to carry out a step S290 similar to step S100 of the first embodiment, which makes it possible to eliminate the surface multi-domain layer and to put the Ferro _lay layer at the desired thickness.

The advantage of process 200 over process 100 is that it is simpler, with only one heat treatment step being required instead of two.

The figures in this document are not necessarily to scale. Some features and components may be shown exaggerated in relation to other components or in a somewhat schematic form, and some details of conventional items may not be shown in the interest of clarity and conciseness.

Of course, the invention is not limited to the embodiments described and variant embodiments can be made without departing from the scope of the invention as defined by the claims.

Claims (11)

Method for manufacturing a structure comprising a negative polarization ferroelectric layer (P1), comprising the steps of:
- providing (S00) a plate of ferroelectric material (Ferro _sub ) having a first face (Top) and presenting a polarization (-P1) oriented towards this face;
- forming (S20) a weakening plane (Frgl) in the ferroelectric material plate (Ferro _sub ) by implanting hydrogen ions through the first face (Top);
- assembling (S40) the plate (Ferro _sub ) of ferroelectric material comprising the embrittlement plane (Frgl) to a support assembly (Sprt.Set) by bringing the first face into contact with a free surface of the support assembly;
- detaching (S60) a portion of the ferroelectric material wafer so as to define a ferroelectric layer (Ferro _lay ) detached from the ferroelectric material wafer (Ferro _sub ) and assembled to the support assembly (Sprt.Set) so as to obtain a structure (Struct) formed from the ferroelectric layer and the support assembly, the ferroelectric layer having a negative polarization (P1);
- carrying out (S120; S270) an additional full-field hydrogen implantation step, configured so as to correct or prevent the occurrence of polarization inversion in the volume of the ferroelectric layer and/or at its interface with the support assembly (Sprt.Ens); and
- apply (S80, S280) at least one first heat treatment (H.Treat-1; H. Treat') to the structure (Strct). The method according to claim 1, wherein the step (S120) of additional hydrogen implantation is carried out after the at least one first heat treatment (S80, Htreat-1), the method further comprising a step (S140) of applying a second heat treatment (H.Treat-2) to the structure after the step (S120) of additional hydrogen implantation. The method according to claim 2, further comprising a step (S100) of thinning (S.Pol) the ferroelectric layer, preferably between the step (S80) of applying the first heat treatment (H.Treat-1) and the step (S140) of applying the second heat treatment (H.Treat-2). The method according to claim 1, wherein the step (S270) of additional hydrogen implantation is carried out before the at least one first heat treatment (S280, Htreat’). The method according to claim 4, further comprising a step (S290) of thinning (S.Pol') the ferroelectric layer, preferably after the step (S280) of applying the first heat treatment (H.Treat'). The method according to any one of the preceding claims 1 to 5, comprising a step (Prep-S10) of characterizing a reference structure (Ref; Ref'), an adjustment of parameters of the additional hydrogen implantation step (S120; S270) being defined in response to this characterization. The method according to any one of the preceding claims 1 to 6, wherein the step of additionally implanting hydrogen ions (S120; S270) comprises a first correction ion implantation in the ferroelectric layer (Ferrolay), parameterized so as to generate a projected path of the implanted hydrogen ions such that a depth (D1(R _P )) of its maximum reaches or exceeds an interface depth between (i) a volume (V _B ) of stable polarization of the ferroelectric layer (Ferro _lay ) and (ii) a volume (V _C ) of reversed polarization due to the formation of the embrittlement plane or polarization which would reverse due to the formation of the embrittlement plane if a heat treatment were applied. The method according to any one of the preceding claims 1 to 7, wherein the additional hydrogen ion implantation step (S120; S270) comprises a second correction ion implantation, parameterized so as to generate a projected path of the implanted hydrogen ions such that a depth (D2(R _P )) of its maximum reaches or exceeds the depth of the interface between (i) the ferroelectric layer (Ferro _lay ) and (ii) the support assembly (Sprt.Ens), preferably at a distance to this interface which is less than the thickness of a polarization volume (V _A ) which is limited by this interface and which is multidomain or inverted with respect to a polarization of a stable polarization volume (V _B ) of the ferroelectric layer (Ferro _lay ) or which would invert or become multidomain if a heat treatment were applied. Method according to any one of the preceding claims 1 to 8, in which the additional full-field hydrogen implantation step is parameterized so as to obtain a hydrogen concentration of between 10 ¹⁹ and 10 ²² at/cm ³ in a volume (V _A , V _B , V _C ) of the structure. A method according to any one of the preceding claims 1 to 9, wherein the ferroelectric layer extends in an extension plane and the polarization (P1) makes an angle with included in an angular range ([Ang]) from -20° to -160° relative to the extension plane. A method according to any one of the preceding claims 1 to 10, wherein the ferroelectric layer (Ferro _lay ) is a layer of lithium niobate or lithium tantalate.