Method for preparing thin ferroelectric material layer

Abstract

The present invention relates to a method for preparing a thin layer of a single-domain ferroelectric material. The method comprises: implanting a light substance into a first surface of a ferroelectric donor substrate so as to form a fragile plane and define a first layer located between the fragile plane and the first surface of the substrate; assembling the first surface of the donor substrate and a supporting substrate together with the aid of a dielectric assembly layer; and breaking the donor substrate at the fragile plane. The dielectric assembly layer (7b) comprises an oxide having a hydrogen concentration lower than that of the first layer (3) or preventing hydrogen from diffusing into the first layer, or the dielectric assembly layer (7b) comprises a barrier that prevents hydrogen from diffusing into the first layer. The preparation method further comprises: heat-treating a free surface of the first layer so as to diffuse hydrogen contained in the first layer and induce a multi-domain transition of a surface portion of the first layer, and then thinning the first layer so as to remove at least the surface portion.

Description

Technical Field

[0001] The present invention relates to a method for preparing a thin layer of ferroelectric material. More specifically, the present invention relates to a method for preparing a thin layer of ferroelectric material capable of maintaining the single domain property of the ferroelectric material in the thin layer of the final product. For example, the preparation method is used in the fields of microelectronics, micromechanics, photonics, etc. Background Art

[0002] In the preamble, it was noted that ferroelectrics are materials that have an electric polarization in their natural state that can be reversed by the application of an external electric field. "Ferroelectric domains" refer to individual regions in a single sheet of material where the polarization is uniform (all dipole moments are aligned parallel to each other in a given direction). Thus, where a ferroelectric material is formed from a single region of uniform polarization, the material may be characterized as "single-domain," or where a ferroelectric material includes multiple regions of polarity that may be different, the material may be characterized as "multi-domain."

[0003] Various methods are known from the prior art for forming thin ferroelectric material layers. For example, it can be a method using molecular beam epitaxy, plasma sputtering, laser pulse deposition or even Smart Cut TM The invention relates to a technique for the removal of a thin layer from a solid substrate of ferroelectric material by fracturing it at a fragile area (or brittle plane) formed in the solid substrate by implanting a light substance.

[0004] The invention more specifically relates to the preparation of a thin ferroelectric layer obtained by applying this method. Document US20100088868 indicates that a lithium tantalate ferroelectric layer transferred by a method implementing hydrogen implantation leads to the formation of a multi-domain layer. This characteristic makes the layer unsuitable for use because it affects the performance of devices to be formed on / in the thin layer, such as surface acoustic wave devices (SAW).

[0005] Document FR3068508 discloses the finishing of a thin ferroelectric layer transferred to a support with the aid of an intermediate layer and by applying the Smart Cut method. In one example, the finishing consists in heat treating the thin layer, followed by polishing. According to the document, this sequence will allow the recovery of the initial monodomain characteristics of the thin ferroelectric layer. However, further analyses carried out by the applicant have shown that this finishing sequence alone does not guarantee the monodomain properties of the layer throughout its thickness, that is to say, from its interface with the intermediate layer to its free surface. More specifically, these analyses, carried out using PFM (“Piezoelectric Response Force Microscopy”) measurements, which typically extend from the measurement surface to a depth of more than about 50 nm, have confirmed that the surface portion of the thin layer is indeed monodomain, but that the buried portion juxtaposed to the intermediate layer can be shown to be multidomain. In order to carry out this “buried” PFM measurement, the layer is gradually thinned by chemical mechanical polishing to approach the measurement surface of the intermediate layer.

[0006] Document FR2914492 also discloses a method for manufacturing a thin ferroelectric material layer using Smart Cut technology. In this document, as in the aforementioned document US20100088868, the layer is prepared by subjecting the layer removed and transferred to the supporting substrate to an electric field to improve or restore its single-domain ferroelectric properties. This treatment requires the presence of a pair of electrodes, which are located on either side of the ferroelectric layer to be treated (for example, in electrical contact with the layer). It is not always possible to have such a pair of electrodes, especially when the ferroelectric layer is transferred to an electrically insulating substrate or an electrically insulating layer.

[0007] Purpose of the Invention

[0008] The object of the present invention is to propose a method for preparing a thin layer of ferroelectric material, which method at least partially solves the above-mentioned disadvantages. More specifically, the present invention relates to a method for transferring a layer of ferroelectric material by implementing a step of implanting a light substance, which method maintains or is able to restore the monodomain quality of the transferred layer throughout its thickness. Thus, in the layer preparation method, it is not necessary to subject the layer to an electric field in order to restore its monodomain quality, and the layer can be transferred to an electrically insulating substrate or provided with a dielectric layer. Summary of the invention

[0009] In order to achieve this object, the subject of the present invention proposes a method for thinly preparing a single-domain layer of a ferroelectric material, the method comprising the following steps:

[0010] - implanting a light substance in a first face of a ferroelectric donor substrate to form an embrittlement plane and to define a first layer between said embrittlement plane and said first face of said donor substrate;

[0011] - assembling said first face of said donor substrate to a support substrate by means of a dielectric assembly layer;

[0012] - breaking the donor substrate at the embrittlement plane to transfer the first layer to the support substrate and expose the free face of the first layer;

[0013] - finishing the first layer, the finishing comprising heat treating the free surface of the first layer and then thinning the first layer to form the thin single domain layer;

[0014] According to the present invention, the dielectric assembly layer comprises an oxide having a lower hydrogen concentration than that of the first layer or preventing hydrogen from diffusing toward the first layer, or the dielectric assembly layer comprises a barrier preventing hydrogen from diffusing toward the first layer.

[0015] Surprisingly, this precise series of preparation steps combined with the specific properties of the assembled layers allows the formation of a final thin layer with a single domain polarization throughout its thickness and with satisfactory crystal and surface quality.

[0016] Other advantageous and non-limiting features according to the invention, taken individually or in any technically feasible combination:

[0017] - said thinning comprises chemical mechanical polishing applied to said free face of said first layer;

[0018] - said heat treatment is carried out at a temperature between 300° C. and the Curie temperature of said ferroelectric material constituting said first layer, and for a period of time between 30 minutes and 10 hours;

[0019] - the heat treatment is carried out in an oxidizing or neutral gaseous atmosphere;

[0020] - the donor substrate comprises a ferroelectric material selected from LiTaO3 or LiNbO3;

[0021] - the ferroelectric material has a crystal orientation between 30°RY and 60°RY;

[0022] - the material of the support substrate is silicon;

[0023] - the donor substrate is a block of solid material;

[0024] - the donor substrate comprises a thick layer of ferroelectric material on a handling substrate;

[0025] - wherein the thermal expansion coefficient of the handling substrate is the same as or close to the thermal expansion coefficient of the support;

[0026] - the material of the first layer and the material of the support substrate have different coefficients of thermal expansion;

[0027] - the support substrate is provided with a charge trapping layer;

[0028] - the dielectric assembly layer comprises an oxide having nitrogen, the nitrogen / oxygen ratio of the oxide being greater than or equal to 0.01 or 0.05;

[0029] - the dielectric assembly layer comprises silicon oxide with nitrogen, the nitrogen / oxygen ratio of the silicon oxide being between 0.01 and 0.25 or between 0.05 and 0.01;

[0030] - the dielectric assembly layer is formed by a stack of layers comprising a first oxide layer arranged on the side of the support substrate and a silicon nitride layer arranged on the side of the first layer;

[0031] According to another aspect, the present invention provides a device, comprising:

[0032] - a thin completely single domain layer of a ferroelectric material;

[0033] a dielectric assembly layer in contact with said thin layer and comprising an oxide preventing the diffusion of hydrogen towards said thin layer, or comprising a barrier preventing the diffusion of hydrogen towards said first layer;

[0034] - a support, said support being in contact with said dielectric layer.

[0035] Other advantageous and non-limiting features according to this aspect of the invention, taken individually or in any technically feasible combination:

[0036] - the support substrate is provided with a charge trapping layer between the solid part of the support and the dielectric assembly layer;

[0037] - the dielectric assembly layer comprises an oxide having nitrogen, the nitrogen / oxygen ratio of the oxide being greater than or equal to 0.01 or 0.05;

[0038] - the dielectric assembly layer comprises silicon oxide with nitrogen, the nitrogen / oxygen ratio of the silicon oxide being between 0.01 and 0.25 or between 0.05 and 0.01;

[0039] The dielectric assembly layer is formed by a stack of layers comprising a first oxide layer arranged on the side of the support substrate and a silicon nitride layer arranged on the side of the first layer. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Other features and advantages of the present invention will become apparent from the following detailed description of the invention with reference to the accompanying drawings, in which:

[0041] Figure 1 1A to 1D show a first embodiment of the method according to the present invention;

[0042] Figure 2 2A to 2D show a second embodiment of the method according to the present invention;

[0043] Figure 3 A method for preparing a layer according to the present invention is schematically shown;

[0044] Figure 4 A surface acoustic device that can be manufactured using the method according to the present invention is shown. DETAILED DESCRIPTION

[0045] In order to simplify the following description, the same reference numerals are used for the same elements or for elements that perform the same function in different embodiments of the described method.

[0046] The drawings are schematic representations which, for the sake of readability, are not drawn to scale. In particular, with respect to their lateral dimensions, the thickness of the layers is not to scale.

[0047] The expression "thermal expansion coefficient" used in the rest of this specification in relation to a layer or substrate refers to the expansion coefficient in a direction defined in a principal plane defining the layer or substrate. If the material is anisotropic, the retained value of the coefficient will be the value of greatest magnitude. The values ​​of the coefficient are those measured at room temperature.

[0048] Figure 4 A surface acoustic device that can be manufactured using the method according to the present invention is shown.

[0049] The thin ferroelectric layer 10 is placed in direct contact with the support 7 via the dielectric layer 7b. Metal electrodes 11a, 11b are arranged on the thin ferroelectric layer 10. As is well known per se, a high-frequency electric signal applied to one of the electrodes propagates in the thin ferroelectric layer 10 in the form of a surface wave to be processed (e.g. filtered) in said one and restored on the other electrode. The characteristics of the thin ferroelectric layer 10 and more generally of all the layers forming the stack in which the thin ferroelectric layer 10 is located are decisive for performing the desired processing of the electric signal, in particular the thickness of the thin ferroelectric layer 10 and the monodomain crystal quality.

[0050] The characteristics of the support 7 can influence the propagation of waves in the thin layers 10 by acoustic coupling. Therefore, the nature and thickness of these layers can also be decisive for performing the desired processing of the electrical signals, or at least influence this processing. Figure 4 In the example shown, the support 7 comprises a solid silicon portion 7a and a dielectric assembly layer 7b, preferably comprising an oxide, in direct contact with the thin ferroelectric layer 10. By way of example, the assembly layer 7b can be made of silicon oxide or silicon nitride, or formed by a stack of layers consisting of these materials. Preferably, the support 7 is resistive so as to avoid electrical coupling with high-frequency signals applied to the electrodes 11a, 11b, which electrical coupling could affect the propagation of high-frequency signals. The solid portion 7a can therefore be formed by a high-resistivity silicon substrate, that is to say, the resistivity of the solid portion 7a is greater than 1000 ohm-cm, and more preferably greater than 3000 ohm-cm.

[0051] In order to support the resistance properties of the support 7, Figure 4In the example of , a charge trapping layer 7c, for example made of polycrystalline silicon, has been inserted between the solid part 7a and the assembly layer 7b. The formation of such a layer and its contribution to the resistivity of the supporting substrate 7 are well known per se, and a detailed description thereof can be found in the documents FR2860341, FR2933233, FR2953640, US2015115480. Of course, the charge trapping layer 7c can be formed using techniques other than the technique of providing a layer made of polycrystalline silicon. The layer can also include carbon or consist of or include silicon carbide or an alloy of silicon and carbon. Alternatively, electric traps can be produced in the layer 7c by ion bombarding a surface portion of the support with a relatively heavy substance (for example, argon) to produce crystal defects therein capable of trapping charges. A charge trapping layer 7c made of a porous material can also be provided, for example, by making the surface portion of the substrate 7 porous when the substrate 7 is made of silicon.

[0052] It has been noted that Figure 4 The device shown has no electrodes underneath the thin ferroelectric layer 10. The presence of the dielectric assembly layer 7b does not allow contact with the back side of the thin ferroelectric layer 10. Therefore, it is not possible to have such electrodes or to make electrical contacts to polarize the thin ferroelectric layer 10, as described in the prior art cited in the introduction of this application.

[0053] Therefore, it is important to provide a method for preparing a thin ferroelectric layer 10 that can preserve or restore the single domain characteristics of the layer 10 without applying a repolarization voltage.

[0054] refer to Figure 1 and Figure 2 , and in order to allow the manufacture of Figure 4 In the device shown, the invention generally transfers the first ferroelectric layer 3 onto the supporting substrate 7, the first layer 3 being obtained from the monodomain ferroelectric donor substrate 1 by a transfer technique based on implantation of light species, such as hydrogen species.

[0055] According to this technique, hydrogen is implanted in the first face 4 of the donor substrate 1 to form therein a buried embrittlement plane 2. In this way, a first layer 3 is defined between the embrittlement plane 2 and the first face 4 of the donor substrate 1. The first face 4 of the donor substrate 1 is then assembled with a support 7, in this case with the aid of a dielectric assembly layer 7b. The donor substrate 1 is then fractured at the embrittlement plane 2, for example using a moderate heat treatment and / or the application of mechanical forces. The first layer 3 is then detached from the donor substrate 1 so as to expose the free face 8 of the first layer, the other face 4 of the thin layer being in direct contact with the assembly layer 7b of the support substrate 7.

[0056] It is generally necessary to provide steps for preparing the first transferred layer 3 thus transferred onto the support 7 so as to form a "useful" thin layer 10. These steps are generally aimed at improving the crystal quality of the thin layer 10 and its surface condition (for example the roughness of the thin layer 10). These preparation steps may comprise a step of thinning the first layer 3 (for example by chemical mechanical polishing indicated as "CMP" in the table below) and / or a step of thermal treatment (in the case of ferroelectric materials, this thermal treatment step may be a thermal treatment of about 500° C. carried out in a neutral atmosphere or in an atmosphere containing oxygen, indicated as "TTH" in the table below). It will be noted that this step of thermal treatment of the free face 8 of the layer 3 is separate from the step of causing the fracture of the donor substrate. In fact, this thermal treatment does not allow the free face of the layer 3 to be treated, since it has not yet been completely released during the fracture annealing.

[0057] However, the Applicant has observed that the method used to prepare the first layer 3 transferred to the support substrate 7 intended to form the thin layer 10 may result in the creation of multiple ferroelectric domains within the thin layer 10, thus conferring a multi-domain nature to the thin layer. As previously mentioned, such a characteristic makes the layer unsuitable for its use, since it affects the devices to be formed on / in the transferred thin layer (e.g., devices similar to Figure 4 The performance of the surface acoustic device shown).

[0058] The inventors of the present application have observed that the nature and order of steps in preparing the first layer 3 can significantly affect the ferroelectric properties of the thin layer 10. Based on these initial observations, the inventors of the present application have analyzed various sequences of preparation steps using the piezoelectric force microscopy (PFM) technique to characterize the single-domain or multi-domain nature of the surface of the thin layer 10. The following table summarizes these experimental results.

[0059] Table 1

[0060]

[0061] It can be seen (row 1 of the table) that the first layer 3 obtained at the end of the separation step and before any preparation exhibits compliant ferroelectric quality (that is to say monodomain). However, the surface condition and the crystal quality of this layer require the application of preparation steps.

[0062] The fourth row of the table corresponds to the method of linked thinning and thermal treatment (CMP+TTH) and it can be seen that at the end of this treatment the thin layer has multi-domain ferroelectric properties.

[0063] Rows 2 and 3 represent the application of a single step of thinning CMP or thermal treatment TTH, respectively. Both steps lead to thin layer 3 having properties that do not meet the requirements.

[0064] The Applicant has realised that by sequentially applying a thermal treatment step TTH followed by a thinning step CMP, a thin layer 10 is obtained having a monodomain distribution of the ferroelectric polarisation (as observed by PFM on the surface of the layer) and an acceptable crystal and surface quality.

[0065] Additional analyses carried out by the Applicant have made it possible to determine that the formation of multiple ferroelectric domains in the thin layer 10 is probably related to the presence of a strong hydrogen concentration gradient in the first layer 3 prior to its preparation. In fact, it has been observed that the application of a heat treatment on a layer having such a gradient leads to the formation of multiple ferroelectric domains. The hydrogen implanted in the donor substrate 1 during the step of defining the first layer 3 is distributed in the substrate according to a profile having a concentration peak at the embrittlement plane 2. After fracture, the first ferroelectric layer 3 transferred to the supporting substrate 7 therefore has a significant hydrogen concentration at its free face 8 (of the order of 10^21 at / cm^3). This concentration decreases over the thickness of the thin layer in the direction of the support 7 to reach a level of about 10^19 at / cm^3.

[0066] When the preparation steps include a first step of thinning the first layer 3 to a target thickness of the thin layer 10 followed by a heat treatment, the heat treatment is applied while the thinned layer (particularly the surface portion of the thinned layer) still has a strong hydrogen gradient. This results in the thickness or surface portion of the thin layer 10 where the hydrogen gradient is large being transformed into a plurality of ferroelectric domains.

[0067] When these two steps are reversed, the surface portion of the first layer 3 with a strong hydrogen concentration gradient is transformed into having multiple ferroelectric domains. However, this multi-domain surface portion of the first layer 3 is eliminated by the following thinning step. This makes it possible to provide a thin layer 10 with the desired single domain quality.

[0068] It should be noted that heat treatment (whether performed before or after the thinning step) results in diffusing hydrogen (the layer is saturated with hydrogen) and reduces the gradient along the thickness of the layer. The heat treatment that may be subsequently applied to prepare the thin layer 10 is therefore unlikely to affect the monodomain nature of the layer.

[0069] Very detailed analysis of the thin layer 10 after it has been thinned using piezoelectric force microscopy also suggests that the thin layer 10 may have a buried multi-domain portion in contact with the assembly layer. This buried multi-domain portion is not easily visible using simple PFM surface observations, which are limited to a depth of about 50 nm.

[0070] Based on PFM observations of thinned layers of several samples, the applicant has determined that the properties of the dielectric assembly layer 7b play an important role in the redistribution of hydrogen in the ferroelectric layer during the thermal treatment of the method for preparing the thin layer 10.

[0071] When the assembly layer 7b is rich in hydrogen, as in the case of some deposited oxides or dielectrics, the assembly layer forms a hydrogen source, and the thermal treatment can cause the hydrogen to migrate into the ferroelectric layer in contact with the assembly layer. The hydrogen can accumulate in the ferroelectric layer, especially at the interface between the layer and the assembly layer 7b. This migration and this accumulation can constitute a hydrogen concentration gradient, so that the part of the ferroelectric layer in contact with the assembly layer can undergo a multi-domain transition during the thermal treatment.

[0072] In contrast, when the assembly layer 7b is relatively hydrogen-poor (that is, the hydrogen concentration of the assembly layer is lower than the hydrogen concentration in the transfer ferroelectric layer), excess hydrogen in the ferroelectric layer can be absorbed into the assembly layer 7b during the diffusion caused by the heat treatment. This therefore prevents the accumulation of hydrogen at the assembly interface and avoids multi-domain transition in the portion of the ferroelectric layer near the interface.

[0073] The present invention takes advantage of these results and observations to propose a method of preparing a thin layer 10. More specifically, the present invention relates to a method of preparing a thin layer 10 of ferroelectric material transferred from a donor substrate 1 to a support substrate 7 by a transfer technique (as generally described above) involving the implantation of a light species (such as a hydrogen species). There are several embodiments to perform this transfer.

[0074] according to Figure 1 In the first embodiment shown in 1A to 1D of , the donor substrate 1 consists of a solid single-domain block of a ferroelectric material (e.g., LiTaO3, LiNbO3, LiAlO3, BaTiO3, PbZrTiO3, KNbO3, BaZrO3, CaTiO3, PbTiO3 or KTaO3). The donor substrate 1 can be in the form of a circular wafer of standardized size (e.g., 150 mm or 200 mm in diameter). However, the present invention is by no means limited to these sizes or this shape. The donor substrate 1 may have been removed from an ingot of ferroelectric material, the removal being performed so that the donor substrate 1 has a predetermined crystal orientation. The orientation is selected according to the intended application. Therefore, in the case where the properties of a thin layer are intended to be used to form a SAW filter, it is common practice to select an orientation between 30°RY and 60°RY or between 40°RY and 50°RY. However, the present invention is by no means limited to a specific crystal orientation.

[0075] Regardless of the crystal orientation of the donor substrate 1, the method comprises introducing hydrogen into the donor substrate 1. This introduction may correspond to an implantation of hydrogen, that is to say a bombardment of the plane 4 of the donor substrate 1 with hydrogen ions. It is naturally foreseeable that this hydrogen introduction may be supplemented (for example by implantation) by the introduction of other light species, such as helium.

[0076] In a manner known per se and as Figure 11B , the purpose of the implanted ions is to form an embrittlement plane 2 which delimits a first layer 3 of ferroelectric material to be transferred, situated on one side of a face 4 , and another portion 5 forming the rest of the substrate.

[0077] The nature of the implanted substance, the dose and the implantation energy are selected according to the thickness of the layer to be transferred and the physicochemical characteristics of the donor substrate 1. In the case of a donor substrate 1 made of LiTaO3, it is thus possible to implant a layer of LiTaO3 at an energy between 30 keV and 300 keV. E 16 at / cm 2 to 5 E 17 at / cm 2 The hydrogen dose is between 100 and 150 nm to define the first layer 3 of about 200 nm to 2000 nm.

[0078] In such Figure 1 In the next step shown in 1C, the flat face 4 of the donor substrate 1 is assembled with the face 6 of the support substrate 7. The support substrate 7 can be of the same size and the same shape as the donor substrate 1. For reasons of availability and cost, the support substrate 7 is a silicon single crystal wafer or a multicrystalline wafer. However, more generally, the support substrate 7 can be formed of any material (e.g., silicon) or even an electrically insulating material (such as sapphire or glass) and can be of any shape. In order to form Figure 4 For the device shown, a support substrate formed of a solid portion 7a made of high-resistance silicon and provided with a charge-trapping layer 7b made of polysilicon, for example, will be selected.

[0079] Prior to the assembly step, it is conceivable to prepare the face of the substrate to be assembled using cleaning, brushing, drying, polishing steps or using plasma activation, for example based on oxygen or nitrogen.

[0080] The assembly step may correspond to bringing the donor substrate 1 into close contact with the support substrate 7 by molecular adhesion and / or electrostatic bonding. In order to facilitate the assembly of the two substrates 1, 7, it is provided that at least one dielectric assembly layer 7b is formed on the face 6 of the support substrate 7 to be assembled. It may also or alternatively be provided that an assembly layer is formed on the face 4 of the donor substrate 1 to be assembled, preferably before forming the embrittlement plane 2. In other words, the dielectric assembly layer 7 may be formed at least partially on one and / or the other of the two substrates 1, 7.

[0081] The dielectric assembly layer 7b is formed, for example, of silicon oxide, or more generally of oxides such as Ta2O5, ZrO2 or HfO (which can be formed by deposition like SiO2) or silicon nitride, and has a thickness ranging from a few nanometers to a few micrometers. The assembly layer 7b can be composed of a stack of different types of dielectric layers. The dielectric assembly layer 7b can be manufactured according to various techniques known in the prior art, such as thermal oxidation or nitridation treatment, chemical deposition (PECVD, LPCVD, etc.), etc.

[0082] Generally, the technique used to deposit or form the assembly layer will be advantageous, resulting in a small amount of hydrogen being incorporated therein. It should be noted that the choice of the material forming the assembly layer, its thickness and its formation technique is not completely free. This choice may be determined in particular by the properties of electrical insulation, the propagation of acoustic waves and / or the adhesion of the layer in the final structure and according to the intended application. Finally, it should be noted that it is generally easier and more common to form the assembly layer by deposition, in particular oxide deposition, since this method of formation can be independent of the nature of the support and carried out at relatively moderate temperatures (below 1000° C. or 950° C. or even below 900° C. (the temperature may sometimes be imposed by the nature of the solid part of the support 7a or by the presence of the capture layer 7c)).

[0083] According to an important feature of one embodiment, care will be taken to ensure that the hydrogen concentration of the dielectric assembly layer 7b is lower than the average hydrogen concentration in the first layer 3. By estimating that most of the hydrogen of the first layer 3 comes from the hydrogen implanted to form the embrittlement plane 2, the average concentration of hydrogen can be determined by dividing the implantation dose (in at / cm^2) by the thickness of the first layer (in cm). The average concentration is typically between 10^20at / cm^3 and 10^22at / cm^3. By way of example, in the case where the assembly layer 7b is formed by silicon oxide deposited on a supporting substrate, this oxide will be annealed before the assembly step so that most of the hydrogen it contains is diffused outward (exodiffuse).

[0084] Typically, a step of annealing the assembly layer 7b may be provided, which is intended to bring the dielectric layer to a higher temperature than the heat treatment used to prepare the first transfer layer, which will be described later in this specification. Thus, the layer may encounter temperatures of 600°C, 700°C or even 800°C or more. After this outdiffusion step, the average hydrogen concentration in the assembly layer 7b may therefore be less than 5.10^20at / cm^3, or advantageously less than 10^18at / cm^3, or more advantageously less than 10^18at / cm^3. It will be noted that, in addition to the hydrogen concentration of the assembly layer, annealing may also change the characteristics of the assembly layer. Annealing may in particular reduce the diffusivity of hydrogen (that is, the ability of this substance to diffuse in the material forming the assembly layer) so that hydrogen is less likely to diffuse towards the first layer 3 even at relatively high concentrations (about 10^20at / cm^3).

[0085] As an alternative or in addition to this embodiment, provision may be made for providing a dielectric assembly layer 7b having a barrier layer preventing the diffusion of hydrogen towards the first layer 3. By way of example, the assembly layer 7b may be formed by a stack of a first oxide (e.g. silicon oxide) layer on the side of the supporting substrate 7a (or the trapping layer 7c, when present) and a silicon nitride layer on the side of the first ferroelectric layer 3. The nitride layer prevents any hydrogen that may be present in the silicon oxide layer from diffusing towards the first ferroelectric layer 3.

[0086] According to yet another approach, the dielectric assembly layer can be provided with any hydrogen concentration, but this layer has a very low hydrogen diffusion rate, so that the hydrogen remains sufficiently trapped in this layer not to diffuse significantly towards the thin layer 3 .

[0087] Such a layer may be formed of an oxide, for example silicon oxide SiON with nitrogen, with a nitrogen / oxygen ratio greater than or equal to 0.01, 0.05 or 0.1. When the assembled dielectric layer 7b is based on silicon oxide SiON (which is very common), since this material has well-known characteristics of formation, adhesion, electrical insulation and acoustic wave propagation, it may be decided not to exceed any nitrogen / oxygen ratio greater than 0.1 or 0.25, so as not to excessively change these characteristics and maintain a behavior equivalent to or close to that of simple silicon oxide SiO2. It should be noted that such a nitrogen-rich oxide layer can be easily formed by a deposition technique (for example, PECVD) in which at least one of the carrier gases can be chosen to be nitrogen, which can be incorporated into the oxide layer in a controlled manner. The nitrogen / oxygen ratio can be measured by a technique called EDX (“energy dispersive X-ray spectroscopy”) or determined from nitrogen and oxygen measurements in the oxide layer 4 determined by SIMS (secondary ion mass spectrometry) measurements.

[0088] The dielectric assembly layer 7 b may be formed of a material having a low diffusion rate (eg, an oxide containing nitrogen as described above) throughout its entire thickness.

[0089] Alternatively, it is possible to provide only a barrier layer preventing the diffusion of hydrogen to the first layer 3, made of such a material with a low diffusion rate, as described in the previous alternative. For example, a silicon oxide layer with a high hydrogen concentration can be formed by deposition on a supporting substrate, wherein the layer is prepared using a nitrogen-based plasma to absorb the nitrogen in a surface manner. Thus, a nitrogen-rich SiO barrier layer is formed at this surface thickness, which prevents the diffusion of hydrogen contained in the rest of the oxide.

[0090] When the solid part of the support 7a is provided on a surface having a charge-trapping layer 7c comprising silicon, it is conceivable to form the dielectric assembly layer 7b by thermally oxidizing the surface part of the trapping layer by treating it in an oxidizing atmosphere at a moderate temperature below 950° C. and preferably between 800° C. and 900° C. This thus avoids damaging the charge-trapping layer 7c, for example by recrystallizing it by subjecting it to high temperatures above 950° C. or 900° C. Optionally, a step of polishing the surface of the trapping layer 7c thus oxidized may be introduced to make it compatible with the subsequent assembly steps.

[0091] Whatever the nature of the dielectric assembly layer 7b and the treatments it undergoes, at the end of this assembly step, a dielectric assembly layer 7b is provided. Figure 1 1C of the drawings, the assembly comprising two associated substrates, the planar face 6 of the support substrate 7 being adhered to the planar face 4 of the donor substrate 1 by means of a dielectric assembly layer 7b.

[0092] The assembly is then processed so as to separate the first layer 3 of ferroelectric material from the donor substrate 1 by fracture at the embrittlement plane 2 .

[0093] Thus, this separation step may consist in applying a heat treatment to the assembly at a temperature ranging from about 80° C. to 300° C., so as to allow the transfer of the first layer 3 to the supporting substrate 7. Instead of or in addition to the heat treatment, this step may consist in applying a knife blade or a jet of a gaseous or liquid fluid at the embrittlement plane 2, or in applying any other force of a mechanical nature.

[0094] After this separation step, one obtains Figure 1 1D shows a structure 9. The structure 9 comprises a first layer 3 of ferroelectric material, the first layer 3 comprising a first free surface 8 and a second surface 4 (corresponding to the first surface of the donor substrate) located on a support substrate 7, and a dielectric assembly layer 7b is located between the first layer 3 and the rest of the support 7.

[0095] Figure 22A to 2D show a second embodiment particularly suitable for producing a heterostructure 9 in which the thermal expansion coefficient of the first layer 3 is very different from that of the support, for example by more than 10%.

[0096] refer to Figure 2 2A, in this case the donor substrate 1 consists of a thick layer of ferroelectric material 1a having the same properties as described for the solid block of ferroelectric material with respect to the first embodiment, and a handling substrate 1b.

[0097] Advantageously, the handling substrate 1 b is formed of a material (or materials) provided with a coefficient of thermal expansion close to that of the material constituting the support substrate 7. "Close" means that the difference between the coefficient of thermal expansion of the handling substrate 1 b and that of the support is smaller as an absolute value than the difference between the coefficient of thermal expansion of the solid block of ferroelectric material and that of the support matrix 7.

[0098] Preferably, the handling substrate 1b and the support substrate 7 have the same coefficient of thermal expansion. During the assembly of the donor substrate 1 and the support 7, a structure is formed that can withstand heat treatment at relatively high temperatures. For simplicity of implementation, this can be obtained by selecting the handling substrate 1b so that it is formed of the same material as the support substrate 7.

[0099] In order to form the donor substrate 1 of the present embodiment, a solid block of ferroelectric material is first assembled with a handling substrate 1b, for example according to a bonding technique by molecular adhesion as described above. Then, a layer of ferroelectric material 1a is formed by thinning (for example, by grinding and / or chemical mechanical polishing and / or etching). Prior to assembly, provision may be made to form an adhesion layer on one and / or the other of the faces to be contacted (for example, by depositing silicon oxide and / or silicon nitride). The assembly may include applying a low-temperature heat treatment (for example, between 50°C and 300°C, typically 100°C) to allow sufficient strengthening of the bonding energy to allow the next step of thinning to be carried out.

[0100] The thickness of the handling substrate 1 b is chosen to be substantially the same as the thickness of the support substrate 7. The thinning step is performed so that the thickness of the thick layer 1 a is small enough to have a low strength against the stresses generated during the heat treatment applied in the rest of the method. At the same time, the thickness is large enough to be able to remove the first layer 3 or a plurality of such layers. For example, the thickness may be between 5 micrometers and 400 micrometers.

[0101] The following steps of the method of this second embodiment are identical to those described in the first embodiment. Hydrogen is implanted into the thick layer 1a to generate an embrittlement plane 2, which marks the separation of the first layer 3 from the rest 5 of the donor substrate 1, as shown in FIG. Figure 2This step is followed by a step of assembling the donor substrate 1 on the support substrate 7, as shown in FIG. Figure 2 2C . A dielectric assembly layer 7 b will be provided between the donor substrate 1 and the support substrate 7, having the same characteristics as those provided in the context of the first embodiment. It should be noted that if it is decided to form a silicon oxide layer (or a dielectric of any other nature requiring an out-diffusion anneal of hydrogen) on the face of the donor substrate 1 to be assembled in this embodiment, the thickness of this silicon oxide layer will be limited to a few nanometers (e.g. less than 10 nm or 50 nm), since it is difficult to apply an out-diffusion anneal to this layer without the risk of damaging the donor substrate 1 consisting of an assembly of materials with very different thermal expansion coefficients.

[0102] The first layer 3 is then separated from the rest of the substrate 5 in order to obtain Figure 2 The 2D structure shown in 9.

[0103] An advantage of this embodiment is that the assembly formed by the donor substrate 1 and the support 7 can be exposed to temperatures much higher than those applied in the first embodiment, without risking uncontrolled fracture of one of the substrates or delamination of the donor substrate 1. The balanced structure, in terms of the coefficient of thermal expansion of the assembly, thus makes it possible to facilitate the step of detaching the first layer 3 by exposing the assembly to relatively high temperatures, for example between 100° C. and 600° C.

[0104] Whatever the embodiment chosen and as described above, a step of preparing the first layer 3 is then required in order to form a thin layer 10 with satisfactory crystal and surface quality.

[0105] like Figure 3 As shown in 3A of , the method initially comprises heat treating the free surface 8 of the first transferred layer 3. This heat treatment allows the correction of crystal defects present in the first layer 3. In addition, this helps to consolidate the bond between the layer 3 and the support 7. The above studies have shown that if the temperature is high enough, this heat treatment also has the effect of diffusing the hydrogen contained in the first layer 3, especially in its surface portion, and causing a multi-domain transformation of the surface portion. The thickness of the surface portion can be about 50nm or less and can be established over the entire range of the layer 3. At the end of the heat treatment, the ferroelectric layer 10 has a relatively constant hydrogen concentration over its thickness and between 10^18at / cm^3 and 10^19at / cm^3. It should be noted that due to the characteristics of the above-mentioned assembly layer 7b, the hydrogen contained in the assembly layer is unlikely to diffuse towards the thin layer 3 during annealing and / or be trapped at the interface between the assembly layer 7b and the layer. Therefore, this avoids the formation of a hydrogen concentration gradient in the buried region near the interface and in this way prevents the region from being transformed into having multi-domain characteristics.

[0106] The heat treatment used to prepare the compliant thin layer 10 subjects the structure to a temperature between 300° C. and the Curie temperature of the ferroelectric material (and preferably greater than or equal to 450° C., 500° C. or 550° C. to promote the diffusion of hydrogen) for a period of between 30 minutes and 10 hours. The heat treatment can provide a gentle slope. The heat treatment is preferably carried out by exposing the free face 8 of the first layer 3 to an oxidizing or neutral gaseous atmosphere (that is, without covering this face of the thin layer with a protective layer that can prevent the outward diffusion of hydrogen).

[0107] refer to Figure 3 3B, the preparation method also includes thinning the thin layer 3 after the heat treatment. This thinning can correspond to polishing the free surface 8 of the thin layer 3, for example, by mechanical, chemical-mechanical thinning techniques and / or chemical etching. This makes it possible to prepare the free surface 8 so that it presents a low roughness, for example, less than 0.5nm RMS 5×5μm by atomic force measurement (AFM), and the multi-domain surface part of the thin layer 10 will be removed. A removal of a thickness of 100nm to 300nm is generally provided to reach the target thickness of the thin layer 10, and in all cases a thickness greater than the thickness of the multi-domain surface part.

[0108] This thus constitutes a thin monodomain layer throughout its thickness having the qualities required for surface quality, crystal quality. The dielectric assembly layer 7b having a hydrogen concentration lower than that of the first layer 3 or comprising a barrier to the diffusion of hydrogen towards the first layer 3 avoids the accumulation of excess hydrogen in the buried region of the thin layer 10 located near the assembly interface. This prevents multidomain transitions in this region, which are impossible to eliminate.

[0109] Of course, the invention is not limited to the embodiments described and variations may be added without departing from the scope of the invention as defined by the claims.

Claims

1. A method for preparing a thin single domain layer (10) of a ferroelectric material, the method comprising the following steps: - implanting a light substance in a first face (4) of a ferroelectric donor substrate (1) to form an embrittlement plane (2) and to define a first layer (3) located between said embrittlement plane (2) and said first face (4) of said donor substrate (1); - assembling the first face (4) of the donor substrate (1) to a support substrate (7) by means of a dielectric assembly layer (7b), wherein the dielectric assembly layer (7b) is annealed prior to the assembling step so that the dielectric assembly layer (7b) reaches a higher temperature than the heat treatment used for producing the first layer (3); - breaking the donor substrate (1) at the embrittlement plane (2) so as to transfer the first layer (3) to the support substrate (7) and expose the free face (8) of the first layer (3); - finishing the first layer (3), the finishing comprising heat treating the free surface (8) of the first layer (3) and then thinning the first layer (3) to form the thin monodomain layer (10); The dielectric assembly layer (7b) comprises an oxide which prevents hydrogen from diffusing towards the first layer, or the dielectric assembly layer (7b) comprises a barrier which prevents hydrogen from diffusing towards the first layer (3).

2. The method according to claim 1, wherein: The thinning comprises a chemical mechanical polishing applied to the free face (8) of the first layer (3).

3. The method according to claim 1 or 2, wherein: The heat treatment is carried out at a temperature between 300° C. and the Curie temperature of the ferroelectric material constituting the first layer ( 3 ) and for a period of time between 30 minutes and 10 hours.

4. The method according to claim 1 or 2, wherein: The heat treatment is performed in an oxidizing or neutral gaseous atmosphere.

5. The method according to claim 1 or 2, wherein: The donor substrate (1) comprises a ferroelectric material selected from LiTaO3 or LiNbO3.

6. The method according to claim 1 or 2, wherein: The ferroelectric material has a crystal orientation between 30°RY and 60°RY.

7. The method according to claim 1 or 2, wherein: The material of the supporting substrate (7) is silicon.

8. The method according to claim 1 or 2, wherein: The donor substrate (1) is a block of solid material.

9. The method according to claim 1 or 2, wherein: The donor substrate (1) comprises a thick ferroelectric material layer (1a) located on a handling substrate (1b).

10. The method according to claim 9, wherein: The thermal expansion coefficient of the handling substrate (1b) is the same as or close to the thermal expansion coefficient of the supporting substrate (7).

11. The method according to claim 1 or 2, wherein: The material of the first layer (3) and the material of the support substrate (7) have different thermal expansion coefficients.

12. The method according to claim 1 or 2, wherein: The dielectric assembly layer (7b) comprises an oxide having nitrogen, the nitrogen / oxygen ratio of the oxide being greater than or equal to 0.01 or 0.

05.

13. The method according to claim 1 or 2, wherein: The dielectric assembly layer (7b) comprises silicon oxide with nitrogen, wherein the nitrogen / oxygen ratio of the silicon oxide is between 0.01 and 0.25 or between 0.05 and 0.

01.

14. The method according to claim 1 or 2, wherein: The dielectric assembly layer (7b) is formed by a stack of layers including a first oxide layer disposed on the side of the support substrate (7) and a silicon nitride layer disposed on the side of the first layer (3).

15. A device having a thin layer of ferroelectric material, the device comprising: - a completely monodomain thin monodomain layer (10) of the ferroelectric material layer; - a dielectric assembly layer (7b) which is in contact with the thin single domain layer (10) and comprises an oxide that prevents hydrogen from diffusing toward the thin single domain layer (10), or comprises a barrier that prevents hydrogen from diffusing toward the thin single domain layer (10), the hydrogen concentration of the dielectric assembly layer (7b) being lower than the average hydrogen concentration in the thin single domain layer (10); - A supporting substrate (7) which is in contact with the dielectric layer.

16. The device according to claim 15, wherein The support substrate (7) is provided with a charge trapping layer (7c) positioned between a solid portion (7a) of the support substrate and the dielectric assembly layer (7b).

17. A device according to claim 15 or 16, wherein The dielectric assembly layer (7b) comprises an oxide having nitrogen, the nitrogen / oxygen ratio of the oxide being greater than or equal to 0.01 or 0.

05.

18. The device according to claim 15 or 16, wherein The dielectric assembly layer (7b) comprises silicon oxide with nitrogen, wherein the nitrogen / oxygen ratio of the silicon oxide is between 0.01 and 0.25 or between 0.05 and 0.

01.

19. The device according to claim 15 or 16, wherein The dielectric assembly layer (7b) is formed by a stack of layers including a first oxide layer on the side of the support substrate (7) and a silicon nitride layer on the side of the thin monodomain layer (10).

Images