Process for preparing a thin layer

Abstract

The present invention relates to a process for preparing a thin layer (1), the process comprising the following steps: a weakening step (S2) for forming a weakened zone (3) in a central portion (2c) of a donor substrate (2), the weakened zone (3) not extending into a peripheral portion (2p) of the donor substrate (2); a bonding step (S3) for bonding a main face (2a) of the donor substrate (2) to a receptor substrate (5) so as to form a component to be divided; and a separating step (S4) for separating the component to be divided, the separating step including a heat treatment which causes the thin layer (1) to be released only from the central portion (2c) of the donor substrate (2). The process further includes a disassembling step (S6) after the separating step, the disassembling step including treating the component to be divided so as to disassemble the peripheral portion of the donor substrate (2) from the receptor substrate (5).

Description

Technical Field

[0001] The present invention relates to a process for preparing a thin layer transferred to a carrier substrate by applying Smart Cut TM technology. Background Art

[0002] In some implementations of Smart Cut TM technology, a thin layer is prepared by introducing a light material through the main surface of a donor substrate to form a buried weakened zone. In this way, in the donor substrate, the thin layer is defined between the buried weakened zone and the main surface of the substrate. Then, the donor substrate is bonded to a second substrate, referred to as the "receptor" substrate, and a separation heat treatment is potentially applied to the assembly with the assistance of mechanical stress to cause a cleavage wave to be initiated and propagated at the buried weakened zone, thereby releasing the thin layer, which is thus transferred to the receptor substrate.

[0003] The role of the separation heat treatment is to promote the growth and pressurization of microcavities, the development of which is due to the presence of the light material in the weakened zone. This role is the source of the initiation and propagation of the cleavage wave, thereby causing the release of the thin layer.

[0004] The state of the exposed surface of the thin layer transferred to the receptor substrate is irregular, which is generally not entirely satisfactory. Therefore, processes implementing Smart Cut technology generally envision additional finishing steps aimed at reducing these surface irregularities. This may mean that in this finishing step, the exposed surface of the thin layer is treated by chemical mechanical polishing or by annealing at a high temperature (usually on the order of 1100 °C) while exposing the exposed surface of the thin layer to a reducing or neutral atmosphere.

[0005] Document FR3061988 proposes a process for finishing a thin layer prepared using Smart Cut technology, as described above. This document proposes that after the thin layer is released from the donor substrate, a moderate heat treatment below 950 degrees is directly applied to the thin layer in a reducing or neutral atmosphere. Since the released surface of the thin layer is not exposed to the ambient atmosphere, the released surface of the thin layer is not oxidized, and the atoms forming the released surface are highly mobile in order to reorganize themselves and smooth the surface. Therefore, even at relatively low temperatures, this heat treatment is particularly effective for smoothing the released surface of the thin layer. It should be noted that in the process proposed in FR3061988, the cleavage and smoothing treatments must be performed without contact with the ambient atmosphere between two steps (i.e., in the same equipment). This imposes extremely stringent limitations on the purity control of the atmosphere in the equipment and the heat capacity.

[0006] Document FR2876307 provides a similar heat treatment directly after the release of the thin layer from the donor substrate.

[0007] Document EP2933828 states that surface irregularities are the result of a combination of phenomena of various different natures. These irregularities can correspond to the traces of microcavities that cause the segmentation of the donor substrate. They contribute to the formation of roughness on the surface of the thin layer, the amplitude and wavelength of which are on the order of about 1 nm or several tens of nm.

[0008] These irregularities can also result from the interaction between the propagation of the segmentation wave during segmentation and the acoustic vibrations encountered in the component to be segmented. These vibrations are caused by the sudden release of energy during the segmentation of the thin layer. Then, depending on the instantaneous stress state of the material through which the segmentation wave is traveling, the segmentation wave is prone to a vertical deflection relative to the plane in which it is traveling. Reference can be made to the article “Crack Front Interaction with Self-Emitted Acoustic Waves”, Physical Review Letters, American Physical Society, 2018, 121.(19), pp.195501, which analyzes this mechanism in detail. The irregularities caused by this phenomenon are relatively large in size and constitute the non-uniformity of the thickness of the thin layer, the amplitude of which can reach the nanometer order of magnitude, and the wavelength of which reaches the millimeter order of magnitude or even the centimeter order of magnitude.

[0009] In addition, and as demonstrated in “Fracture dynamics in implanted silicon” published by Massy et al., Applied Physics Letters 107.9 (2015), the sudden release of energy, the oscillation and impact of the two substrates during segmentation may also cause damage to the separated surface. This damage may in particular be caused by the impact between these two surfaces or their relative movement.

[0010] More generally, after the weakening step and after the two parts have been separated from each other, the treatment of the carrier substrate carrying the thin layer and the remaining part of the donor substrate may also cause damage, such as scratches or particles, on the released surface of the thin layer.

[0011] It is difficult to remove surface irregularities other than roughness form from the thin layer only by heat treatment. Specifically, under the action of heat treatment, the size of these irregularities may exceed the average migration distance of the atoms forming the surface. This is especially the case in the smoothing treatment proposed in document FR3061988 cited above.

[0012] Thus, the thin layer prepared using the prior art process may have a damaged or imperfect surface state, and it is impossible to remove all irregularities from such a surface state.

[0013] The present invention seeks to overcome at least in part the above limitations of the prior art. Summary of the Invention

[0014] To achieve this objective, the subject matter of the present invention relates to a process for preparing a thin layer, the process comprising the following steps:

[0015] - A weakening step, which includes introducing a light substance into the main surface of the donor substrate, and this step is used to form a weakened area in the central part of the donor substrate, so as to define the thin layer by means of the main surface of the donor substrate, and this weakened area does not extend into the peripheral part of the donor substrate;

[0016] - A bonding step, which bonds the main surface of the donor substrate to the receptor substrate to form a component to be divided, so that the central part and the peripheral part of the main surface of the donor substrate are in surface contact with the surface of the receptor substrate;

[0017] According to the present invention, the process further includes:

[0018] - A separation step, which separates the component to be divided, and this separation step includes a heat treatment, which causes a splitting wave to be initiated and propagated in the weakened area so that the thin layer is released only from the central part of the donor substrate, and this splitting wave does not completely propagate through the peripheral part, such that the donor substrate and the receptor substrate remain attached to each other at the peripheral part of the donor substrate;

[0019] - A disassembling step, which is different from the separation step and is applied after the separation step, and this disassembling step includes treating the component to be divided so as to disassemble the peripheral part of the donor substrate (2) from the receptor substrate (5), and thereby transfer the thin layer (1) to the receptor substrate (5).

[0020] By positioning the weakened area only in the central part of the donor substrate, the thin layer can be released without causing complete disassembly of the donor substrate and the receptor substrate, which makes it possible to improve the surface state of the thin layer.

[0021] According to other advantageous and non-limiting features of the present invention (taken individually or in any technically feasible combination):

[0022] - The process includes a finishing step, which includes a heat treatment for bringing the component to be divided to a temperature higher than the temperature of the heat treatment in the separation step, so as to smooth the released surface of the thin layer.

[0023] - This finishing step is applied between the separation step and the disassembling step;

[0024] - The finishing step involves subjecting the donor substrate to a heat treatment at a temperature lower than 1000°C, lower than 950°C, lower than 900°C, or lower than 600°C.

[0025] - The separation step and the finishing step are performed in the same apparatus.

[0026] - The separation step and the finishing step are performed in different apparatuses.

[0027] - The finishing step is applied after the disassembling step.

[0028] - The donor substrate is a wafer having a circular shape, and the peripheral portion is an annular portion, and the width of the annular portion obtained from the edge of the substrate is between 1 mm and 5 cm.

[0029] - The donor substrate includes a peripheral chamfer on the main surface side of the donor substrate, and the peripheral portion extends from the chamfer by a width between 1 mm and 5 cm.

[0030] - The light substance is selected from the list formed by hydrogen and helium.

[0031] - The donor substrate is made of single crystal silicon.

[0032] - The separation step involves subjecting the donor substrate to a temperature between 250°C and 500°C.

[0033] - The disassembling step is performed by chemical etching or by applying a mechanical force to separate the carrier substrate and the donor substrate. Description of the Drawings

[0034] With reference to the accompanying drawings, further features and advantages of the present invention will become apparent from the following detailed description of the present invention, wherein:

[0035] Figure 1 Steps of the process according to the present invention are shown;

[0036] Figure 2 The central portion and the peripheral portion of the donor substrate are shown. Detailed Description of the Invention

[0037] With reference to Figure 1 , a process for preparing the thin layer 1 according to the present specification is now given.

[0038] The process includes: setting a donor substrate 2 in a preparatory step S1, the substrate having a main face 2a. The donor substrate 2 can be composed of a single-crystalline silicon wafer, especially in the case where the process for preparing the thin layer is intended to produce a silicon-on-insulator substrate. The wafer can undergo pre-treatment, such as oxidation, formation of a surface layer by epitaxy or by deposition or any other surface preparation step.

[0039] However, the present invention is in no way limited to a donor substrate 2 composed of a silicon wafer, and the nature and shape of this substrate can be freely chosen according to the target application. Thus, it can be a substrate including any single-crystalline semiconductor material, for example, silicon carbide or germanium. It can also be an insulating material such as a piezoelectric material, such as lithium tantalate or lithium niobate. Thus, the donor substrate can take the shape of a circular wafer with any diameter from a few millimeters to 300 mm or 450 mm or more. It can take another shape, such as square or rectangular. As specified above, the donor substrate 2 can undergo pre-steps such as polishing, deposition, oxidation, etc., and is thus formed by a stack of heterogeneous layers.

[0040] The donor substrate 2 can have a chamfered edge 2b (as is conventional in the field of semiconductor wafers) in order to prevent it from breaking during processing. The chamfer 2b can extend several millimeters beyond the periphery of the substrate.

[0041] As Figure 2 shown, the donor substrate 2 according to the present specification has a distinction between the following two parts: a central part 2c and a peripheral part 2p. These parts 2c, 2p extend from one face of the substrate to the other face, and project onto the main face 2a of the substrate in any shape, as long as the peripheral part 2p completely surrounds the central part 2c. Thus, as Figure 2 shown, the central part 2c can be projected onto the main face 2a of the donor substrate 2 (here in the shape of a circular wafer) so as to form a disk centered on this face, and the peripheral part 2p can be projected onto this main face so as to form an annular surface that is concentric with the disk defining the central part 2c and extends up to the edge of the substrate 2. The central part 2c and the peripheral part 2p are intended to come into contact with a receptor substrate in order to bond the two substrates to each other. When the donor substrate 2 has a chamfer 2b, the peripheral part 2p thus extends radially from the edge of the substrate towards the center of the substrate beyond the chamfer in order to allow this contact.

[0042] For example, the peripheral part 2p of the donor substrate 2 can have a width between 1 mm and several centimeters (for example 5 cm), which width is obtained starting from the edge of the chamfer when the chamfer is present, or otherwise starting from the edge of the substrate. Thus, in Figure 2In the case of the example shown, the donor substrate 2 is a circular wafer having a chamfer with a width of approximately 1 mm at its outer edge. The peripheral portion 2p is defined as an annular portion, and the width of this annular portion obtained from the inner edge of the chamfer ranges between 1 mm and 5 cm, such that this peripheral region can come into contact with a receptor substrate having at least the same dimensions. The portion provided inside the annular portion forms the central portion 2c of the donor substrate 2.

[0043] According to Figure 1 the description, the process according to the present specification includes a weakening step S2, which includes introducing a light substance through the main face 2a of the donor substrate 2. This step aims to form a weakened region 3 buried in the donor substrate 2. The weakened region 3 extends in a plane of the donor substrate 2 that is substantially parallel to the main face 2a of the substrate 2. The weakened region 3, together with the main face 2a of the donor substrate 2, defines at least a part of the thin layer 1 that is intended to be prepared. It is well known that light substances are usually selected from a list formed by hydrogen and helium, but this is not restrictive. These light substances are usually introduced into the donor substrate 2 by ion implantation, although other introduction methods are also possible (by plasma, by diffusion). For the case of implantation, the implantation energy defines the depth to which the ions penetrate into the material of the donor substrate 2 and thus defines the depth of the buried plane in which the weakened region 3 is located.

[0044] When the introduction is performed by ion implantation, it can be envisaged that the introduction is performed in a single step, for example, for implanting a single dose of hydrogen or a single dose of helium; or it is performed in multiple consecutive steps, for example, for implanting a dose of hydrogen and then implanting another dose of helium. Regardless of the nature and manner of introducing the substance into the donor substrate, the term "dose" refers to the total amount of the substance introduced into the unit area of the donor substrate, for example, the number of light substances implanted per cm 2

[0045] In the context of the present specification, the weakening step S2 is performed such that, at the end of this step, the weakened region 3 is positioned in the central portion 2c of the donor substrate 2 and does not extend into the peripheral portion 2p. As will be clear in the remainder of the present specification, by positioning the weakened region only in the central portion 2p, the thin layer can be released without causing the complete disassembly of the donor substrate and the receptor substrate to which the donor substrate will be bonded, which makes it possible to improve the surface state of the thin layer.

[0046] There are many methods that can be used to position the weakened region 3 only in the central portion 2c. Thus, and according to the first method, the weakening step S2 is performed by differentiating the dose introduced into the central portion 2c from the dose introduced into the peripheral portion 2p. Reference can be made to application FR3063176 in this regard.

[0047] ​Thus, the weakening step S2 may include: introducing a first dose of a light substance into the central portion 2c of the donor substrate 2, and introducing a second dose of the light substance, which is lower than the first dose, into the peripheral portion 2p of the substrate 2.

[0048] The first dose of the light substance is selected such that it is sufficient to release the thin layer 1 from the weakened zone provided in the central portion 2c of the donor substrate 1 in a later separation step S4. Thus, this dose results in the formation of a weakened zone 3 in the central zone 2c. Then, the second dose of the light substance introduced into the peripheral portion 2p is selected such that such disassembly is not allowed in the peripheral portion 2p. Thus, this dose does not result in the formation of a weakened zone in the sense of this specification. For complete clarity, the peripheral portion 2p of the donor substrate 2 may include light substances injected or introduced in any other way, but these substances do not have a sufficient concentration in the portion 2p to form a weakened zone in the sense of this application, and in particular, it may not cause the separation of the donor substrate 2 at the portion 2p.

[0049] By way of example, and according to Figure 1 a preferred implementation shown, the second dose of the light substance may be zero, and in this case, the light substance is only introduced into the central portion 2c of the donor substrate 2 and not into the peripheral portion 2p.

[0050] To achieve this, when the introduction of the light substance is performed by ion implantation, a shielding material that forms a barrier to the penetration of the light substance may be used to mask the peripheral portion 2p. The mask may be made of Teflon, aluminum, or carbon. The mask may also be formed by a sacrificial mask 4 made of resin, hard oxide, or nitride, which is at least provided at the peripheral portion 2p of the main surface 2a of the donor substrate 2, and it should be understood that the mask will be removed when the weakening step S2 is completed.

[0051] When the weakening step S2 contemplates multiple consecutive ion implantation stages, such as in the case of co-implantation of hydrogen and helium, the mask may only be present during some of these implantation stages.

[0052] According to a variant of this first method, the introduction of the light substance into the donor substrate 2 may be performed by scanning a beam comprising such a light substance over the main surface 2a of the donor substrate 2. In this implementation of the weakening step S2, the movement of the beam may be controlled such that the peripheral portion 2p of the donor substrate 2 is excluded from any implantation, or the dose of the substance introduced into the portion 2p is less than the dose of the substance introduced into the central portion 2c. Thus, in this variant, the mask may be omitted, which is advantageous.

[0053] The properties of the substances introduced into the central portion 2c and the peripheral portion 2p, as well as the exact dosage of these substances, are selected according to the properties of the donor substrate and according to the characteristics of the subsequent separation step S4.

[0054] For example, when the donor substrate 2 is made of single-crystalline silicon and the separation step involves subjecting the donor substrate 2 to a heat treatment at a temperature between 250 °C and 500 °C, the following conditions can be envisaged:

[0055] Only hydrogen

[0056] Central portion: between 4E16 at / cm 2 and 1E17 at / cm 2 in between

[0057] Peripheral portion: between 0 and 3E16 at / cm 2 in between

[0058] Hydrogen and helium

[0059] Central portion: H: between 0.5E16 at / cm 2 and 2E16 at / cm 2 in between; He: between 0.5E16 at / cm 2 and 2E16 at / cm 2 in between.

[0060] Peripheral portion: H: between 0 and 1E16 at / cm 2 in between; He: between 0.5E16 at / cm 2 and 2E16 at / cm 2 in between.

[0061] According to an alternative method, the weakening step S2 can include introducing the same dosage of a light substance into the central portion 2c of the donor substrate 2 and the peripheral portion 2p of this substrate 2 during a first sub-step. When the donor substrate 2 is made of silicon, this conventional dosage (e.g., about 5E16 at / cm 2An amount of hydrogen sufficient to form a weakened zone 3 extending in the central zone 2c and entering the peripheral zone 2p. In a second sub-step, the peripheral portion 2p of the donor substrate 2 is processed to render ineffective the light substances introduced into the peripheral portion 2p. This may include, for example, using a laser to process this peripheral portion in order to diffuse the light substances previously introduced into this zone and effectively remove the weakened zone 3 from the peripheral portion 2p. Alternatively, the peripheral portion may be processed to damage the weakened zone present in this portion, for example, by injecting a relatively heavy substance (such as a silicon substance). When the processing performed in this second sub-step is completed, the weakened zone 3 is only located in the central zone 2c of the donor substrate 2 and no longer extends into the peripheral zone 2p.

[0062] Regardless of the manner in which this weakening step S2 is implemented, what results is a donor substrate 2 including an embedded weakened zone 3 located in the central portion 2c of the donor substrate 2, which weakened zone 3 together with the main face 2a of the donor substrate 2 defines a thin layer 1, and the weakened zone 3 does not extend into the peripheral portion 2p.

[0063] In a subsequent bonding step S3 of the process according to the present specification, the thus-prepared donor substrate is bonded to the receptor substrate 5 by bringing the main face 2a of the donor substrate 2 into face contact with the face of the receptor substrate 5. The receptor substrate 5 can be of any nature and of any desired shape, provided that it has a face of sufficient size to contact at least a part of the central portion 2c and the peripheral portion 2p of the donor substrate 2.

[0064] This bonding operation can be carried out by any method, for example, by molecular adhesion, by applying an adhesive material to at least one of the faces to be bonded, or by eutectic bonding when the faces to be bonded have been pre-prepared to have a metallic surface. The bonding step S3 can also envisage conditioning the faces of the two substrates by cleaning, activation or any other preparation step in order to facilitate this bonding operation and fix the two substrates to each other with sufficient adhesion energy.

[0065] Alternatively, the bonding operation can include gradually forming the receptor substrate 5 by depositing the constituent material of the substrate 5 on the main face 2a of the donor substrate 2 at a moderate temperature.

[0066] The term "component to be divided" will be used in the remainder of the present specification to refer to the component formed by the donor substrate 2 and the receptor substrate 5 when this step S3 is completed.

[0067] In the next step of the process according to the present specification, the component to be divided is heat-treated during the separation step S4. The heat treatment causes the development of microcavities, the coalescence of these microcavities when the number of these microcavities is sufficient, and their pressurization via available light substances, as conventionally achieved in Smart Cut technology. However, as previously mentioned, the buried weakening zone 3 does not extend over the entire extent of the donor substrate 2. Thus, the heat treatment causes the thin layer 1 to be released from the donor substrate only at the central zone 2c of the donor substrate 2 and not at the peripheral zone 2p.

[0068] The heat treatment of the separation step S4 (which is typically between 200 °C and 500 °C when the donor substrate 2 is made of silicon) can be applied by placing the component to be divided in a conventional oven for a duration that can range from several minutes to several hours. In fact, and for productivity reasons, a large number of components to be divided are usually placed in a cassette, and the cassette is placed in a large-capacity oven or furnace in order to apply the heat treatment to the components to be divided together at a selected temperature and for a selected duration.

[0069] The division wave initiated in this separation step S4 does not propagate from one end of the donor substrate 2 to the other end, and the remaining part of the donor substrate (i.e., the donor substrate from which the thin layer has been taken) is not suddenly released from the receptor substrate, as is the case in the separation steps of the prior art. In particular, the division wave does not completely propagate into the peripheral part 2p of the donor substrate 2, and thus the peripheral part of this donor substrate remains firmly joined to the receptor substrate 5 at this part 2p. Figure 1 In the context of the described process, at the end of the separation step S4, the receptor substrate 5 and the donor substrate 2 thus also remain joined to each other even though the thin layer 1 has been effectively released. In the process of the present specification, the generation of sound waves, the oscillation and impact of these substrates are thus limited. The adverse interaction between the division wave and the sound waves is reduced or eliminated, and the impact or sliding of the substrates is also prevented, which in the processes of the prior art would cause irregularities or damage to occur at the surface of the thin layer, and which is difficult to reduce by applying only heat treatment.

[0070] It should be noted that the feature of the process according to which the division wave does not completely propagate into the peripheral part 2p of the donor substrate 2 (and thus the carrier substrate and the receptor substrate still remain joined to each other) is fully verifiable because in this case it is not possible to remove the remaining part of the donor substrate from the receptor substrate, i.e., to separate them from each other. This fully verifiable nature of this feature allows a person skilled in the art to determine the parameters of the steps of the process, in particular the extent of the peripheral zone, the doses of light species to be introduced into the central zone and the peripheral zone respectively, the intensity of the heat treatment in the separation step S4, etc., using several simple experiments and together with the examples provided in the present specification.

[0071] The space 6 between the released surface of the thin layer 1 and the released surface of the donor substrate 2 is filled with light substances that cause the release of the thin layer 1. These substances are confined within this enclosed and sealed space 6 because the cleavage does not propagate into the peripheral portion 2p of the donor substrate 2, and thus this peripheral portion of the donor substrate remains firmly bonded to the receptor substrate 5. Accordingly, this enclosed and sealed space 6 is also isolated from the surrounding atmosphere by the peripheral portion 2p of the donor substrate 2. As a result, external contaminants (e.g., oxygen) are prevented from penetrating into the space 6 and passivating the released surface of the thin layer 1, which would make it more difficult to smooth. It should be noted in this regard that the light substances (usually hydrogen or helium) introduced during the weakening step S2 that allow the cleavage of the donor substrate 2 are extremely pure and they do not interact with the atoms forming the released surface of the thin layer, thus passivating the released surface and restricting the surface mobility of the atoms, which results in an atmosphere particularly suitable for thermal smoothing.

[0072] To take advantage of this very specific state of the component to be cleaved after the separation step S4, a preferred implementation of the process according to the present specification contemplates applying a heat treatment during the finishing step S5, which is intended to bring the component to be cleaved to a temperature higher than the temperature of the heat treatment of the separation step S4 in order to smooth the released surface of the thin layer 1.

[0073] In this preferred implementation, the heat treatment exposes the component to be cleaved to a plateau temperature that is higher than the temperature between 200 °C and 500 °C during the separation step S4 when the donor substrate is made of silicon. Advantageously, the heat treatment remains moderate, below 1000 °C, or 950 °C, or even below 900 °C or 600 °C. It can last for a duration ranging from a few seconds to a few hours at the selected plateau temperature.

[0074] This heat treatment of the finishing step S5 causes the component to be cleaved to reach a treatment temperature, including the light substances confined within the enclosed space 6 delimited by the released surface of the thin layer 1 and the released surface of the donor substrate 2. These surfaces, which are not contaminated by reactive substances from, for example, the external atmosphere, are composed of atoms that are highly mobile due to the temperature increase, which enables the surfaces, in particular the released surface of the thin layer 1, to be smoothed in a highly efficient manner even at low temperatures below or equal to 1000 °C or 950 °C. It should be noted that this treatment of the component to be cleaved can be implemented in a device including any atmosphere because it is not this atmosphere that contributes to smoothing the surface of the thin layer 10. The device (e.g., a furnace) is intended only to raise the temperature of the component to be cleaved.

[0075] Thus, the experiments conducted by the applicant have shown that after a heat treatment in the finishing step S5 including exposing the component to be segmented to 950 °C for 2 minutes, a thin silicon layer 1 with a surface roughness of approximately 1.6 nm RMS can be obtained. In the absence of any finishing step and thus in the absence of any heat treatment at a temperature higher than the temperature reached during the separation step S4, the roughness of the thin layer was measured to be at 5.5 nm RMS. The effectiveness of the finishing step just described is clearly evident.

[0076] The finishing step S5 can be carried out in the same equipment (e.g., the same oven or the same furnace) as the equipment used for the separation step S4. However, contrary to the prior art solutions outlined in the introduction of this application, the finishing step S5 can also be carried out in another equipment different from the equipment used to perform the separation step S4. Specifically, the two substrates forming the component to be segmented remain attached to each other, and the risk of damaging the thin layer 1 during the processing of the component in order to store it or move it to another equipment is reduced. Additionally, since the release surface of the thin layer 1 is completely restricted, it cannot be exposed to the ambient atmosphere during the equipment change. Therefore, it is not oxidized or passivated, and when the finishing step is carried out in an environment other than the environment used in the separation step, all possibilities of its smoothing are retained even at low temperatures. Moreover, as previously mentioned, the equipment for carrying out the finishing step S5 in this implementation can be chosen very freely.

[0077] In a subsequent step, the component to be segmented undergoes a disassembly step S6, which aims to disassemble the donor substrate 2 (or more precisely, the remaining part of the donor substrate after the thin layer 1 has been removed from the donor substrate 2) from the receptor substrate 5, and thus complete the transfer of the thin layer 1 to the receptor substrate 5. It should be noted that this disassembly step S6 is clearly different from the separation step S4. In fact, a very advantageous feature of the process of this specification is that the propagation of the segmentation wave initiated in the separation step S4 and the propagation of the segmentation wave resulting in the release of the thin layer 1 according to this disassembly step S5 are clearly distinguished, thereby causing the donor substrate to be completely disassembled from the receptor substrate. In the conventional segmentation steps of the prior art, the propagation of the segmentation wave completely propagates through the plane of the donor substrate, disassembling the donor substrate from the receptor substrate in this single step.

[0078] The process implemented to obtain disassembly during step S6 can be changed: it can involve chemically etching (e.g., wet etching) the edges of the component to be divided in order to disassemble them at the bonding interface of the two substrates. To this end, the component to be divided can be positioned vertically so as to immerse a part of its edge in a solution of KOH or TMAH (when the donor substrate is made of silicon), or in a solution containing HF (when there is a silicon oxide layer at the interface between the donor substrate 2 and the receptor substrate 5). By rotating the component, the component is continuously processed all the time so as to disassemble the two substrates from each other. Alternatively, and as Figure 1 schematically shown, the process can include applying a blade at the bonding interface and inserting the blade between the two substrates in order to separate the component. More generally, it can involve applying any means or any force that causes the carrier substrate 5 and the donor substrate 2 to disassemble at their contact surfaces.

[0079] In any case, different from the splitting step of the prior art process, this disassembly step S6 can be perfectly controlled so that the risk of damaging the thin layer 1 can be minimized.

[0080] According to another implementation, after the separation step S4, a finishing step is not directly performed between this step S4 and the disassembly step S6, but the finishing step is performed after the disassembly step S6.

[0081] The finishing step S5 of this implementation can be a conventional finishing step. For example, when the thin layer 1 is made of silicon, the released surface of the thin layer 1 is exposed to a reducing or neutral atmosphere at a temperature between 950 °C or 1000 °C and 1100 °C. More generally, the exposed surface of the thin layer will be subjected to a neutral or reducing atmosphere, and its temperature is higher than the temperature of the heat treatment in the separation step.

[0082] In this implementation, the advantage that there are no or fewer large-sized irregularities or damages on the surface of the thin layer 1 is utilized. The roughness exhibited by the thin layer 1 can be effectively smoothed via the above-mentioned heat treatment.

[0083] When these processes are completed (regardless of the implementation selected), and as Figure 1 shown in the final step S7, the result is that the thin layer 1 is transferred to the carrier substrate 5. The thin layer 1 exhibits an improved surface state (compared with the thin layer 1 directly obtained after splitting using the prior art process). In particular, due to the smoothing effect of the finishing step S5, it exhibits a relatively low roughness. When the receptor substrate 5 and the donor substrate 2 are completely disassembled from each other by the propagation of the splitting wave, due to the interaction between the splitting wave and the sound wave propagating through the receptor substrate and the donor substrate, it also exhibits very little or no non-uniformity. The damage to the thin layer 1 caused by the impact between the two substrates or their relative sliding on each other is also limited.

[0084] Of course, the present invention is not limited to the described implementation, and its variations may fall within the scope of the present invention defined by the claims.

[0085] In particular, other steps in addition to the described steps can be envisaged. In particular, these may involve applying additional finishing processes to the thin layer, such as a sacrificial oxidation step, or performing polishing to improve the quality of the layer or adjust its thickness.

[0086] It is also possible to combine the two proposed implementations, and perform the finishing step S5 in the first stage via a first heat treatment between the separation step S4 and the disassembly step S6, and supplement the treatment by applying a second stage (e.g., a second heat treatment) to the thin layer 1 after the disassembly step S6.

Claims

1. A process for preparing a thin layer (1), the process comprising the following steps: - A weakening step (S2), the weakening step comprising introducing light substances through a main face (2a) of a donor substrate (2), the weakening step being for forming a weakened zone (3) in a central portion (2c) of the donor substrate (2) so as to define the thin layer (1) by means of the main face (2a) of the donor substrate (2), the weakened zone (3) not extending into a peripheral portion (2p) of the donor substrate (2); - A bonding step (S3), the bonding step bonding the main face (2a) of the donor substrate (2) to a receptor substrate (5) so as to form a component to be divided, a central portion (2c) and a peripheral portion (2p) of the main face (2a) of the donor substrate (2) being in contact with a face of the receptor substrate (5); The process is characterized in that the process further comprises: - A separation step (S4), the separation step separating the component to be divided, the separation step comprising a heat treatment which causes a division wave to be initiated and propagated in the weakened zone so as to release the thin layer (1) only from the central portion (2c) of the donor substrate (2), the division wave not completely propagating through the peripheral portion (2p) such that the donor substrate (2) and the receptor substrate (5) remain attached to each other at the peripheral portion (2p) of the donor substrate (2); - A disassembly step (S6), the disassembly step being different from the separation step (S4) and being applied after the separation step, the disassembly step comprising treating the component to be divided so as to disassemble the peripheral portion of the donor substrate (2) from the receptor substrate (5) and thereby transfer the thin layer (1) to the receptor substrate (5).

2. The process for preparing a thin layer (1) according to claim 1, the process comprising a finishing step (S5), the finishing step comprising a heat treatment which is for bringing the component to be divided to a temperature higher than the temperature of the heat treatment of the separation step (S4) so as to smooth a released surface of the thin layer (1).

3. The process for preparing the thin layer (1) according to claim 2, wherein, The finishing step (S5) is applied between the separation step (S4) and the disassembly step (S6).

4. The process for preparing the thin layer (1) according to claim 3, wherein, The finishing step (S5) effects a heat treatment which exposes the donor substrate (2) to a temperature lower than 1000 °C, lower than 950 °C, lower than 900 °C, or lower than 600 °C.

5. The process for preparing the thin layer (1) according to any one of claims 2 to 4, wherein, The separation step (S4) and the finishing step (S5) are carried out in the same equipment.

6. The process for preparing the thin layer (1) according to any one of claims 2 to 4, wherein, The separation step (S4) and the finishing step (S5) are carried out in different equipment.

7. The process for preparing the thin layer (1) according to claim 2, wherein, The finishing step (S5) is applied after the disassembly step (S6).

8. A process for preparing a thin layer (1) according to any one of claims 1 to 4, wherein, The donor substrate (2) is a wafer of circular shape and the peripheral portion (2p) is an annular portion, the width of the annular portion obtained from the edge of the donor substrate (2) being between 1 mm and 5 cm.

9. The process for preparing the thin layer (1) according to any one of claims 1 to 4, wherein, The donor substrate (2) includes a peripheral chamfer (2b) on the main surface (2a) side of the donor substrate, and the peripheral portion (2p) extends from the peripheral chamfer by a width between 1 mm and 5 cm.

10. The process for preparing the thin layer (1) according to any one of claims 1 to 4, wherein, The light substances are selected from a list formed by hydrogen and helium.

11. The process for preparing the thin layer (1) according to any one of claims 1 to 4, wherein, The donor substrate (2) is made of single-crystalline silicon.

12. The process for preparing the thin layer (1) according to any one of claims 1 to 4, wherein, The separation step (S4) is implemented by exposing the donor substrate (2) to a temperature between 250 °C and 500 °C.

13. The process for preparing the thin layer (1) according to any one of claims 1 to 4, wherein, The separating step is performed by chemical etching or by applying mechanical force to separate the receptor substrate (5) and the donor substrate (2).

Images