A method for manufacturing a homoepitaxial layer of silicon carbide capable of limiting the formation of BPD defects and related composite structure

Abstract

This invention relates to a method for homoepitaxially fabricating an active layer of single-crystal silicon carbide on a composite structure, the method comprising the following steps: 1) providing a composite structure comprising a growth layer of single-crystal silicon carbide extending in a principal plane and disposed on a carrier substrate, the growth layer being defined by an outer peripheral contour and having a first thickness along an axis perpendicular to the principal plane; 2) forming a local barrier within or above the growth layer, the local barrier extending along the outer peripheral contour at a distance from the outer peripheral contour and corresponding to a physical discontinuity in the growth layer selected from: - a protrusion caused by a material present on the growth layer, the material being different from the material of the growth layer; - a depression corresponding to an etched region of the growth layer; or - an amorphous region or a region having a different crystallinity than the rest of the growth layer; the thickness of the local barrier along the axis perpendicular to the principal plane is less than the first thickness; 3) epitaxially growing the active layer on the growth layer.

Description

Technical Field

[0001] This invention relates to the field of semiconductor materials, and particularly to composite structures comprising an active layer of silicon carbide (SiC) homoepitaxially grown on a free surface made of SiC. The invention further relates to a manufacturing method that enables the restriction of the formation of basal plane dislocations (BPDs) and Shockley stacking faults (SSFs) type extended defects. The invention also relates to a composite structure on which a high-quality active layer can be homoepitaxially grown. Background Technology

[0002] Silicon carbide (SiC) is a material particularly suitable for manufacturing power and radio frequency devices, or devices that operate at very high temperatures. To fabricate these devices, an active layer 150 of SiC (e.g., 4H or 6H polytype) is typically grown on the surface of a solid substrate 1 (made of single-crystal SiC) or a composite structure 100 comprising a thin layer 10 (made of single-crystal SiC) added on a carrier substrate 20 (advantageously lower-quality single-crystal or polycrystalline SiC), such as... Figure 1 As shown. The composite structure 100 can be made, in particular, by known thin-layer transfer techniques (e.g., SmartCut). TM )preparation.

[0003] During the epitaxial growth of the active layer 150, certain defects (referred to as initial defects) present on the surface of the solid substrate 1 or thin layer 10 of the composite structure 100 may lead to the formation of local inclusions in the SiC 3C polytype or off-center regions in SiC 4H. The stress field generated by the local inclusions induces the slip of basal plane dislocations (BPDs), thereby forming a complex network of Shockley stacking faults (SSFs) and partial dislocations in the epitaxial active layer. These defects are very serious because they can damage subsequently fabricated devices in the active layer through bipolar degradation mechanisms. Furthermore, each defect can easily propagate along a direction parallel to the epitaxial surface through thermal activation and electron-hole recombination. This makes the situation even worse, as each propagating defect can affect multiple devices.

[0004] The subject of this invention This invention relates to a method for fabricating a single-crystal silicon carbide active layer via homoepitaxial growth, the method enabling the limitation of the density of basal plane dislocations (BPDs) and Shockley stacking faults (SSFs) in the layer. The fabrication method specifically includes the step of forming a peripheral local barrier on or within the growth layer of the composite structure to prevent the propagation of inclusions generated in the active layer due to BPDs and SSFs. The invention also relates to a composite structure incorporating said local barrier. Summary of the Invention

[0005] This invention relates to a method for fabricating an active layer of single-crystal silicon carbide on a composite structure via homoepitaxial growth, the method comprising the following steps: 1) Provide a composite structure comprising a growth layer of monocrystalline silicon carbide extending in a principal plane and disposed on a carrier substrate, the growth layer being defined by an outer peripheral profile and having a first thickness along an axis perpendicular to the principal plane; 2) A local barrier is formed within or above the growth layer, the local barrier extending along the outer perimeter contour at a distance from the outer perimeter contour and corresponding to a physical discontinuity in the growth layer selected from the following: - Protrusions caused by a material present on the growth layer, the material being different from the material of the growth layer; - A depression corresponding to the etched area of ​​the growth layer; or - Amorphous regions or regions with different crystallinity from the rest of the growth layer; The thickness of the local barrier along the axis perpendicular to the main plane is less than the first thickness; 3) The active layer is epitaxially grown on the growth layer, and the active layer has a second thickness.

[0006] Other advantageous and non-limiting features according to the invention, individually or in any technically feasible combination: The initial thickness is between 50 nanometers and 1 micrometer; The location of the outer periphery of the growth layer is on average between 0.5 mm and 2 mm from the outer periphery of the carrier substrate of the composite structure; The location of the local barrier is greater than 0.1 mm, greater than 0.5 mm, greater than 1 mm, greater than 2 mm, greater than 3 mm, or even greater than 5 mm from any point on the outer periphery of the growth layer; The thickness of the local barrier is less than 70%, 50%, 30%, 20%, or even 10% of the first thickness; The depressions are created by mechanical wear, laser wear, wet etching, or dry etching of the growth layer; Amorphous regions, or regions with different crystallinity from the rest of the growth layer, are generated by laser amorphization or ion implantation. The protrusions are created by localized deposition of materials such as tungsten nitride or tungsten disilicide; The second thickness is greater than or equal to 5 micrometers, greater than or equal to 10 micrometers, or even greater than or equal to 30 micrometers.

[0007] The present invention also relates to a composite structure comprising a growth layer of single-crystal silicon carbide extending in a principal plane and disposed on a carrier substrate, the growth layer being defined by an outer peripheral profile and having a first thickness along an axis perpendicular to the principal plane; the composite structure is characterized in that it includes a local barrier within or above the growth layer, the local barrier extending along the outer peripheral profile at a distance from the outer peripheral profile and corresponding to a physical discontinuity in the growth layer selected from the following: - Protrusions caused by a material present on the growth layer, the material being different from the material of the growth layer; - A depression corresponding to the etched area of ​​the growth layer; or - Amorphous regions or regions with different crystallinity from the rest of the growth layer.

[0008] The thickness of the local barrier along the axis perpendicular to the main plane is less than the first thickness.

[0009] Other advantageous and non-limiting features according to the invention, individually or in any technically feasible combination: - The location of the local barrier is greater than 0.1 mm, greater than 0.5 mm, greater than 1 mm, greater than 2 mm, greater than 3 mm, or even greater than 5 mm from any point on the outer periphery of the growth layer; - The thickness of the local barrier is less than 70%, 50%, 30%, 20%, or even 10% of the first thickness; - The composite structure also includes an active layer of monocrystalline silicon carbide homoepitaxially grown on the growth layer, the active layer having a second thickness; - The second thickness is greater than or equal to 5 micrometers, greater than or equal to 10 micrometers, or even greater than or equal to 30 micrometers. Attached Figure Description

[0010] Further features and advantages of the present invention will become apparent from the following detailed description of the invention, taken in conjunction with the accompanying drawings, wherein: [ Figure 1 ] Figure 1 The study shows a solid substrate, a composite structure with an epitaxial layer, and a composite structure without an epitaxial layer. [ Figure 2 ] Figure 2 An example of the surface morphology of the active layer grown on the composite structure is shown; the dark areas correspond to extended defects (BPD, SSF, networks of partial dislocations). [ Figure 3 ] Figure 3The image shows an extended defect obtained by photoluminescence imaging: a jagged region can be seen in the lower left corner of the image (the black area is the annular part of the composite structure without a growth layer), in which inclusions cause extended defects; [ Figure 4a ] [ Figure 4b ] [ Figure 4c ] [ Figure 4d ] [ Figure 4e ] Figure 4a , Figure 4b , Figure 4c , Figure 4d and Figure 4e The various sub-steps for manufacturing the composite structure according to the method of the present invention are shown; [ Figure 5 ] Figure 5 An example of a local barrier formed above and / or within the growth layer of a composite structure according to the method of the present invention is shown. Detailed Implementation

[0011] The present invention relates to a method for fabricating an active layer 150 of single-crystal silicon carbide on a growth layer 10 of a composite structure 100 by homoepitaxial growth.

[0012] The applicant discovered that the extended defects caused by local inclusions formed during epitaxy at the initial defects in growth layer 10, as mentioned in the introduction, become extremely important for epitaxial active layers with a thickness greater than 10 micrometers, and critical for epitaxial active layers with a thickness greater than 30 micrometers. In composite structure 100, the initial defects can specifically correspond to: - Pores penetrating growth layer 10 caused by localized migration problems. - Serrated areas on the outline of the growth layer, or - Local deformation of the growth layer.

[0013] Figure 2 An example of the morphology of defects detected on the surface of a 30-micron active layer grown on composite structure 100 is shown: Note the very high density of extended defects that begin primarily at the contour of growth layer 10 and extend toward the center of the structure with thermal action. Figure 3 An example of an inclusion is shown, from which a BPD and SSF defect network extends outward, resulting in extended defects. This image was obtained using photoluminescence microscopy. The black area in the lower left corner of the image lacks growth layer 10, and the outline 10c of growth layer 10 is jagged. As mentioned above, these jagged areas constitute initial defects, which are prone to generating local inclusions during epitaxial growth, thus forming extended defects.

[0014] The method according to the invention aims to provide an improved active layer 150 by providing a local barrier 130 in the growth layer 10, which can prevent the propagation of BPD and SSF defects; the definition of the local barrier takes into account a specific aspect of the composite structure 100, namely that the growth layer 10 includes an irregular peripheral contour 10c, which is the main source of initial defects that are prone to cause fatal propagation defects after the extension of the active layer 150.

[0015] The first step of the method corresponds to providing a composite structure 100 comprising a single-crystal silicon carbide growth layer 10 extending in a principal plane (x, y) and disposed on a carrier substrate 20. The preferred form of the composite structure 100 is a circular wafer with a diameter of 100 mm, 150 mm, 200 mm, or larger. However, it may also take any other form for subsequent processing to manufacture components. The thickness of structure 100 extends along the z-axis in the figure.

[0016] As mentioned in the introduction, the composite structure 100 can be constructed using layer transfer techniques (such as SmartCut). TM )preparation.

[0017] In the first sub-step a), a donor substrate 1 and a support substrate 20 of single-crystal silicon carbide are provided. Figure 4a The single-crystal SiC can be 4H, 6H, or 3C polytype. The donor substrate 1 is preferably in wafer form, with a diameter that is the same as or very similar to the diameter of the carrier substrate 20 to which it will be subsequently assembled, and a thickness typically in the range of 300 micrometers to 800 micrometers. It has a front surface 1a and a back surface 1b. The surface roughness of the front surface 1a is advantageously selected to be less than 1 nanometer RMS, or even less than 0.5 nanometer RMS, which is measured by atomic force microscopy (AFM) within a 20 micrometer × 20 micrometer scanning range. The doping type and resistivity of the donor substrate 1 depend on the application and the target device. The carrier substrate 20 corresponds to the mechanical support of the future composite structure 100. It is advantageously formed from polycrystalline silicon carbide (p-SiC) or single-crystal silicon carbide (SiC) of lower crystallinity. Its electrical properties (doping level and type) can also be selected according to the intended application.

[0018] The second sub-step (b) includes injecting a light material into the donor substrate 1 to form an embedded brittle surface 11, which defines the surface layer 10' to be transferred with the front surface 1a of the donor substrate 1. Figure 4bThe light material, preferably hydrogen and / or helium, is injected into the donor substrate 1 to a depth consistent with the desired thickness of the growth layer 10. This light material will form microcavities near the given depth, distributed in a thin layer parallel to the free surface 1a of the donor substrate 1, i.e., parallel to the plane (x, y) in the figure. For simplicity, this thin layer is referred to as the embedded brittle surface 11. The injection energy of the light material is selected to achieve the desired depth. For example, the energy level of hydrogen ion injection ranges from 10 keV to 250 keV, and the dose ranges from 5... E 16 / cm 2 and 1 E 17 / cm 2 This defines a surface layer 10' with a thickness of approximately 100 nanometers to 1500 nanometers. Note that a protective layer may be deposited on the front surface 1a of the donor substrate 1 prior to the ion implantation step. For example, this protective layer may consist of a material such as silicon oxide or silicon nitride. It may be removed before the next sub-step.

[0019] The third sub-step c) corresponds to assembling one side of the front surface 20a of the carrier substrate 20 with one side of the front surface 1a of the implanted donor substrate 1. Figure 4c The lateral dimensions (specifically its diameter) of the carrier substrate 20 in the principal plane (x, y) are the same as those of the composite structure 100. The thickness of the carrier substrate 20 is typically between about 50 micrometers and several hundred micrometers, for example, between 50 micrometers and 650 micrometers, or between 100 micrometers and 450 micrometers, or even between 200 micrometers and 350 micrometers.

[0020] Assembly is achieved through molecular adhesion along the bonding interface 40, i.e., direct bonding. Optionally, an intermediate layer may be formed on the front surface 1a of the donor substrate 1 before or after the introduction of the light material, and in any case before the assembly stage. This intermediate layer may be made of a dielectric material, a semiconductor material, or a metallic material (e.g., silicon oxide, silicon, silicon carbide, tungsten, titanium, etc.). Optionally, an intermediate layer may also be deposited on the assembly surface 20a of the carrier substrate 20 prior to assembly; it may be selected to have the same or different properties as the intermediate layer mentioned for the donor substrate 1. The intermediate layer may optionally be deposited on either of the two substrates 1 and 20 to be assembled. After assembly, it is intended to embed one or more intermediate layers into the bonding assembly 50 and ultimately into the composite structure 100.

[0021] Direct bonding via molecular adhesion does not require adhesive materials because the bonds are established at the atomic scale between the assembled surfaces. Several types of molecular adhesion bonding exist, which vary significantly in terms of temperature, pressure, or atmospheric conditions or pretreatment before surface contact. Room temperature bonding (regardless of whether the surfaces to be assembled are pre-plasma activated), atomic diffusion bonding (ADB), surface activated bonding (SAB), etc., can be used as references.

[0022] The assembly sub-steps may include: performing conventional sequential chemical cleaning (e.g., RCA cleaning), surface activation (e.g., by oxygen or nitrogen plasma), or other surface preparation (e.g., cleaning by scrubbing) before bringing the surfaces 1a and 20a to be assembled into contact. These steps are conducive to improving the quality of the bonding interface 40 (low defect rate, high adhesion energy).

[0023] It should be noted that there are chamfers at the outer peripheral edge 20c of the carrier substrate 20 and the outer peripheral edge of the donor substrate 1, which will form an unbonded peripheral ring. Figure 4c (Not shown in the image), surface layer 10' will not be transferred to the outer ring.

[0024] The fourth sub-step d) corresponds to separation along the embedded brittle surface 11 to form an intermediate composite structure 100', which includes, on one hand, the transferred surface layer 10' and the carrier substrate 20, and on the other hand, the remaining portion 1' of the donor substrate. Figure 4d Separation along the embedded brittle surface 11 is typically achieved by heat treatment at a temperature between 800°C and 1200°C. This heat treatment causes cavities and microcracks to appear in the embedded brittle surface 11, subjecting it to pressure from a light substance in a gaseous state until the cracks propagate along the brittle surface 11. Alternatively or in combination, mechanical stress may be applied to the bonded assembly, particularly to the embedded brittle surface 11, to propagate or facilitate the mechanical propagation of the cracks leading to separation.

[0025] After separation, the free surface 10'a of the surface layer 10' is generally rough: for example, it exhibits a roughness ranging from 5 nm RMS to 100 nm RMS. The surface layer 10' is defined in the plane (x, y) by a peripheral profile 10'c defined rearward from the outer peripheral edge 20c of the carrier substrate 20. In particular, the surface layer 10' does not shift at the (potential) chamfer at the outer peripheral edge 20c of the carrier substrate 20: the peripheral profile 10'c is generally defined rearward from the chamfered region of the edge 20c, in other words, offset towards the center of the substrate.

[0026] As mentioned above, the outer periphery profile 10'c can be serrated and irregular; on average, the distance between it and the outer periphery edge 20c (the width of the unbonded outer periphery ring) is between 0.5 mm and 2 mm, depending on the chamfer profile of the outer periphery edge 20c, the physical properties of the assembled substrate, the bonding type, etc.

[0027] Finally, the fifth sub-step e) includes subjecting the free surface 10'a of the surface layer 10' to one or more thermal, mechanical, and / or chemical treatments to form a composite structure 100 having a growth layer 10 of monocrystalline silicon carbide. Figure 4e Specifically, sub-step e) may include chemical mechanical polishing of the free surface 10'a of surface layer 10'. Removal of material, for example, between 50 nm and 300 nm, enables effective restoration of the surface finish of the layer, typically resulting in a roughness of less than or equal to 0.5 nm RMS, or even less than or equal to 0.1 nm RMS (10 x 10 μm). 2 Or 20x20 μm 2 (AFM scan). Sub-step e) may also include at least one heat treatment at a temperature ranging from 1200°C to 1800°C. This heat treatment is used to remove residual light material in the surface layer 10' and promote its lattice rearrangement, thereby forming a high-quality growth layer 10. It can also enhance the bonding interface 40.

[0028] In this stage of the method, the first thickness of the growth layer 10 of the composite structure 100 is typically in the range of tens to hundreds of nanometers, for example, between 50 nanometers and 1000 nanometers. The growth layer 10 is defined by an outer peripheral contour 10c, which is irregular, for example, serrated. Typically, the outer peripheral contour 10c is located at an average distance of 0.5 mm to 2 mm from the outer peripheral edge 20c of the carrier substrate 20 of the composite structure 100.

[0029] The method according to the invention then includes a second step corresponding to forming a local barrier 130 within or above the growth layer 10. This local barrier corresponds to a physical discontinuity in the growth layer 10 and extends continuously along the outer perimeter profile at a distance from the outer perimeter profile 10c. Figure 5 (i) and Figure 5 (ii) Preferably, its location is greater than 0.1 mm, greater than 0.5 mm, greater than 1 mm, or even greater than 2 mm, greater than 3 mm, or even greater than 5 mm from any point on the outer periphery contour 10c of the growth layer 10.

[0030] Within the principal plane (x, y), the barrier 130 may extend along the contour line of the outer perimeter contour 10c, or along a fundamentally different contour line, such as more regular (perfectly circular) or more irregular, to bypass specific initial defects present in the vicinity of the outer perimeter contour 10c adjacent to the growth layer 10.

[0031] According to the first embodiment, the physical discontinuity can take the form of a protrusion caused by the presence of material disposed on the growth layer 10 (the material being different from the material of the growth layer 10). For example, such a protrusion can be generated by localized deposition of a material such as tungsten nitride or tungsten disilicide (WSi2).

[0032] According to the second embodiment, the physical discontinuity of the local barrier 130 can be characterized as a depression, which can be generated, in particular, by locally mechanically abrading (sawing, grinding), laser abrasion, wet etching, or dry etching of the growth layer 10. Figure 5 (ii) and Figure 5 (iii)

[0033] Finally, according to the third embodiment, the physical discontinuity may correspond to an amorphous region or a region with a different degree of crystallinity from the rest of the growth layer 10, said region being generated, for example, by laser amorphization or ion implantation.

[0034] The thickness of the local barrier 130 along the z-axis perpendicular to the principal plane (x, y) is less than the first thickness, specifically less than 70%, 50%, 30%, 20%, or even 10% of the first thickness. Its width in the principal plane (x, y) and in the radial direction can range from 0.1 micrometers to 1000 micrometers.

[0035] The applicant has determined that the aforementioned features of the local barrier 130 prevent basal plane dislocations (BPDs) and Shockley stacking faults (SSFs) from extending from local inclusions of the SiC 3C polytype or the SiC 4H off-center region, which are prone to form during epitaxy at irregularities in the outer contour 10c of the growth layer 10. Therefore, the presence of this local barrier 130 prevents fatal defects of the BPD and SSF types from extending from the contour 10c toward the center of the growth layer 10, thereby enabling the acquisition of a high-quality active layer 150 after epitaxial growth.

[0036] The physical discontinuity of the growth layer 10 formed by the local barrier 130 effectively prevents the propagation of BPD and SSF defects. Although not limited to this interpretation, it can be considered that the thickness of the barrier 130 (less than the thickness of the growth layer 10) is an important parameter for preventing new defects from extending radially from the barrier 130.

[0037] The method according to the invention finally includes a third step, which corresponds to the epitaxial growth of an active layer 150 on a growth layer 10, the active layer having a second thickness that is typically greater than or equal to 5 micrometers, greater than or equal to 10 micrometers, or even greater than or equal to 30 micrometers.

[0038] This silicon carbide epitaxial growth is carried out within a conventional temperature range (i.e., between 1500°C and 1900°C).

[0039] The physical discontinuity formed by the local barrier 130 leads to locally interrupted epitaxial growth: Specifically, the barrier 130 extends along the thickness direction as the active layer 150 grows, taking the form of a “wall” including stacking faults (SF), which helps to prevent the slippage of BPD dislocations.

[0040] However, the interruption of epitaxial growth associated with barrier 130 remains localized, while the remainder of active layer 150 grows according to the crystal structure of growth layer 10.

[0041] exist Figure 5 (iii) and Figure 5 In the embodiment shown in (iv), the composite structure 100 provided in the first step of the method comprises a growth layer 10 of c-SiC with a thickness of 600 nanometers, placed on a carrier substrate 20 of p-SiC with a thickness of 350 micrometers. The local barrier 130 generated during the second step of the method corresponds to a depression formed by local etching through mechanical abrasion: its depth is approximately 300 nanometers and its width is approximately 2 micrometers. It is located 1 millimeter away from the outer peripheral contour 10c of the growth layer 10. As in Figure 5 As seen in (iii), a hole (initial defect) exists above the barrier 130, penetrating the growth layer 10. In the third step of the method, a 30-micrometer active layer 150 is generated by homoepitaxial growth at a temperature of 1550°C. Local inclusions are formed at the initial defects. Figure 5 (iv) ): BPD and SSF type extended defects appear on either side of the package. Barrier 130 prevents the defects from extending toward the center (downward in the figure) of the active layer 150.

[0042] Of course, the present invention also relates to a composite structure 100, which includes a growth layer 10 disposed on a carrier substrate 20, and a local barrier 130 formed on or within the growth layer 10. The present invention also relates to a composite structure 100 having an active layer 150 homoepitaxially grown on the growth layer 10, wherein the local barrier 130 of the growth layer 10 enables the acquisition of an active layer 150 of excellent quality.

[0043] Of course, the present invention is not limited to the described embodiments and examples, and alternative embodiments may be applied without departing from the scope of the invention as defined by the claims.

Claims

1. A method for fabricating an active layer (150) of single-crystal silicon carbide on a composite structure (100) by homoepitaxial growth, the method comprising the following steps: 1) A composite structure (100) is provided, the composite structure (100) including a growth layer (10) of single crystal silicon carbide extending in a principal plane (x, y) and disposed on a carrier substrate (20), the growth layer (10) being defined by a peripheral profile (10c) defined rearwardly from the outer peripheral edge (20c) of the carrier substrate (20) and having a first thickness along an axis (z) perpendicular to the principal plane (x, y), the composite structure (100) including a bonding interface (40) between the growth layer (10) and the carrier substrate (20). 2) A local barrier (130) is formed within or above the growth layer (10), the local barrier (130) extending along the outer periphery at a distance from the outer periphery (10c) and corresponding to a physical discontinuity in the growth layer (10) selected from the following: - Protrusions caused by a material present on the growth layer (10), the material being different from the material of the growth layer; - A depression corresponding to the etched area of ​​the growth layer (10); or - Amorphous regions or regions with different crystallinity from the rest of the growth layer (10), The thickness of the local barrier (130) along the axis (z) perpendicular to the principal plane (x, y) is less than the first thickness; 3) The active layer (150) is epitaxially grown on the growth layer (10) having a local barrier (130), and the active layer (150) has a second thickness.

2. The manufacturing method according to claim 1, wherein, The first thickness is between 50 nanometers and 1 micrometer.

3. The manufacturing method according to any one of claims 1 and 2, wherein, The location of the outer periphery contour (10c) of the growth layer (10) is on average between 0.5 mm and 2 mm from the outer periphery edge (20c) of the carrier substrate (20) of the composite structure (100).

4. The manufacturing method according to any one of claims 1 to 3, wherein, The location of the local barrier (130) is greater than 0.1 mm, greater than 0.5 mm, greater than 1 mm, greater than 2 mm, greater than 3 mm, or even greater than 5 mm from any point on the outer periphery contour (10c) of the growth layer (10).

5. The manufacturing method according to any one of claims 1 to 4, wherein, The thickness of the local barrier (130) is less than 70%, 50%, 30%, 20%, or even 10% of the first thickness.

6. The manufacturing method according to any one of claims 1 to 5, wherein, The depression is created by mechanical wear, laser wear, wet etching or dry etching of the growth layer (10).

7. The manufacturing method according to any one of claims 1 to 5, wherein, The amorphous regions, or regions with different crystallinity from the rest of the growth layer (10), are generated by laser amorphization or ion implantation.

8. The manufacturing method according to any one of claims 1 to 5, wherein, The protrusions are created by localized deposition of materials such as tungsten nitride or tungsten disilicide.

9. The manufacturing method according to any one of claims 1 to 8, wherein, The second thickness is greater than or equal to 5 micrometers, greater than or equal to 10 micrometers, or even greater than or equal to 30 micrometers.

10. A composite structure (100) comprising a single-crystal silicon carbide growth layer (10) extending in a principal plane (x, y) and disposed on a carrier substrate (20), the growth layer (10) being defined by a peripheral profile (10c) defined rearwardly from an outer peripheral edge (20c) of the carrier substrate (20) and having a first thickness along an axis (z) perpendicular to the principal plane (x, y), the composite structure (100) comprising a bonding interface (40) between the growth layer (10) and the carrier substrate (20), characterized in that, The composite structure includes a local barrier (130) within or above the growth layer (10), the local barrier extending along the outer perimeter contour at a distance from the outer perimeter contour (10c) and corresponding to a physical discontinuity in the growth layer (10) selected from the following: - Protrusions caused by a material present on the growth layer (10), the material being different from the material of the growth layer; - A depression corresponding to the etched area of ​​the growth layer (10); or - Amorphous regions or regions with different crystallinity from the rest of the growth layer (10), The thickness of the local barrier (130) along the axis (z) perpendicular to the principal plane (x, y) is less than the first thickness.

11. The composite structure (100) according to claim 10, wherein, The location of the local barrier (130) is greater than 0.1 mm, greater than 0.5 mm, greater than 1 mm, greater than 2 mm, greater than 3 mm, or even greater than 5 mm from any point on the outer periphery contour (10c) of the growth layer (10).

12. The composite structure (100) according to any one of claims 10 and 11, wherein, The thickness of the local barrier (130) is less than 70%, 50%, 30%, 20%, or even 10% of the first thickness.

13. The composite structure (100) according to any one of claims 10 to 12, further comprising an active layer (150) of single-crystal silicon carbide homoepitaxially grown on the growth layer (10), the active layer (150) having a second thickness.

14. The composite structure (100) according to claim 13, wherein, The second thickness is greater than or equal to 5 micrometers, greater than or equal to 10 micrometers, or even greater than or equal to 30 micrometers.

Images