Silicon carbide-based composite structure having good vertical electrical conduction

Abstract

The invention relates to a composite structure having a front face and a rear face extending parallel to a main plane, which structure comprises: - a useful layer of single-crystal silicon carbide, a free face of which constitutes the front face, which layer has a first concentration of N-type dopants; - a carrier substrate made of polycrystalline silicon carbide, a free face of which constitutes the rear face, which substrate has a second concentration of N-type dopants, the second concentration being higher than the first concentration; - an intermediate region extending along the main plane and including an interface zone between an assembled face of the useful layer and an assembled face of the carrier substrate; the composite structure being characterised in that the intermediate region comprises inclusions of single-crystal silicon carbide in direct contact with the useful layer and extending, in a direction normal to the main plane, between grains of the carrier substrate, the inclusions having a third concentration of N-type dopants which is between the first concentration and the second concentration. The invention also relates to a method for producing such a composite structure.

Description

SILICON CARBIDE-BASED COMPOSITE STRUCTURE WITH GOOD VERTICAL ELECTRICAL CONDUCTION FIELD OF THE INVENTION The present invention relates to the field of microelectronics and semiconductors. In particular, the invention relates to a composite structure comprising a useful layer of monocrystalline silicon carbide, a support substrate of polycrystalline silicon carbide, and an intermediate region, between the useful layer and the support substrate, promoting vertical electrical conduction. TECHNOLOGICAL BACKGROUND OF THE INVENTION Silicon carbide (SiC) is increasingly used for the manufacture of high-performance power devices. However, high-quality monocrystalline SiC (c-SiC) substrates for the microelectronics industry remain expensive and difficult to source in large quantities. It is therefore advantageous to use layer transfer solutions to develop composite structures typically comprising a thin c-SiC layer (derived from a high-quality c-SiC substrate and intended to accommodate the sensitive functional parts of the devices) on a lower-cost support substrate, for example polycrystalline SiC (p-SiC). It should also be noted that a composite structure can provide additional functionalities and performances, in particular by providing a support substrate with advantageous mechanical, electrical and / or thermal properties. A well-known thin-film transfer solution is the Smart Cut® process, based on the implantation of light ions into a donor substrate (c-SiC) and on assembly, by direct bonding, at a bonding interface between the donor substrate and a support substrate (e.g. p-SiC). In the field of power electronics, vertical components require good electrical conduction between the thin layer (also called useful layer) and the support substrate of the composite structure. In particular, the bonding interface between a useful layer and a support substrate must have as low a resistivity as possible, preferably less than 1 mohm.cm 2 , or even less than 0.1mohm.cm 2 . Document WO2022 / 008809 proposes to form an interface zone from a very thin layer of a metallic material, for example tungsten, in a SiC / SiC semiconductor structure. Considering a useful layer and a support substrate in monocrystalline SiC with a resistivity of 20 mohm.cm, and a layer of 1 nm to 3 nm of tungsten (W) sandwiched between the useful layer and the support substrate, it is possible to obtain a resistivity of the interface zone less than or equal to 0.1 mohm.cm 2 , after application of an adequate heat treatment and formation of very fine nodules largely made of metal (W), responsible for the good vertical conduction of the interface area. Document WO2022 / 129726 proposes forming an interface zone, in a SiC / SiC semiconductor structure, comprising regions of direct contact between the useful layer and the support substrate, and agglomerates comprising a semiconductor material different from that or those of the useful layer and the support substrate, and having a thickness less than or equal to 250 nm. The semiconductor material is chosen for its particular affinity with oxygen. Thus, the agglomerates are capable of effectively trapping any oxygen possibly present at the bonding interface; the regions of direct contact between the useful layer and the support substrate, devoid of native oxide residues in particular, allow efficient and good quality electrical conduction and / or vertical semiconductor / semiconductor contact. Considering a useful layer and a support substrate in c-SiC with a resistivity of 20 mohm.cm, and a silicon semiconductor material, it is possible to obtain a resistivity of the interface zone less than or equal to 0.1 mohm.cm. 2 , for example 0.032 mohm.cm 2 (after annealing the structure at 1370°C) or 0.0076 mohm.cm 2 (after annealing at very high temperature, namely 1900°C). SUBJECT OF THE INVENTION The present invention proposes an alternative solution particularly favorable to obtaining very low interface resistivities in composite structures whose useful layer is made of c-SiC and the support substrate of p-SiC, and whose doping levels are not necessarily extremely high. The invention relates to a composite structure comprising an intermediate region including an interface zone between an assembled face of the useful layer and an assembled face of the support substrate, and comprising inclusions of monocrystalline silicon carbide in direct contact with the useful layer and extending between grains of the support substrate. These inclusions are particularly favorable to vertical electrical conduction. The invention also relates to a method for manufacturing said composite structure. BRIEF DESCRIPTION OF THE INVENTION The present invention relates to a composite structure having a front face and a rear face extending parallel to a main plane, and comprising: - a useful layer of monocrystalline silicon carbide, a free face of which constitutes the front face, and having a first concentration of N-type dopants, - a support substrate made of polycrystalline silicon carbide, a free face of which constitutes the rear face, and having a second concentration of N-type dopants, the second concentration being greater than the first concentration, - an intermediate region extending along the main plane and including an interface zone between an assembled face of the useful layer and an assembled face of the support substrate. The composite structure is remarkable in that the intermediate region comprises inclusions of monocrystalline silicon carbide in direct contact with the useful layer and extending, in a direction normal to the main plane, between grains of the support substrate, said inclusions having a third concentration of N-type dopants between the first and second concentrations. According to advantageous characteristics of the invention, taken alone or in any feasible combination: the support substrate, at least on the side of its assembled face, comprises silicon carbide grains having a size less than or equal to 500 nm, said grains covering in the main plane, a surface greater than or equal to 20% of the assembled face; the silicon carbide grains having a size less than 500 nm cover, in the main plane (x,y), a surface greater than or equal to 50%, or even 80% of the assembled face of the support substrate; at least 20% of the silicon carbide grains, on the side of the assembled face of the support substrate, have a disorientation, between them or with the orientation

[0111] , greater than or equal to 30°; the intermediate region comprises silicon nodules, arranged in the interface zone or under the inclusions between grains of the support substrate; the first concentration is between 1.10 18 / cm 3 and 5.10 18 / cm 3 ;the second concentration is between 1.10 20 / cm 3 and 5.10 20 / cm 3 , preferably between 1.10 20 / cm 3 and 2.10 20 / cm 3 ;the inclusions have a thickness, along an axis normal to the principal plane, of between 20 nm and 70 nm, and a width, in the principal plane, of between 100 nm and 700 nm;the intermediate region has a density of inclusions, in the principal plane, of between 0.1 / μm 2 and 1 / μm 2 . The invention also relates to a method of manufacturing a composite structure as above, comprising the following steps: a) providing a monocrystalline silicon carbide donor substrate having a first concentration of N-type dopants, and comprising a buried fragile plane induced by ion implantation of light species; (b) the provision of the polycrystalline support substrate; c) the deposition of a silicon film on a face to be assembled of the donor substrate and / or on the face to be assembled of the support substrate; d) assembling the donor substrate and the support substrate at their faces to be assembled, so as to obtain a bonded assembly comprising a silicon interlayer, derived from the film(s) deposited in step c), between the donor substrate and the support substrate, said interlayer having a thickness of between 8 nm and 30 nm; e) separation along the buried fragile plane to form a transferred structure comprising the useful layer, transferred from the donor substrate, arranged on the interlayer, itself arranged on the support substrate; (f) applying a heat treatment to the transferred structure, at a holding temperature between 1500°C and 2000°C, to form the composite structure. According to advantageous characteristics of the invention, taken alone or in any feasible combination: the support substrate comprises, at least on the side of a face to be assembled, silicon carbide grains having a size less than or equal to 500 nm, said grains covering, in the main plane, a surface greater than or equal to 20% of the face to be assembled; at least 20% of the silicon carbide grains, on the side of the assembled face of the support substrate, have a disorientation, between them or with the

[0111] orientation, greater than or equal to 30°; the plateau temperature is maintained for a period of between 30 min and 4 h; the treatment of step f) has a temperature rise ramp less than or equal to 20°C / min, or even less than or equal to 10°C / min; the second concentration is of the order of 1.10 20 / cm 3 , the first concentration is of the order of 3.10 18 / cm 3, and the silicon carbide grains having a size less than or equal to 500 nm cover, in the main plane, a surface greater than or equal to 50%, or even 60% of the face to be assembled of the support substrate; the second concentration is of the order of 1.10 20 / cm 3 , the first concentration is of the order of 1.10 18 / cm 3 , and the silicon carbide grains having a size less than or equal to 500 nm cover, in the main plane, a surface greater than or equal to 80% of the face to be assembled of the support substrate. Other characteristics and advantages of the invention will emerge from the detailed description which follows with reference to the appended figures in which: The present invention presents a composite structure in accordance with the invention; Illustrates the intermediate region of a composite structure in accordance with the invention, on the left in the figure schematically, on the right with a transmission electron microscopy image; Figures 3a to 3f show steps of a method of manufacturing a composite structure, in accordance with the present invention; The intermediate region of a composite structure according to the invention is shown schematically on the left in the figure, and on the right with a transmission electron microscopy image; The present intermediate region of a composite structure in accordance with the invention, with a transmission electron microscopy image; Presents theoretical curves (numerical simulations) linking the resistivity of the intermediate region of a composite structure as a function of the percentage of surface area of ​​the support substrate in which the p-SiC grains have a size less than or equal to 500 nm, for different doping concentrations of the useful layer (N monoSiC ) and the support substrate (N polySiC ). Some figures are schematic representations which, for readability purposes, are not to scale. In particular, the layer thicknesses along the z axis are not to scale with the lateral dimensions along the x and y axes. The same references in the figures or in the description may be used for elements of the same nature. DETAILED DESCRIPTION OF THE INVENTION The invention relates to a composite structure 100 and an associated manufacturing method. The composite structure 100 has a front face 100a and a rear face 100b extending parallel to a main plane (x,y). It comprises a useful layer 10 of monocrystalline silicon carbide, preferably of polytype 4H, a free face of which constitutes the front face 100a, and a support substrate 20 of polycrystalline silicon carbide 3C, a free face of which constitutes the rear face 100b (). The composite structure 100 also comprises an intermediate region 30 extending along the main plane (x,y) and including an interface zone between an assembled face of the useful layer 10 and an assembled face of the support substrate 20. The intermediate region 30 comprises inclusions 31 of monocrystalline silicon carbide in direct contact with the useful layer 10 and extending, in a direction (z) normal to the main plane (x,y), between grains 21i, 21j of the support substrate 20 (). To facilitate understanding of the particular properties of the composite structure 100, linked in particular to this intermediate region 30, we will describe, firstly, the manufacturing method of said structure 100. The method according to the invention comprises a step a) of providing a donor substrate 1 made of monocrystalline silicon carbide (for example of 4H polytype) having a given concentration of N-type dopants (called the first concentration). Typically, the first concentration is between 1.10 18 / cm 3 and 5.10 18 / cm 3 , corresponding to a resistivity range between 20 and 30 mohm.cm. The donor substrate 1 comprises a front face 1a intended to be assembled, and a rear face 1b (). It comprises a buried fragile plane 11 induced by ion implantation of light species (in reference to the Smart Cut process). These species are for example hydrogen, helium or a combination of these two species. The buried fragile plane delimits, with the front face 1a of the donor substrate 1, the useful layer 10 to be transferred. The method comprises a step b) of providing a support substrate 20 made of polycrystalline silicon carbide (polytype 3C). The support substrate 20 comprises a front face 20a intended to be assembled and a rear face 20b (). It has a concentration of N-type dopants (called the second concentration), higher than the first concentration. Typically, the second concentration is between 1.10 20 / cm 3 and 5.10 20 / cm 3. To simplify the supply of the support substrate 20, the second dopant concentration can preferably be limited to the range 1.10 20 / cm 3 and 2.10 20 / cm 3 . The N-type dopant, in the donor substrate 1 or in the support substrate 20, may in particular be nitrogen. Usually, in the field of microelectronics, the support substrate 20 and the donor substrate 1 are in the form of circular wafers with a diameter typically between 100 mm and 200 mm (or more). The front (circular) faces 1a, 20a and rear 1b, 20b of the substrates 1, 20 extend parallel to the main plane (x, y). The thicknesses of the donor substrates 1 and support 20 are typically between 200 μm and 800 μm. Advantageously, the support substrate 20 has the particularity of comprising, at least on the side of its face to be assembled 20a, grains of silicon carbide 21i, 21j having a size less than or equal to 2 μm, less than or equal to 1 μm, or even less than or equal to 500 nm, said grains covering in the main plane (x, y), a surface greater than or equal to 20% of the face to be assembled 20a. The size of a grain 21i, 21j, delimited by the grain boundaries 22, corresponds to the largest dimension of said grain 21i, 21j, in the main plane (x, y). To measure the dimensions of the grains 21i, 21j or distances between grain boundaries 22, it is possible to rely on images obtained by conventional scanning electron microscopy (SEM) or involving electron diffraction (EBSD for "Electron Back Scattered Diffraction"). It is also possible to use X-ray crystallography. Advantageously, the grains 21i, 21j of p-SiC having a size less than or equal to 2 μm, less than or equal to 1 μm, or even less than or equal to 500 nm, cover, in the main plane (x,y), a surface greater than or equal to 30%, 40%, 50%, 60%, 70%, or even greater than or equal to 80% of the front face to be assembled 20a of the support substrate 20. It is understood that the rest of the surface is occupied by grains of larger size. The manufacturing method then comprises a step c) of depositing a 3' silicon film (amorphous or polycrystalline) on the face to be assembled 1a of the donor substrate 1, on the face to be assembled 20a of the support substrate 20 or on both faces 1a, 20a (). Advantageously, this deposition is carried out by sputtering at low temperature, and involves a high vacuum, typically between 3x10 -6 Pa and 8x10 -6 Pa, before the injection of the spray gas (Ar), then an atmosphere around 10 -1 Pa and 10 -2Pa during deposition. The thickness of the deposited 3' film can vary between a few nm and a few tens of nm. The following step d) corresponds to the assembly, by direct bonding, of the donor substrate 1 and the support substrate 20 at their faces to be assembled 1a, 20a, so as to obtain a bonded assembly 50 comprising an interlayer 3 made of silicon, resulting from the film(s) 3' deposited in step c), between the donor substrate 1 and the support substrate 20 (). This interlayer 3 advantageously has a thickness of between 8 nm and 30 nm. The assembly is carried out by molecular adhesion bonding, consisting of bringing the faces to be assembled 1a, 20a into contact, without adding adhesive material. This may involve direct bonding between the useful layer 10 and the silicon film 3', when the latter has been deposited solely on the support substrate 20, or direct bonding between the support substrate 20 and the film 3', when the latter has been deposited solely on the donor substrate 1, or even direct bonding between two silicon films 3', when they have been deposited on the donor substrate 1 and on the support substrate 20. The direct assembly is preferably carried out under a high vacuum, of the order of 10 -6 Not or less. Advantageously, the deposition of step c) and the direct assembly of step d) are chained together without breaking the vacuum, in situ or in multi-chamber equipment. Mention may be made, for example, of the BV7000 atomic diffusion bonding equipment from Canon, in which it is possible to successively carry out the deposition of the 3' film and the direct bonding, while maintaining a controlled atmosphere. The method then comprises a step e) involving a separation along the buried fragile plane 11 to form a transferred structure 100' comprising the useful layer 10, transferred from the donor substrate 1, arranged on the interlayer 3, itself arranged on the support substrate 20. Such a separation can be carried out during a heat treatment, typically around 900°C-1000°C, capable of growing cavities and microcracks, induced by the implanted species, in the buried fragile plane 11. The separation can also be carried out by applying a mechanical stress, or by the combination of thermal and mechanical stresses, as is well known with reference to the Smart Cut process. Sequences of cleaning, smoothing, polishing or etching of the separated face 10a of the useful layer 10 and / or of the separated face 1'a of the remainder 1' of the donor substrate may be carried out so as to restore good surface quality, in particular in terms of roughness, defects and other contamination. Finally, the method comprises a step f) corresponding to the application of a heat treatment to the transferred structure 100', at a holding temperature between 1500°C and 2000°C, advantageously between 1750°C and 2000°C, to form the composite structure 100 (, left). This temperature range induces the melting of the silicon of the interlayer 3. The interlayer 3 will segment in the form of drops to optimize its surface energy and the liquid silicon 3'' of at least some of these drops will diffuse into the grain boundaries 22. It will then be able to interact with the SiC grains 21i, 21j of the support substrate and dissolve a certain quantity thereof. The dissolved carbon, whose diffusion coefficient in the liquid silicon is 1.7.10 -8 m 2 / s, is then transported from the dissolution zone to the interface between the liquid silicon and the useful monocrystalline layer 10. By recombining with the silicon, the carbon will allow a growth of monocrystalline SiC from the useful layer 10, the latter acting as a seed (, right). This is similar to growth in a TSSG process ("top seeded solution growth"), which consists of placing liquid silicon in a graphite crucible and then positioning a high-quality monocrystalline SiC seed on the surface of the liquid silicon; the carbon present on the walls of the crucible will dissolve in contact with the liquid silicon and diffuse into the latter by temperature gradient to the SiC seed. This results in a growth of monocrystalline SiC from this seed.In the present invention, this growth process takes place on a very local scale and at the interface zone between the useful layer 10 and the support substrate 20: the graphite crucible is replaced by the grains 21i, 21j of the p-SiC support substrate, the seed by the useful layer 10 in c-SiC and the source of silicon comes from the interlayer 3. The composite structure 100 which results from this step f) therefore has an intermediate region 30, which extends along the main plane (x,y) and which comprises inclusions 31 made of monocrystalline silicon carbide in direct contact with the useful layer 10 and extending, in a direction (z) normal to the main plane (x,y), between grains 21i, 21j of the support substrate 20. The growth mode of these inclusions 31 promotes the diffusion of the dopants, present in the grains 21i, 21j, into said inclusions 31. Thus, the latter have a third concentration of N-type dopants between the first and second concentrations, which is very favorable to the establishment of good electrical conduction between the useful layer 10 and the support substrate 20. For example, for a first dopant concentration of 5.10 18 / cm 3 and a second dopant concentration of 5.10 20 / cm 3, the third dopant concentration is expected to be between 1.10 20 and 1.5.10 20 / cm 3 . The inclusions 31 typically have a thickness, along an axis (z) normal to the main plane (x,y), of between 20 nm and 70 nm, and a width, in the main plane (x,y), of between 100 nm and 700 nm. The size of the inclusions 31 is in particular a function of the thickness of the interlayer 3 made of silicon. The intermediate region 30 has a density of inclusions 31, in the main plane (x,y), typically between 0.1 / μm 2 and 1 / μm 2This density is all the greater as grains 21i, 21j of small size (less than or equal to 2 μm, less than or equal to 1 μm, or even less than or equal to 500 nm) cover a large surface area of ​​the assembled face 20a of the support substrate 20. According to the invention, these small grains cover more than 20% of the surface area of ​​said assembled face 20a, more than 30%, more than 40%, more than 50%, more than 60%, more than 70%, or even 80% and more. The crystallographic orientations of the grains 21i, 21j of the assembled face 20a of the support substrate 20 can also influence the formation, size and density of the inclusions 31. The density of inclusions 31 is also linked to the thickness of the interlayer 3 of silicon, a greater quantity of silicon being capable of increasing said density in the intermediate region 30. It should be noted that a significant disorientation of the grains between them or with respect to the

[0111] orientation is favorable to the interaction between the silicon drops and the grain boundaries 22, because this disorientation is accompanied by grain boundaries 22 that are not very stable and it is then energetically favorable for the silicon to dissolve carbon atoms at the level of said grain boundaries 22. A disorientation greater than 10°, 20°, or even 30° appears particularly advantageous. The orientation of the grains can be observed by transmission electron microscopy (TEM) or by electron diffraction (EBSD for “Electron Back Scattered Diffraction”). Such disorientation is expected between the grains 21i, 21j covering at least 20%, 30%, 40% or even more than 50% of the surface of the assembled face 20a of the support substrate 20; this objective may be evaluated by statistical measurements on said face 20a. The inclusions 31 in c-SiC between the p-SiC grains (3C) have a particular geometry. For example, in the case of a useful layer 10 in c-SiC of polytype 4H, it has been observed that the base of inclusions 31 is perpendicular to the direction

[0004] of the c-SiC of the useful layer 10, parallel to the {0004} planes. Thus, the base of at least some inclusions 31 is not parallel to the main plane (x,y), but disoriented by 4° relative to it. This is due to the fact that the {0004} planes of the SiC-4H monocrystal are disoriented by 4° relative to the assembled face 1a of the useful layer 10 (and therefore relative to the main plane) before step e). Since the growth of the SiC in the form of inclusions 31 is carried out by epitaxy from the useful layer 10, the crystal planes keep the same orientation. The growth therefore ends on {0004} planes, at 4° from the main plane (x,y). The orientation of the flanks of the inclusion 31 c-SiC seems to depend on the stability of the facets between the hexagonal single crystal and the cubic polycrystal. The inclination of the flanks also depends on the respective orientations of the grains 21i, 21j in contact with said inclusion 31. In general, the flanks of the inclusions 31 have an inclination between 110° and 135° relative to the base. The heat treatment of step f) plays an important role in the formation of these inclusions 31 in c-SiC. Thus, beyond a plateau temperature between 1500°C and 2000°C, the duration of this plateau and the temperature rise and fall ramps can also play a role. The holding temperature is preferably 1600°C, 1700°C, 1750°C, 1800°C, 1850°C, 1900°C, or 1950°C. The duration of the holding temperature is defined so as to allow the diffusion of carbon in the liquid silicon of the drops at the grain boundaries 22, and the growth of the inclusions 31 from the useful layer 10. It is these inclusions 31 which promote vertical electrical conduction; thus their size and density are to be developed as best as possible. Preferably, the holding temperature is maintained for a duration of 30 min, 1 h, 2 h, 3 h, 4 h or even longer. The temperature rise (up to the plateau) is advantageously chosen to be less than or equal to 20°C / min, less than or equal to 10°C / min, or even less than or equal to 5°C / min, because this promotes the creep of the silicon from the interlayer 3, towards the grain boundaries 22. The same ramp ranges can be applied to the temperature drop. In addition, these ramp ranges reduce the risk of formation of “bubble” type defects at the assembly interface. The intermediate region 30 may also comprise silicon nodules 33, 32, arranged in the interface zone, or under the inclusions 31 between grains 21i, 21j of the support substrate 20 (). Indeed, certain silicon drops, resulting from the segmentation of the interlayer 3 during the heat treatment of step f), may be located opposite a grain 21i, 21j and not a grain boundary 22: in this case, they form, following their solidification during the temperature reduction ramp of the heat treatment, silicon nodules 33 arranged between the useful layer 10 and the support substrate 20, in the composite structure 100. Such a nodule 33 may have a thickness, along an axis (z) normal to the main plane (x,y), of between 50 nm and 200 nm, and a width, in the main plane (x,y), of between 200 nm and 400 nm. It is also observed that the liquid silicon 3'' having diffused between grains 21i, 21j, along grain boundaries 22, can form a nodule 32 if it is not completely consumed in the growth of an inclusion 31: in this case, we find a silicon nodule 32 arranged under an inclusion 31 and between grains 21i, 21j of the support substrate 20, in the composite structure 100 at the end of step e). This type of nodule 32 can have dimensions, in the main plane (x,y), of the order of 200 nm to 400 nm, and a typical thickness, along the axis (z), of a few nm to a few μm. The interface between the c-SiC inclusion 31 and the underlying silicon nodule 32 is disoriented by 4° relative to the main plane (x,y). The intermediate region 30 further includes an interface zone between an assembled face 1a of the useful layer 10 and an assembled face 20a of the support substrate 20, said interface zone being due to the segmentation of the interlayer 3 which causes the appearance of direct contact zones between the useful layer 10 in c-SiC and the support substrate 20 in p-SiC. These direct contact zones can form steps 34 as can be seen in the. It is assumed that the solubilization of the p-SiC (3C) in the liquid silicon 3'' during step f) makes it possible to reconstruct the c-SiC (4H) of the useful layer 10 in the form of steps 34 ("step bunching") in at least part of the direct contact zones between the thin layer 10 and the support substrate 20.It appears that the steps 34 are located at the level of direct contact zones where the grains 21i, 21j of the support substrate 20 have low disorientations between them, whereas the inclusions 31 in c-SiC described previously are located at the level of grains 21i, 21j which are very disoriented and / or thermodynamically unstable. The total thickness of the intermediate region 30, along an axis (z) normal to the main plane (x,y), is greater than or equal to the maximum thickness of the inclusions 31. The presence of silicon nodules 32 under all or part of the inclusions 31 can of course increase this total thickness. To achieve good vertical electrical conduction, in other words, a resistivity of the intermediate region 30 as low as possible, preferably less than 1 mohm.cm 2 , or even less than 0.1 mohm.cm 2, certain preferred embodiments are proposed, depending on the doping levels of the useful layer 10 and the support substrate 20. According to a first embodiment, the first concentration (useful layer 10) of N-type dopants (nitrogen) is of the order of 5.10 18 / cm 3 , and the second concentration (support substrate 20) in N-type dopants (nitrogen) is of the order of 1.10 20 / cm 3. To obtain the desired low resistivity, the support substrate 20 is chosen such that the silicon carbide grains 21i, 21j having a size less than or equal to 500 nm cover, in the main plane (x, y), a surface greater than or equal to 20% of its face to be assembled. This contributes to obtaining a density of inclusions 31 in c-SiC, favorable to vertical electrical conduction, compatible with the desired objective. The third concentration of dopants (inclusions) is between the first and second concentrations. According to a second embodiment, the first concentration of N-type dopants (useful layer) is of the order of 3.10 18 / cm 3 , and the second concentration of N-type dopants (support substrate) is of the order of 1.10 20 / cm 3. To obtain the desired low resistivity, the support substrate is chosen such that the silicon carbide grains having a size less than or equal to 500 nm cover, in the main plane, a surface greater than or equal to 50%, or even 60% of its face to be assembled. This contributes to obtaining a density of inclusions 31 in c-SiC with a third concentration of dopants between the first and the second concentration, inclusions favorable to vertical electrical conduction, compatible with the desired objective. According to a third embodiment, the first concentration of N-type dopants (useful layer) is of the order of 1.10 18 / cm 3 , and the second concentration of N-type dopants (support substrate) is of the order of 1.10 20 / cm 3. To obtain the desired low resistivity, the support substrate is chosen such that the silicon carbide grains having a size less than or equal to 500 nm cover, in the main plane, a surface greater than or equal to 80% of its face to be assembled. Here again, this contributes to obtaining a density of inclusions 31 in c-SiC with a third concentration of dopants between the first and the second concentration, inclusions favorable to vertical electrical conduction, compatible with the desired objective. The present theoretical curves obtained by numerical simulations of the electronic transport vertically in the composite structure 100. They illustrate the evolution of the resistivity of the intermediate region 30 as a function of the surface percentage of the support substrate 20 (on the side of its assembled face 20a) in which the grains 21i, 21j have a size less than or equal to 500 nm. When the second dopant concentration of the support substrate 20 (denoted N polySiC on the) is very high (typically 5.10 20 / cm 3 ), it is not critical that the support substrate 20 has a certain percentage of small grains (< 500 nm), a resistivity of less than 10 -5 ohm.cm 2 can be achieved regardless of the doping level of the useful layer 10 (between 1.10 18 / cm 3 and 5.10 18 / cm 3 ). On the other hand, when the second dopant concentration is of the order of 1.10 20 / cm 3, obtaining a resistivity lower than 10 -5 ohm.cm 2 will be conditioned by the presence of small-sized grains favorable to the formation of inclusions 31 in c-SiC, which benefit from a third concentration of dopants higher than the first concentration of the useful layer 10. The lower the first concentration, the higher the percentage of surface area of ​​the assembled face 20a of the support substrate 20 comprising grains 21i,21j of a size less than or equal to 500 nm must be. The invention also relates to a composite structure 100 having a front face 100a and a rear face 100b extending parallel to the main plane (x,y) (). It comprises a useful layer 10 having a first concentration of N-type dopants. Typically, the first concentration is between 1.10 18 / cm 3 and 5.10 18 / cm 3 . It also comprises a support substrate 20 having a second concentration of N-type dopants, higher than the first concentration. Typically, the second concentration is between 1.10 20 / cm 3 and 5.10 20 / cm 3 , or even preferentially limited to the 1.10 range 20 / cm 3 and 2.10 20 / cm 3 . Advantageously, the support substrate 20, at least on the side of its assembled face 20a, comprises silicon carbide grains 21i, 21j having a size less than or equal to 2 μm, less than or equal to 1 μm, or even less than or equal to 500 nm, said grains covering in the main plane (x,y), a surface greater than or equal to 20% of the assembled face 20a. In particular embodiments, the silicon carbide grains 21i, 21j having a size less than or equal to 2 μm, less than or equal to 1 μm, or even less than 500 nm cover, in the main plane (x,y), a surface greater than or equal to 30%, 40%, 50%, 60%, 70%, or even 80% of the assembled face 20a of the support substrate 20. Advantageously, the grains of the support substrate 20, on the side of its assembled face 20a, are significantly disoriented between them (i.e. between two adjacent grains) or with respect to the

[0111] orientation: as mentioned previously in the method according to the invention, this is favorable to the interaction between the silicon drops and the grain boundaries 22 (in step f), because the low stability of the grain boundaries 22, linked to this disorientation, makes the dissolution of the carbon atoms by the silicon at said grain boundaries 22 energetically favorable. A disorientation greater than 10°, 20°, or even 30° appears particularly favorable. Such a disorientation is expected between the grains 21i, 21j covering at least 20%, 30%, 40% or even more than 50% of the surface of the assembled face 20a of the support substrate 20. The composite structure 100 also comprises an intermediate region 30 extending along the main plane (x,y) and including an interface zone between an assembled face 1a of the useful layer 10 and an assembled face 20a of the support substrate 20. The intermediate region 30 comprises inclusions 31 made of monocrystalline silicon carbide in direct contact with the useful layer 10 and extending, in a direction (z) normal to the main plane (x,y), between grains 21i, 21j of the support substrate 20. Said inclusions 31 have a third concentration of N-type dopants between the first and the second concentration. The presence of these inclusions 31 provides a significant advantage in terms of electrical conduction of the intermediate region 30 of the composite structure 100. The intermediate region 30 may have a density of inclusions 31, in the main plane (x,y), of between 0.1 / μm 2 and 1 / μm 2 . Typically, the inclusions 31 have a thickness, along an axis (z) normal to the main plane (x,y), of between 20 nm and 70 nm, and a width, in the main plane (x,y), of between 100 nm and 700 nm. As mentioned previously in the description of the method for manufacturing the composite structure 100, the intermediate region 30 may also comprise silicon nodules 32, 33 and steps 34 at the interface zone between the assembled face of the useful layer 10 and that of the support substrate 20. Of course, the invention is not limited to the embodiments described and variant embodiments can be made without departing from the scope of the invention as defined by the claims.

Claims

Composite structure (100) having a front face (100a) and a rear face (100b) extending parallel to a main plane (x,y), and comprising:- a useful layer (10) of monocrystalline silicon carbide, a free face of which constitutes the front face (100a), and having a first concentration of N-type dopants,- a support substrate (20) of polycrystalline silicon carbide, a free face of which constitutes the rear face (100b), and having a second concentration of N-type dopants, the second concentration being greater than the first concentration,- an intermediate region (30) extending along the main plane (x,y) and including an interface zone between an assembled face (1a) of the useful layer (10) and an assembled face (20a) of the support substrate (20);the composite structure (100) being characterized in that the intermediate region (30) comprises inclusions (31) of monocrystalline silicon carbide in direct contact with the useful layer (10) and extending, in a direction (z) normal to the main plane (x,y), between grains (21i,21j) of the support substrate (20), said inclusions (31) having a third concentration of N-type dopants between the first and the second concentration.; Composite structure (100) according to claim 1, in which the support substrate (20), at least on the side of its assembled face (20a), comprises grains (21i, 21j) of silicon carbide having a size less than or equal to 500 nm, said grains (21i, 21j) covering in the main plane (x, y), a surface greater than or equal to 20% of the assembled face (20a). Composite structure (100) according to one of the preceding claims, in which the grains (21i, 21j) of silicon carbide having a size less than 500 nm cover, in the main plane (x, y), a surface greater than or equal to 50%, or even 80% of the assembled face (20a) of the support substrate (20). Composite structure (100) according to one of the preceding claims, in which at least 20% of the grains (21i, 21j) of silicon carbide, on the side of the assembled face (20a) of the support substrate (20), have a disorientation, between them or with the orientation [111], greater than or equal to 30°. Composite structure (100) according to one of the preceding claims, in which the intermediate region (30) comprises silicon nodules (32, 33), arranged in the interface zone or under the inclusions (31) between grains (21i, 21j) of the support substrate (20). Composite structure (100) according to one of the preceding claims, in which the first concentration is between 1.10 18 / cm 3 and 5.10 18 / cm 3 . Composite structure (100) according to one of the preceding claims, in which the second concentration is between 1.10 20 / cm 3 and 5.10 20 / cm 3 , preferably between 1.10 20 / cm 3 and 2.10 20 / cm 3 . Composite structure (100) according to one of the preceding claims, in which the inclusions (31) have a thickness, along an axis (z) normal to the main plane (x,y), of between 20 nm and 70 nm, and a width, in the main plane (x,y), of between 100 nm and 700 nm. Composite structure (100) according to one of the preceding claims, in which the intermediate region (30) has a density of inclusions (31), in the main plane (x,y), of between 0.1 / μm 2 and 1 / μm 2 . A method of manufacturing a composite structure (100) according to one of claims 1 to 9, comprising the following steps: a) providing a donor substrate (1) made of monocrystalline silicon carbide having a first concentration of N-type dopants, and comprising a buried fragile plane (11) induced by ion implantation of light species; b) providing the polycrystalline support substrate (20); c) depositing a silicon film (3') on a face to be assembled (1a) of the donor substrate (1) and / or on the face to be assembled (20a) of the support substrate (20); d) assembling the donor substrate (1) and the support substrate (20) at their faces to be assembled (1a, 20a), so as to obtain a bonded assembly (50) comprising an interlayer (3) made of silicon, derived from the film(s) (3') deposited in step c), between the donor substrate (1) and the support substrate (20), said interlayer (3) having a thickness of between 8 nm and 30 nm;e) separating along the buried fragile plane (11) to form a transferred structure (100') comprising the useful layer (10), transferred from the donor substrate (1), arranged on the interlayer (3), itself arranged on the support substrate (20); f) applying a heat treatment to the transferred structure (100'), at a holding temperature between 1500°C and 2000°C, to form the composite structure (100).; Method for manufacturing a composite structure (100) according to claim 10, in which the support substrate (20) comprises, at least on the side of a face to be assembled (20a), grains (21i, 21j) of silicon carbide having a size less than or equal to 500 nm, said grains (21i, 21j) covering, in the main plane (x, y), a surface greater than or equal to 20% of the face to be assembled (20a). Method for manufacturing a composite structure (100) according to one of claims 10 and 11, in which at least 20% of the grains (21i, 21j) of silicon carbide, on the side of the assembled face (20a) of the support substrate (20), have a disorientation, between them or with the orientation [111], greater than or equal to 30°. Method for manufacturing a composite structure (100) according to one of claims 10 to 12, in which the bearing temperature is maintained for a period of between 30 min and 4 h. Method for manufacturing a composite structure (100) according to one of claims 10 and 13, in which the treatment of step f) has a temperature rise ramp less than or equal to 20°C / min, or even less than or equal to 10°C / min. Method of manufacturing a composite structure (100) according to one of claims 10 to 14, in which:- the second concentration is of the order of 1.10 20 / cm 3,- the first concentration is of the order of 3.10 18 / cm 3 , and- the grains (21i, 21j) of silicon carbide having a size less than or equal to 500 nm cover, in the main plane (x, y), a surface greater than or equal to 50%, or even 60% of the face to be assembled (20a) of the support substrate (20). Method of manufacturing a composite structure (100) according to one of claims 10 to 14, in which:- the second concentration is of the order of 1.10 20 / cm 3 ,- the first concentration is of the order of 1.10 18 / cm 3 , and- the grains (21i, 21j) of silicon carbide having a size less than or equal to 500 nm cover, in the main plane (x, y), a surface greater than or equal to 80% of the face to be assembled (20a) of the support substrate (20).

Images