Reduced truncation angle variation element 13 nitride plate

Abstract

The invention relates to a two-dimensional element 13 or III nitride crystal platelet, delimited by an N-oriented face and an opposite E-oriented face with respect to element 13 or III, E being preferably chosen from Ga, In, Al, or a combination thereof, characterized in that the variation of the crystal truncation angle along the largest dimension of said platelet is less than 5 × 10⁻³° / mm, and its curvature or geometric deformation of its faces has an absolute value less than 10⁻³ mm / mm of the largest dimension of said platelet. Figure for the abstract: Fig. 3

Description

Title of the invention: Reduced truncation angle variation element 13 nitride plate Technical field

[0001] The invention relates to the general technical field of semiconductor materials based on nitride of elements from columns 13 or III of the periodic table - in particular such as gallium nitride GaN.

[0002] In particular, the present invention relates to element 13 nitride wafers (or "wafer" according to Anglo-Saxon terminology), their manufacturing process and their use.

[0003] Such a nitride wafer of element 13 can be intended for the realization of semiconductor structures such as light-emitting diodes (LEDs) or laser diodes (LDs) or other electronic components. PRESENTATION OF PRIOR ART

[0004] Current manufacturing processes for element 13 nitride-based semiconductor materials rely on the so-called heteroepitaxy technique, which consists of growing a crystal (such as a gallium nitride crystal GaN) on a starting substrate of a different nature (such as a sapphire substrate).

[0005] However, the hetero-epitaxy technique can induce many crystal defects in the element 13 nitride structure, such as dislocations, limiting the performance and lifespan of element 13 nitride-based components, as well as geometric deformation of the crystal and a high variation in the truncation angle.

[0006] These defects are due to differences in material between the starting substrate and the element 13 nitride layer, and in particular to the mismatch:

[0007] - of crystal lattice parameters (i.e., lattice parameter) between the starting substrate and the element 13 nitride layer, and - coefficient of thermal expansion between the starting substrate, usually sapphire, and the element 13 nitride layer.

[0008] Different ways of obtaining by heteroepitaxy a crystal exhibiting crystalline and geometric deformations are known.

[0009] One solution consists of allowing the crystalline layer to grow freely on the substrate so as to form a sphere from which platelets of element 13 nitride material can be cut. This method leads to a crystal with little gradient in crystalline quality, but the yield of this process, calculated on the basis of the number of spheres of good crystalline quality divided by the total number of spheres formed, is low. By elsewhere, the dimensional quality and the variation of the truncation angle still remain very improvable.

[0010] Crystalline quality can be measured by the full width at half maximum of the X-ray diffraction (XRD) peak of the line (002) around the angle co under symmetric conditions of the GaN (0001) planes.

[0011] Another approach consists of first depositing and forming a constrained layer bonded to the flat substrate, followed by separation of said substrate to produce a self-supporting crystalline layer. Unlike the previous process, this technique leads to a crystal exhibiting a crystalline quality gradient typically of at least 5% between the two opposite faces of the self-supporting layer. This latter self-supporting crystal growth process can lead to very good process yield, but the dimensional quality and the variation in truncation angle remain improvable.

[0012] Document JP 56059699 describes the possibility of fabricating a nitride crystal of element 13 with a truncation angle between 0.5 and 4 degrees, said crystal being obtained by epitaxy on a sapphire substrate having a truncation angle between 0.5 and 4 degrees. However, the crystalline quality of such a crystal is no longer sufficient to allow the fabrication of optoelectronic and electronic components, particularly due to the presence of stacking faults and dislocations.

[0013] The article "Effect of Slight Misorientation of Sapphire Substrate on Metalorganic Chemical Vapor Deposition Growth of GaN" by Takayuki Yuasa et al., published in the Journal of Applied Physics, vol. 38 (1999), pages L703-705, Part 2, no. 7A, 1 July 1999, demonstrates the advantage of using a sapphire substrate with a truncation angle of approximately 0.1 to 0.2 degrees to produce a GaN crystal with a uniform morphology. Using a starting substrate with too high a truncation angle results in a GaN crystal with a scale-shaped growth face (or "scale-shaped" according to Anglo-Saxon terminology). This approach requires very strict substrate selection and is not necessarily sufficient because, as explained previously, the substrate truncation angle is not the only determining parameter.

[0014] US patent 2012 / 0184090 proposes a method in which element-13 nitride wafers having a non-zero truncation angle are obtained by cutting a thick element-13 nitride ingot with a zero truncation angle. More specifically, the method described in US 2012 / 0184090 consists of:

[0015] - to grow a nitride ingot of element 13 with a zero truncation angle, and to - to cut the ingot along cutting planes inclined parallel to the crystallization plane at the periphery of said ingot.

[0016] This makes it possible to obtain element 13 nitride platelets having an angle of Non-zero truncation occurs in the central region of the wafer, with the truncation angle decreasing in a peripheral region of the wafer compared to the diametrically opposite region. However, cutting a thick ingot along an inclined cutting plane results in significant material losses.

[0017] Document EP1528591A2 proposes a method to reduce the geometric deformation of a self-supporting semiconductor crystal with a deflection of + / - 100 pm at the end of the growth stage to achieve a range of at most +30 pm / -20 pm after the finishing stage. The method consists of mechanically grinding the concave face to obtain a damaged layer and flatten the crystal. This damaged layer can then be removed by chemical etching or frosting to achieve the aforementioned geometric specification. The damaged layer is very thin (a few micrometers). However, this document is silent regarding the variation in the truncation angle. For the final application, this layer must be removed. However, during the removal of the damaged layer, particularly, for example, by chemical etching, the crystal deforms again, and the variation in the truncation angle remains high.

[0018] There is therefore an unmet need for a nitride crystal wafer of element 13 exhibiting high crystalline quality, and in particular in the case of self-supporting crystals said wafer exhibiting a crystalline quality gradient greater than 5%, whose truncation angle variation and geometric deformation are very low permanently. Summary of the invention

[0019] An object of the present invention is to provide a crystal wafer showing these characteristics obtained through a new manufacturing process that can be implemented with any type of starting substrate, allowing the production of a final crystal exhibiting low geometric deformation and truncation angle variation despite the multiple deformations generated by the different finishing stages between obtaining the raw crystal and the final wafer ready for the production of an optoelectronic or electronic component.

[0020] The inventors have discovered that a particular process makes it possible to obtain a two-dimensional structure, preferably self-supporting, nitride crystal of element 13 or III, in particular a plate delimited by a face with N orientation, an opposite face with E orientation according to element 13 or III, E being preferably chosen from Ga, In, Al or a combination of these elements:

[0021] - of high crystalline quality, in particular and preferably, such as - the full width at half maximum (FWHM) of the X-ray diffraction peak of the (002) line around the angle co measured on face E (0001), where FWHME, is less than 130 Arcsec - and the quality gradient defined by the ratio (FWHM'N-FWHM'E) / thickness of said wafer, where FWHM'N designates the full width at half maximum of the X-ray diffraction peak of the (201) line around the angle co measured on the N (Nitrogen) face and FWHM'E is that measured on the face of element 13, is in absolute value between 0.005 and 5 arcsec / micrometer. - whose variation in the crystalline truncation angle along the largest dimension of said platelet is less than 5x10 3° / mm, i.e. less than 0.005° per mm, preferably less than 4x10 3° / mm, preferably less than 2x10 3° / mm, or even less than 1x10 3° / mm, and - whose curvature or geometric deformation of its faces has a deflection less than 103 mm / mm in absolute value of the largest dimension of said plate, preferably less than 6x10 4 mm / mm, - a wafer surface having a crystalline quality and roughness allowing the growth of a subsequent layer in a single direction.

[0022] According to one possible mode, the face E is convex, as shown in figure 1a, and the geometric deflection of said face E is negative and greater than -10-3 mm / mm of the largest dimension of said wafer, preferably greater than -0.7x10-3 mm / mm of the largest dimension of said wafer.

[0023] The method according to the invention comprises in particular the following steps:

[0024] a: obtaining a platelet of rough crystal, preferably self-supporting, of element 13 nitride by epitaxial growth,

[0025] b: determination of the variation of the truncation angle of said rough crystal at the end of growth, optionally after cutting said crystal, preferably by grinding or by cutting with a wire saw, to a relative thickness greater than 2x103, preferably greater than 5x103 and less than 102, said relative thickness corresponding to the ratio of the thickness of the wafer to its largest dimension, in particular to the largest width of its faces,

[0026] c: work hardening of a face in order to reduce the variation of the truncation angle, in particular by sandblasting, exposure to laser radiation, shot peening, ion bombardment or other technique allowing the creation of a surface mechanical stress on said crystal, so as to obtain a non-zero geometric deflection, in particular a geometric deflection with an absolute value between 0.001 and l, preferably between 0.01 and 0.1 mm / mm of the largest dimension of said wafer

[0027] d: after work hardening, grinding of face N by machining,

[0028] e: after work hardening, grinding of face E by machining,

[0029] f: polishing of face E,

[0030] g: chemical roughening of face N after machining of face N or after polishing of face E,

[0031] h: cleaning to decontaminate the surfaces of the wafer

[0032] i: preparation of face E, in particular by plasma etching, for subsequent deposition of at least one layer necessary for the realization of an optoelectronic or electronic component.

[0033] In the sequence of steps of the process according to the invention, the inventors found that it is possible to obtain a very small variation in truncation angle by combining work hardening, not for the purpose of geometrically straightening the crystal, but in order, later in the process, to maintain equality of the geometric and crystalline deflections after rectification and polishing of the N face, with the aim of reducing the variation in the truncation angle of the crystal after the finishing steps. Here, "equality of deflections" means that the difference between the value of the geometric deflection and that of the crystalline deflection divided by the geometric deflection is less than 5%, preferably less than 2% in algebraic value, that is to say, in absolute value and sign.The geometric deformation of the crystal due to this work hardening step c is then "corrected" by subsequent grinding steps which will allow a reduction of the geometric deformation of the crystal.

[0034] In particular, the final crystal wafer obtained after finishing steps f), g), h), and i) exhibits a crystal deflection or curvature, corresponding to the deformation of its crystal planes, less in absolute value than 1 x 10³ mm / mm of the largest dimension of said wafer, preferably less in absolute value than 0.5 x 10³ mm / mm of the largest dimension of said wafer. According to a preferred method, this crystal deflection remains less in absolute value than 1 x 10³ mm / mm at all stages following the work-hardening step c.

[0035] According to a particular method, the work hardening is carried out until the variation in truncation angle is less than 5x10 3° / mm, preferably less than 4x103 ° / mm or even less than 3x10 3° / mm of the largest dimension of the insert.

[0036] Preferably, work hardening is carried out on the face of the plate opposite which the crystal deformation is concave or for which the curvature of the crystal lattice is concave. Unlike the teaching of EP1528591A2, work hardening is not carried out until a geometric deflection equal to 0 is obtained, but until the geometric deflection is aligned with or equal to the crystal deflection.

[0037] As shown for example in example 1 below, the preliminary step c) of work hardening thus makes it possible to anticipate the deformations generated by the following finishing operations, in particular the grinding operation and subsequent physicochemical treatments of the insert.

[0038] In the case where the N face of the plate is concave, the work hardening step c is carried out until a crystal deflection of absolute value between 0.2 and 2, preferably 0.3 and 1.5, or even between 0.4 and 1.0 µm / mm of the largest dimension of the platelet,

[0039] In the case where the face N of the plate is convex, the work hardening step c is carried out until a crystal deflection is obtained in absolute value between 0.1 and 1.5, preferably 0.2 and 1.2, or even between 0.2 and 1.0 pm / mm of the largest dimension of the plate.

[0040] According to one possible mode, the thinning by machining in step b) is predetermined based on the initial geometric curvature of the raw crystal. This advantageously makes it possible to obtain a curvature corresponding to a deflection of less than 6 x 10⁴ mm / mm of the largest dimension of said crystal.

[0041] In particular, the inventors have determined that it is possible by prior iteration on a series of wafers, for example in the case of a Gallium nitride wafer, to define the optimal thickness of the wafer as a function of the geometric deflection of the rough crystal before the work hardening step c.

[0042] According to one possible mode, the optimal thickness in pm of the crystal before work hardening follows the following relationship:

[0043] ef = ah + be;

[0044] where f in pm is the initial deflection of the crystal in absolute value before rectification;

[0045] ei is the initial thickness in pm of the crystal before cutting;

[0046] where a is between -3.5 and -2 depending on the chemical composition of the crystal, and where b is between 0.1 and 1.

[0047] The invention also relates to the use of an element 13 nitride wafer as described above for the manufacture of optoelectronic or electronic components, such as light-emitting diodes, laser diodes, vertical or horizontal geometry transistors, power electronics or electronic components for radio frequency transmission, reception or amplification, current rectifier diodes or sensors. Definitions:

[0048] With reference to Figures IA and IB, the geometric curvature of a rough crystal 5, in particular the convexity or concavity, of face N is defined with respect to face N. The curvature of face N is determined by measuring the geometric sagitta (distance between face N and a plane tangent to that face). Symmetrically, the geometric curvature of face E, in particular its convexity or concavity, can be determined by measuring the geometric sagitta of the opposite face E. The sagittas given in the examples were determined according to this convention. The sagitta is negative with respect to a convex face and positive with respect to a concave face.

[0049] With reference to Figures 2A, 2B, 2C and 2D, it is illustrated for a rough crystal 5, a crystalline curvature of a crystal plane 11 and a geometric curvature. Thus, the [Fig. 2A] represents an ideal rough crystal 5, geometrically flat and without crystal curvature. In this crystal, the crystal deflection, representing the curvature or geometric deformation of a reference crystal plane, is zero. The crystal deflection is negative with respect to a convex crystal curvature, and positive with respect to a concave crystal curvature.

[0050] Fig. 2B illustrates a rough crystal 5, geometrically flat and with a crystal curvature. In this crystal, the crystal arrow, representing the curvature or geometric deformation of a crystal plane, is non-zero.

[0051] Fig. 2C illustrates a rough crystal 5 which has a concave geometric curvature and a concave crystalline curvature (positive crystalline arrow).

[0052] Fig. 2D also illustrates a rough crystal 5 which exhibits a convex geometric curvature and a concave crystalline curvature (positive crystalline arrow).

[0053] Thus, as illustrated in these figures, the crystalline curvature may be different from the geometric curvature of the crystal.

[0054] The geometric deflection or curvature of the plate can be measured using a micrometric comparator of the profilometer type as described by EP1528591A2

[0089] ,

[0055] The crystal bow or curvature of the wafer is measured by X-ray diffraction as described in the publication "Curvature and bow of bulk GaN substrates" by Humberto M. Foronda et al., published in the Journal of Applied Physics 120, 035104 (2016). The crystal curvature radius Rc is then defined by the following relation: Rc = D² / (8fc) where fc denotes the crystal bow and D the wafer diameter, both expressed in meters.

[0056] In the context of the present invention, the truncation angle at a point on the surface of a plate is also defined as the angle between:

[0057] - the normal to a front face of the element 13 nitride wafer (normal to the point for which the truncation angle is defined) and - the normal (also known as the C axis) to a crystal plane with orientation C (0001) (normal to the crystal plane at the aforementioned point).

[0058] The truncation angle can be measured by X-ray diffraction according to SEML MF26, entitled "Test Method for Determining the Orientation of a Semiconductive Single Crystal." The variation in truncation angle is defined as the difference between the highest and lowest truncation angles.

[0059] Within the framework of the present invention, the crystalline quality is also defined by the full width at half maximum of the X-ray diffraction peak (XRD) of the line (002) around the angle co in symmetrical condition of the planes (0001) of the crystal (see for example the publication "AMC 2017 - Advanced Materials Characterization Workshop June 6 & 7, X-ray analysis methods" by Mauro Sardela).

[0060] Grinding is defined as any cutting or machining operation consisting of removing more than 5%, or even more than 10% by mass of material from the crystal plate. Conversely, the so-called work hardening operation, even when it involves grinding with fine abrasive typically grade 200 mesh or higher according to the FEPA standard, results in very low material loss, typically less than 5%, or even less than 2%, or less than 1% by mass.

[0061] Wafer cleaning refers to any operation that removes organic and / or metallic contamination from EN surfaces to the ppb (parts per billion) level. It may consist of a chemical bath well-known in the semiconductor industry and chosen from among the SCI, SC2, Piranha, or carrot baths. Brief description of the drawings

[0062] Other advantages and features of the process according to the invention and of the associated product will become clearer from the following description of several embodiments, given by way of non-limiting examples, based on the accompanying drawings in which:

[0063] [fig.lA]

[0064] [fig.1B]

[0065] - Figures IA and IB illustrate the determination of the geometric arrow of a rough crystal.

[0066] [fig.2A]

[0067] [fig.2B]

[0068] [fig.2C]

[0069] [fig.2D]

[0070] - Figures 2A, 2B, 2C and 2D describe crystals exhibiting a crystal curvature or crystalline arrow different from their geometric curvature or arrow.

[0071] [fig.3]

[0072] - Figure 3 illustrates a method for manufacturing a raw crystal according to the invention.

[0073] [fig.4]

[0074] - Figure 4 illustrates the obtaining of a rough crystal according to the invention.

[0075] [fig.5]

[0076] - Figure 5 describes the rough crystal according to a first example of embodiment.

[0077] [fig.6]

[0078] - Figure 6 describes the rough crystal according to a second embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0079] With reference to figures 2 and 3, the main phases of the GaN wafer manufacturing process are illustrated.

[0080]

[0081]

[0082]

[0083]

[0084]

[0085]

[0086]

[0087]

[0088] The following will describe the process according to the invention with reference to the fabrication of gallium nitride (GaN) wafers. However, it is obvious to those skilled in the art that the process described below can be used to grow a material comprising a nitride layer of element 13 other than gallium nitride (GaN). Preferred but not exhaustive aspects of the product and process described above are described in the following sections: - The crystal is a gallium nitride crystal. - The crystal has a radius of curvature greater than 15 meters, preferably greater than 18 meters, more preferably greater than 20 meters, or even greater than or equal to 25 meters. - The crystal has a dislocation density of lOAm-2 or less. Advantageously, the truncation angle is non-zero in the vicinity of the center of the face E, for example of the Gallium face, is between 0.1 and 5 degrees, preferably between 0.2 and 0.8 degrees, even more preferably between 0.3 and 0.6 degrees, the cutting planes extending perpendicularly to the principal axis A of the crystal, in particular along the plane a and m. 1. Manufacturing process The process according to the invention comprises more particularly the following successive phases: - a growth phase (10) of a first layer of element 13 nitride (5a) on a starting substrate (1) having a crystallization plane whose truncation angle is between 0.1 and 5 degrees, - a formation phase (20) of a separation zone (4); - a growth resumption phase (30) to form a second layer of element 13 nitride (5b), - a separation phase (40) to obtain a self-supporting crystal (5) - a finishing phase (50) of the raw crystal obtained, characterized in that it understand : -at least one work-hardening step (52) by damaging the surface corresponding to a concave crystalline curvature of said crystal, in order to straighten the crystalline planes, and -at least one rectification step (53) allowing the crystal to be geometrically straightened, in order to obtain a plate according to the invention. -optionally a preliminary thinning step (51) by removing material from the rough crystal according to the initial deflection of said crystal, in order to obtain a final plate whose maximum deflection is less than 6xl0 4mm / mm of the greatest length of said plate.

[0089] 1. Growth Phase 10

[0090] The optional growth phase 10 consists of forming a sublayer of GaN 5a by lateral overgrowth.

[0091] Lateral overgrowth makes it possible to minimize the density of defects contained in the sublayer of GaN 5a.

[0092] The approach used to reduce the dislocation density in the GaN 5a sublayer consists of:

[0093] - initiate an island-based growth mode of GaN, then to - promote the coalescence of the islands to obtain the GaN 5a sublayer.

[0094] Advantageously, the lateral overgrowth is implemented on a starting substrate 1 having a non-zero truncation angle.

[0095] Using a starting substrate 1 with a non-zero truncation angle allows a first layer of GaN 5a with a non-zero truncation angle to be grown.

[0096] The starting substrate 1 can be chosen from Si, AIN, GaN, GaAs, A12O3 (sapphire), ZnO, SiC, LiA102, LiGaO2, MgAl2O4, 4H-SiC, or any other type of starting substrate known to those skilled in the art to implement gallium nitride growth.

[0097] The starting substrate 1 may have a thickness of a few hundred micrometers, generally 350 micrometers.

[0098] Advantageously, the starting substrate 1 can be treated by nitriding prior to any deposition step. This improves the quality of the resulting GaN crystal.

[0099] The truncation angle can be between 0.1 and 5 degrees, preferably between 0.2 and 0.8 degrees, and even more preferably between 0.3 and 0.6 degrees (in particular to limit stacking errors).

[0100] The growth of the GaN 5a sublayer can be implemented according to different variants. In particular, lateral overgrowth can be based on:

[0101] - on the use of a dielectric mask 3 including openings 3a in which islands are formed, as described in document WO99 / 20816; - on the use of a dielectric layer without openings on which islands spontaneously form, as described in document EP 1 338 683. - First variant of lateral overgrowth

[0102] In a first variant, phase 10 of growth consists of an epitaxial lateral overgrowth (hereinafter referred to as "ELO" according to the acronym of the Anglo-Saxon expression "Epitaxial Lateral Overgrowth").

[0103] The ELO includes a step of depositing a thicker planar layer 2 onto the starting substrate 1.

[0104] This deposition is preferably carried out by organometallic vapor phase epitaxy (or “MOVPE”, acronym for the Anglo-Saxon expression “MetalOrganic Vapor Phase Epitaxy”), for example at a temperature between 500°C and 700°C, in particular 600°C.

[0105] The deposition of a layer 2 reduces the stresses between the starting substrate 1 and the GaN 5a sublayer subsequently epitaxially deposited on it. Indeed, the deposition of layer 2 on substrate 1 ensures a "smooth" transition between substrate 1 and the GaN 5a sublayer, whose crystalline structures are different.

[0106] The deposition of layer 2 also facilitates the subsequent separation of the GaN 5 crystal, as will be shown in the following description. Layer 2 is, for example, a GaN layer, an AIN layer, or an AlGaN layer.

[0107] In another step, a mask 3 including openings 3a is formed. The openings 3a can be point-like or in the form of bands, and allow positions to be defined for the subsequent selective growth of GaN islands.

[0108] The mask 3 can be a mask made of a dielectric material, such as SiNx (SiN, Si3N4, etc.) or SiO2 or TiN. This minimizes defects created at the mask edge and thus improves the quality of the GaN layer subsequently epitaxially deposited on it.

[0109] The formation of mask 3 can be achieved by any technique known to a person skilled in the art. For example, mask formation may consist of:

[0110] * the deposition of a dielectric layer 3b from gaseous precursors of silane and ammonia directly onto layer 2, and • etching by photolithography of the dielectric layer 3b to form openings 3a.

[0111] This gives us a starting substrate 1 covered with a layer 2 and a mask 3. In addition to its function of improving the quality of the GaN 5a sublayer (by filtering through defects), the mask 3 makes it possible to weaken the interface between the starting substrate 1 and the first layer of GaN 5a.

[0112] Another step consists of forming GaN islands through the openings 3a of the mask 3. The growth rate along an axis orthogonal to the principal plane of the starting substrate 1 is kept higher than the lateral growth rate. This results in islands or bands with triangular cross-sections (depending on the shape of the openings 3a). Within these triangular bands, the through-dislocations are bent at 90°.

[0113] Lateral growth is then carried out to ultimately produce a flat ELO layer. At the end of this step of the process, a first layer of GaN 5a is obtained with a dislocation density less than 1OAm-2.

[0114] 1. Second variant of lateral overgrowth

[0115] In a second variant, the growth phase 10 consists of a universal lateral overgrowth (hereinafter referred to as "ULO" according to the acronym for the Anglo-Saxon expression "Universal Lateral Overgrowth") as described in document EP 1 977 028.

[0116] The ULO includes a step of depositing a nucleation layer on the starting substrate 1.

[0117] The nucleation layer is, for example, a very thin film of silicon nitride (SiN), on the order of a few atomic planes, in other words, on the order of 10 nm to 20 nm thick. The deposition of SiN based on silane and ammonia can last 360 seconds.

[0118] A continuous buffer layer 2 – for example, of GaN – is then deposited on the nucleation layer. The deposition of the GaN buffer layer 2 filters out crystalline defects and thus minimizes, from the beginning of the process, the defect density that will be present in the subsequently epitaxially grown GaN 5a sublayer. The thickness of this GaN buffer layer 2 can be between 10 and 100 nm. The temperature during this operation can be between 500 and 700°C.

[0119] The process then involves high-temperature annealing at a temperature between 900 and 1150°C. Under the combined effect of the temperature increase, the presence of a sufficient quantity of hydrogen in the gaseous vehicle, and the presence of the very thin SiN film, the morphology of the GaN buffer layer 2 undergoes a profound modification resulting from solid-phase recrystallization by mass transport. The initially continuous GaN buffer layer 2 is then converted into a discontinuous layer of GaN motifs. This yields GaN motifs or islands of very high crystalline quality that maintain an epitaxial relationship with the starting substrate thanks to the very small thickness of the nucleation layer.

[0120] The areas where the silicon nitride SiN is exposed then function as a mask, and the GaN motifs function like the GaN areas located in the ex-situ openings in the mask. Lateral growth is then carried out to ultimately result in a planar ULO layer.

[0121] This method, where the silicon nitride mask forms spontaneously, and which involves the same dislocation bending mechanisms as in ELO, is identified as "ULO" (or "spontaneous ELO").

[0122] 1. Phase 20 of the formation of a separation zone 4

[0123] The process further includes a phase 20 of forming a separation zone 4.

[0124] This phase 20 of the formation of a separation zone can be implemented according to different variants. In particular, phase 20 of the separation zone formation can be implemented:

[0125] - prior to phase 10 of GaN sublayer growth (first variant), or - subsequent to phase 10 of GaN sublayer growth (second variant), or - during phase 10 of GaN sublayer growth (third variant). - First variant of the formation of the separation zone 4

[0126] In a first variant, phase 20 of formation of a separation zone 4 may consist of depositing a sacrificial intermediate layer, prior to phase 10 of growth of the GaN 5a sublayer, as described in document EP 1 699 951.

[0127] The intermediate layer, which can be SiO2; SiN; Si; SiON; A1ON; Al, ZnO, Ti; TiO2; TiN; TiC; Ni; InN; MoSi2 or ZrN, serves as a sacrificial layer intended to be spontaneously vaporized during the subsequent growth phase by epitaxy of the GaN 5a sublayer.

[0128] 1. Second variant of separation zone formation

[0129] In a second embodiment, phase 20 of the formation of a separation zone 4 includes an implantation step carried out after phase 10 of growth of the GaN 5a sublayer. This implantation allows the creation of a weakened zone in the GaN 5a sublayer.

[0130] The implantation consists of bombarding the GaN 5a sublayer with ions so as to create in the semiconductor, at a depth close to the average penetration depth of these ions, a layer of microcavities (or bubbles).

[0131] The implanted ions may be selected from tungsten, helium, neon, krypton, chromium, molybdenum, iron, hydrogen, or boron. Preferably, the implanted ions are tungsten ions. These have the property of decomposing GaN.

[0132] In terms of dose, when the implanted ions are H+ ions, the dose of implanted ions can be between 1016 and 1017 cm2, and the implantation depth can vary between 0 nm and 50 nm from the free surface - called growth face - of the GaN 5a sublayer.

[0133] Embrittling ion implantation can be carried out in a single step or in successive steps. The temperature can be between 4 K and 1400 K during the implantation step. Implantation can be followed by an annealing phase to repair the crystal damage created during ion implantation; this annealing can be carried out at a temperature between 500°C and 1500°C.

[0134] 1. Third variant of separation zone formation

[0135] In a third variant, the separation zone 4 can be formed during phase 10 of growth of the GaN 5a sublayer.

[0136] In particular when the growth phase is carried out according to the first embodiment variant called ELO (i.e. deposition of a dielectric mask 3), the phase 20 of formation of the separation zone 4 can include the implantation of the buffer layer 2 prior to the deposition of the mask 3.

[0137] This allows the separation zone 4 to be placed at a precisely desired depth because the first layer of GaN 5a deposited during the ELO overgrowth step does not "hinder" ion implantation.

[0138] Of course, implantation can be carried out at different stages of the lateral overgrowth phase ELO (or ULO), either in the islands, or at an intermediate stage where the islands are not fully coalesced, or after total coalescence of the islands.

[0139] 1. Epitaxial resumption phase 30

[0140] Following the phases 20 of formation of a separation zone 4 and growth 10 of the sublayer of GaN 5a, the process includes a resumption of epitaxy 30 to form a first thick layer of GaN 5b undoped, a second thick layer 5c of GaN doped n and a third thick layer of GaN 5d undoped.

[0141] The process can also be started directly at this stage 30, by forming a thick layer of GaN 5b, the growth stages 10 and the formation of the separation zone 20 being optional. In the following text, it is assumed that these stages 10 and 20 are applied.

[0142] This re-establishment by epitaxy can be implemented by:

[0143] - Organometallic Vapor Epitaxy (“MOVPE”); - Halide Vapor Epitaxy (or "HVPE", acronym for the Anglo-Saxon expression "Hydride Vapor Phase Epitaxy"); - Closed-Space Vapor Transport (CSVT); or by - Liquid Phase Epitaxy (or "LPE" abbreviation for the Anglo-Saxon expression "Liquid Phase Epitaxy"). - Oxide Vapor Phase Epitaxy (or "OVPE," an acronym for the English expression "Oxide Vapor Phase Epitaxy")

[0144] In this step, we prefer to implement the "HVPE" technology which allows us to obtain three main interesting effects:

[0145] - A first effect is that the GaN 5a sublayer is thickened without losing its crystalline qualities (no new dislocation or crack is generated), - A second effect is that the dislocation density is further reduced during resumption of epitaxy in HVPE, by a factor of at least 2, beyond 100pm of GaN (0001) growth (ref https: / / doi.Org / 10.1143 / APEX.5.095503) - A third effect is that the thick layer of GaN 5 thus obtained can in In certain cases, allow spontaneous separation of its starting substrate 1 at the separation zone 4 in the event of sublimation or mechanical fracture of said zone during HVPE growth.

[0146] More specifically, the regrowth is carried out according to the following process: the temperature rise is achieved in a mixed atmosphere of nitrogen, ammonia, and hydrogen. As soon as a temperature of approximately 1000°C is reached and maintained stably, the growth phase of a thick layer of GaN is initiated by introducing, for example, gallium chloride (GaCl) into the vapor phase. The GaCl and ammonia are partially pyrolyzed in the growth chamber, the temperature of which is maintained at approximately 1000°C. Thus, a single-crystal deposit of GaN gradually forms at the nucleation substrate (formed during the first growth phase).

[0147] It is necessary to obtain a sufficiently thick GaN layer, and therefore sufficiently strong from a mechanical point of view, to prevent the GaN layer from fracturing into small pieces during separation and to facilitate its handling without risk of breakage. Growth is thus continued for several hours under these experimental conditions in order to reach a thickness of at least 200 microns of the GaN layer and preferably a thickness greater than 1 mm.

[0148] Growth is then definitively completed by diverting the HCl flow outwards and cooling takes place in an atmosphere consisting of nitrogen and ammonia.

[0149] The growth conditions of these first, second, and third single-crystal layers 5b, 5c, 5d are typically a growth temperature between 900 and 1200°C, with a growth rate between 50 and 500 micrometers / h, preferably between 70 and 200 micrometers / h.

[0150] The raw self-supporting GaN crystal thus obtained has a thickness greater than 200 µm and preferably greater than 1 mm. Its maximum thickness is less than 10 mm or even less than 5 mm.

[0151] The diameter of the raw GaN self-supporting crystal thus obtained is greater than 50 mm and preferably greater than 100 mm. Its maximum diameter is less than 250 mm or even less than 200 mm.

[0152] With reference to [fig.4], layer 5b has areas of permanent defects corresponding to hollows or pits of decreasing cross-section in the direction opposite to the direction of growth.

[0153] In a second step, under the same growth conditions, doping is carried out by adding the dopant element n, to obtain a second single-crystal layer of GaN 5c on the 5b layer, according to the following process:

[0154] - Oxygen supply and concentration are controlled by controlling the purity of the element III precursor and carrying out very thorough purging of the reactor before growth under vacuum at a residual pressure of less than 500 torrs. - For Germanium: from a solid source, GeCl4, germane, tetramethylgermanium, and isobutylgermane, as well as their derivatives. These dopant gases are then vaporized in the reaction chamber. Preferably, these dopant gases can be pre-mixed in the vapor phase with the GaCl stream to improve the homogeneous distribution of the doping flux in the growth chamber.

[0155] In the case of a gaseous precursor, the gas reservoir is maintained at a pressure between 1 bar and 3 bars and a carrier gas flow (N2 and / or H2) with a flow rate between 0.25 sccm and 20 sccm is applied.

[0156] - For Silicon, starting from silane, dichlorosilane, silicon tetrachloride, and their derivatives, which are vaporized in the reaction chamber. In the case of dichlorosilane (1% diluted in 99% N2 (or H2)), a flow rate of between 1 and 20 sccm is applied. Preferably, these doping gases can be premixed in the vapor phase with the GaCl flow rate to improve the homogeneous distribution of the doping flow rate in the growth chamber. - Silicon and germanium can be introduced together, leading to a 3-dopant system.

[0157] Typically the thickness of this first single-crystal layer of GaN 5b is from 100 to 1000 micrometers.

[0158] The permanent defect zones 6 of layer 5b result in the defects also propagating in layer 5b. Thus, during the growth of the second doped layer 5c, depressions of material are obtained in layer 5c in depression zones coinciding with the defect zones of layer 5b.

[0159] Other forms of material depression may appear linked to inclusions or nucleations which locally modify the growth rate without reaching the upper surface of layer 5b or even 5c.

[0160] In a third step, the supply of dopant gas n is stopped, and the growth conditions are maintained, allowing a third single-crystal layer of GaN 5d, typically 200 to 5000 micrometers thick, to be obtained on the previous doped 5c layer typically 100 to 2000 micrometers thick.

[0161] The thickness obtained of the single-crystal layer and the growth mode of GaN 5d allows the depression zones of the 5c layer to be filled by the deposition of this third layer.

[0162] 1. Separation phase 40

[0163] A separation phase 40 is also implemented, this is dependent on the variant implemented for phase 20 of the formation of the separation zone 4.

[0164] In the case of ion implantation, the spontaneous separation phase 40 takes place due to the thermal cycle (resumption of epitaxy at high temperature and cooling) which the thickened layer of GaN 5 undergoes, which, because of the difference in coefficients of thermal expansion between the starting substrate 1 and the thick layer of GaN 5, generates stresses causing its separation.

[0165] In the case of the deposition of a sacrificial intermediate layer, this separation occurs during the resumption of epitaxy by spontaneous vaporization of the intermediate layer or by mechanical fracture at the level of the so-called sacrificial layer.

[0166] In the case of post-growth separation, a laser can be used to vaporize the sacrificial layer.

[0167] This gives us a self-supported GaN 5 crystal as illustrated in figures IA or IB.

[0168] Such a crystal, possibly convex, has a radius of curvature typically greater than 5 meters and less than 25 meters, and preferably less than 20 meters. Furthermore, crystal 5 also has a dislocation density of 107 cm⁻² or less, preferably less than 5 × 10⁶ cm⁻²

[0169] The GaN 5 crystal, having been formed on a starting substrate with a non-zero truncation angle (or "miscut" in English), also has a non-zero truncation angle, the orientation of the crystal planes propagating from one layer to the next. For example, in the case of a sapphire substrate 1 with a truncation angle of 4 degrees, the growth face of the crystal 5 has a truncation angle of 4 degrees, preferably between 0.1 and 1 degree over its entire surface.

[0170] 1. Definition phase 50

[0171] Once the GaN 5 crystal has been separated from the starting substrate 1, a finishing phase 50 is carried out on the resulting rough crystal to obtain a final wafer with a maximum deflection of less than 6 x 10⁴ mm / mm crystal width. The finishing phase 50 may include:

[0172] -Optionally, a preliminary thinning step 51 by removing material from the rough crystal according to the initial deflection of said crystal. Preferably, the thinning is carried out by machining to a relative thickness greater than 5x103 and less than 102, said thickness corresponding to the ratio of the thickness of the wafer to the greatest length of its faces.

[0173] -At least one work hardening step 52 by damaging the surface corresponding to a concave crystal curvature of said crystal, in order to straighten the crystal planes to reduce the variation in the truncation angle. The work hardening may be Work hardening is achieved by surface abrasion, for example by sandblasting, shot peening, grinding, honing, or exposure to laser radiation to generate a work-hardened layer. Work hardening can also be achieved by another technique adapted to the mechanically stressed cutting of element 13 nitride single crystals. According to a specific method, work hardening is carried out until the variation in the truncation angle is less than 5 x 10³° / mm of the largest dimension of said wafer.

[0174] Preferably, the work hardening is carried out on the face of the plate opposite which the crystal deformation is concave or for which the curvature of the crystal lattice is concave.

[0175] Advantageously, work hardening is carried out until a crystal deflection is obtained, in absolute value, between 0.2 and 2, preferably 0.3 and 1.5, or even between 0.4 and 1.0 pm / mm of the largest dimension of the insert, in the case where the N face of the insert is concave. In the case where the N face of the insert is convex, work hardening is preferably carried out until a crystal deflection is obtained, in absolute value, between 0.1 and 1.5, preferably 0.2 and 1.2, or even between 0.2 and 1.0 pm / mm of the largest dimension of the insert.

[0176] -At least one rectification step enabling the crystal to be geometrically straightened. The N and / or E face and the sides or edges of the wafer are rectified, and polished to obtain a surface finish acceptable for the application of optoelectronic or electronic components.

[0177] The finishing step may also include other steps, such as chemical defoaming of the N face after machining of the N face or after polishing of the E face, cleaning to decontaminate the surfaces of the wafer, and / or preparation of the E face, in particular by plasma etching, for subsequent deposition of at least one layer necessary for the realization of an optoelectronic or electronic component.

[0178] Thus, the proposed process is particularly suitable for the manufacture of wafers or plates of semiconductor material, in particular wafers or plates of nitride of group 13 or III elements of the periodic table, more particularly wafers or plates composed of nitride of element 13, preferably of GaN, of large diameter, greater than 50 mm, more than 100 mm or even 150 to 200 mm.

[0179] The final wafer or plate of semiconductor material according to the invention has a thickness of 200 to 2000 micrometers and an excellent crystalline quality such that the full width at half maximum of the X-ray diffraction (XRD) peak of the line (002) around the angle co in symmetrical conditions of the GaN films (0001) is less than 130 Arc.sec in the case of a Gallium nitride wafer.

[0180] The crystal undergoes a trimming operation followed by cutting into several slices or plates typically having a thickness of 100 to 600 micrometers using a wire saw either "loose wire saw" (abrasive particles in a slurry which impregnates the wire before cutting), or "fixed wire saw" (abrasive particles previously fixed on the wire).

[0181] The crystal is cut along transverse cutting planes P orthogonal to the principal axis A of the crystal. This allows for a material yield much better than that described in US patent 2012 / 0184090. Current techniques allow for control of layer thickness removal to within 10 micrometers.

[0182] The present invention and its advantages are illustrated by means of the following examples. The examples according to the invention should not, of course, be considered as limiting its implementation. Examples of completed projects

[0183] Raw GaN crystals of 100mm diameter were obtained by growth via HVPE on the basis of a substrate as described for example by WO / 03100839A2 according to step 3. The flow (volume) ratio of N2(N2+H2) is 0.2. The crystals typically have a thickness between 500 and 2000 micrometers.

[0184] In the following examples, crystals according to [Fig. 5] having a concave N face and curved crystal planes following the geometric curvature of the crystal were obtained. A first series was treated according to the finishing process according to the invention and the other series according to a known process for comparison.

[0185] The final crystals in the examples below exhibit a ratio (FWHMN-FWHMGa) / wafer thickness where FWHMN denotes the full width at half maximum of the X-ray diffraction (XRD) peak of the (201) line around the angle co under asymmetric conditions of the (0001) GaN planes on the Nitrogen face and FWHME that on the element 13 face, which is -0.4 arcsec / micron of wafer thickness.

[0186] The geometric bow or curvature of the wafer for a wafer diameter of 100 mm was measured on the N face using a micrometric comparator of profilometer type as described by EP1528591A2

[0089] and the crystal bow for a wafer diameter of 100 mm was measured by ray diffraction as described in the publication "curvature and bow of bulk GaN substrates" by Humberto M.Foronda et al. published in the Journal of Applied Physics 120, 035104 (2016).

[0187] The truncation angle, expressed in degrees per 100 mm wafer width, was measured on the wafer by X-ray diffraction according to SEML MF26. At each point on the wafer, the truncation angle corresponds to the difference between the crystal plane m and the plane normal to the surface at that point. The variation in truncation angle is defined as the difference between the highest and lowest angles, for a 100 mm wafer diameter. Steps f), h), and i) were not performed. detailed in the examples below because the differences in angle variation are small but an overall angle variation has been specified for all of these steps.

[0188] The results are presented in Table 1 below:

[0189] [Tables 1] Process steps: Definition of resulting deflection at each successive processing step (geometric and crystalline deflections are measured with reference to the N face) Example 1 invention Comparative example Comparative example 1b Initial step - raw crystal Geometric deflection (qm) Crystalline deflection (qm) Truncation angle variation +115 +115 0.20 +121 +121 0.20 +115 +115 0.20 Thickness adjustment (optional) Geometric deflection (qm) Crystalline deflection (qm) Truncation angle variation (Machining Ga face) +155 +155 0.71 Na (not applied) (Machining Ga face) +155 +155 0.71 c) Work hardening N face Geometric deflection (qm) Crystalline deflection (qm) Truncation angle variation (Sandblasting N face) -50 -50 0.23 Na Sandblasting face N until a zero geometric deflection is obtained dj-rectification face N Geometric deflection (qm) Crystalline deflection (qm) Truncation angle variation -40 -40 0.18 +55 +140 0.64 +48 + 135 0.62 gj-Chemical attack or frosting of the N face Geometric deflection (qm) Crystalline deflection (qm) Truncation angle variation -20 -20 0.10 -21 +110 0.50 -13 +105 0.48 ej-Ga face rectification Geometric deflection (qm) Crystalline deflection (qm) Truncation angle variation +70 +70 0.32 +61 +150 0.69 +52 +140 0.64 f) h) i)-Finished product after cleaning and +40 +41 +27 . Plasma engraving +40 +115 +107 Geometric deflection (pm) -0.8 -0.8 -0.8 Crystalline deflection (pm) Crystalline quality gradient (arcsec / micron) Truncation angle variation 0.15 0.51 0.49

[0190] Na = not applied

[0191] As can be seen in Table 1, the proposed process, compared to a process without a work-hardening step (a), allows for a significant reduction in crystal deflection and truncation angle variation. In particular, the preliminary work-hardening step makes it possible to straighten or even reverse the crystal curvature, in order to anticipate the deformations generated by the subsequent finishing operations, especially the plate grinding operation. This preliminary work-hardening step thus makes it possible to obtain a plate with very low crystal deflection and geometric deformation. In the following examples, crystals according to [Fig. 6] with a convex N face and curved crystal planes following the geometric curvature of the crystal were obtained. One series was treated according to the finishing process of the invention and the other series according to a known process for comparison.

[0192] Also in comparison with a process where in step c) the wafer is straightened in a sandblasting step until a zero geometric deflection is obtained (1b), a significant crystalline deflection is obtained for the finished product.

[0193] The results are presented in Table 2 below:

[0194] [Tables 2] Process steps: Definition of resulting deflection at each successive processing step (geometric and crystalline deflections are measured with reference to the N face) Example 2 invention Example 2 comparison Initial step - raw crystal Geometric deflection (qm) Crystalline deflection (qm) Truncation angle variation -141 -141 0.6 -132 -132 0.5 Thickness adjustment (optional) Geometric deflection (qm) Crystalline deflection (qm) Truncation angle variation (Machining Ga face) -60 -60 0.28 Na c) Work hardening Ga face Geometric deflection (qm) Crystalline deflection (qm) Truncation angle variation (Sandblasting Ga face) -30 -30 0.14 Na N face deflection Geometric deflection (qm) Crystalline deflection (qm) Truncation angle variation -90 -91 0.41 -50 -178 0.82 gf Chemical attack or frosting face N Geometric deflection (qm) Crystalline deflection (qm) Truncation angle variation -60 -61 0.28 -20 -145 0,66 e-rectification face Ga Geometric deflection (qm) Crystalline deflection (qm) Truncation angle variation +74 +12 0.06 +102 -60 0.28 f), h), i)-Finished product after cleaning and plasma engraving Geometric deflection (qm) Crystalline deflection (qm) +23 -42 -0.8 0.19 +50 -115 -0.8 0.53 , Crystalline quality gradient (arcsec / micron) Truncation angle variation

[0195]

[0196]

[0197]

[0198]

[0199]

[0200]

[0201] As can be seen in Table 2, the proposed process, compared to a traditional method, allows for a significant reduction in crystal deflection and truncation angle variation. The proposed process thus enables the production of a 13-element nitride crystal wafer with high crystalline quality and a crystalline quality gradient exceeding 5%, where truncation angle variation and geometric deformation are consistently very low despite the multiple deformations generated by the various finishing steps between obtaining the rough crystal and the final wafer ready for the fabrication of an optoelectronic or electronic component. In particular, it is possible to obtain a very small variation in the truncation angle by combining work hardening, not with the aim of geometrically straightening the crystal, but with the aim of reducing its truncation angle variation. The geometric deformation of the crystal due to this work hardening step is then "corrected" by subsequent grinding steps, which will reduce the geometric deformation of the crystal. Advantageously, the curvature or geometric deformation of its faces has a deflection less than 103 mm / mm of the largest dimension of said plate. Also, advantageously, the quality gradient is between -0.005 arcsec / micron and -5 arcsec / micron. Also advantageously, the crystal has a dislocation density of 107 cm⁻² or less Also advantageously, the proposed process makes it possible to obtain a plate in which: - at least one face, preferably face E, has a mean square surface roughness (defined according to standard NF EN ISO 4287 December 1998) of less than 0.5 nm over an area of ​​4 pm2, and / or, in which the crystal has a dislocation density of 107 cm-2 or less; and / or - the measured electrical resistivity is less than 25 mOhms.cm; and / or - the absorption coefficient measured at a wavelength of 400nm is less than 4 cm -1.

Claims

Claims

1. Two-dimensional element 13 or III nitride crystal wafer, delimited by a face of orientation N, an opposite face of orientation E according to element 13 or III, E preferably being chosen from Ga, In, Al or a combination of these elements, characterized in that, a diameter of said wafer is greater than 50 mm, a variation in the crystalline truncation angle according to a larger dimension of said wafer is less than 5x10 3° / mm, and - a geometric curvature or geometric deformation of its faces has an arrow less in absolute value than 103 mm / mm of the larger dimension of said wafer.

2. Plate according to the preceding claim, the crystal curvature or crystal deflection of which is less in absolute value than 103 mm / mm of the largest dimension of said plate.

3. Plate according to one of the preceding claims, the face E of which is convex and the geometric deflection of said face E is negative and greater than -103 mm / mm of the largest dimension of said plate, preferably greater than -0.7x103 mm / mm of the largest dimension of said plate.

4. Wafer according to one of the preceding claims in which FWHME denotes the width at half-height of the X-ray diffraction peak of the line (002) around the angle co measured on the E face, and where FWHME is less than 130 Arcsec and the quality gradient defined by the ratio (FWHM'N-FWHM'E) / thickness of said wafer, where FWHM'N denotes the width at half-height of the X-ray diffraction peak of the line (201) around the angle co measured on the N face (Nitrogen) and FWHM'E is that measured on the E face, is in absolute value between 0.005 and 5 arcsec / micrometer.

5. Wafer according to one of the preceding claims, in which at least one face, preferably face E, comprises a surface having a crystalline quality and a roughness allowing the growth of a subsequent layer in a single direction.

6. Platelet according to one of the preceding claims, the crystal truncation angle of which along the largest dimension of said platelet is greater than 2x10 3° / mm.

7. Plate according to one of the preceding claims, the angle of which crystal truncation along the largest dimension of said wafer and measured at the center of said wafer is greater than 3x10 3° / mm and / or less than 7x10 3° / mm.

8. Platelet according to one of the preceding claims, in which element 13 or III is chosen from Ga, In, Al, preferably element 13 or III comprises Gallium.

9. A wafer according to any preceding claim, wherein element 13 or III is Gallium.

10. A method of manufacturing the wafer according to one of the preceding claims comprising the following steps: a: obtaining a self-supporting raw crystal wafer of element 13 nitride by epitaxial growth, b: determining the variation of the truncation angle of said raw crystal at the end of growth, optionally after cutting said crystal, preferably by grinding or by cutting with a wire saw, to a relative thickness greater than 5x103 and less than 102, said relative thickness corresponding to the ratio of the thickness of the wafer to the greatest width of its faces, c: work hardening of a face in order to reduce the variation of the truncation angle, in particular by sandblasting, exposure to laser radiation, shot peening, ion bombardment or another technique making it possible to create a surface mechanical stress on said crystal, so as to obtain a non-zero geometric deflection,preferably between 0.001 and 1 in absolute value, preferably between 0.01 and 0.1 mm / mm of the largest dimension of said wafer d: after work hardening, grinding of the face N by machining, e: after work hardening, grinding of the face E by machining, f: polishing of the face E, g: chemical etching of the face N after machining of the face N or after polishing of the face E, h: cleaning to decontaminate the surfaces of the wafer i: preparation of the face E, in particular by plasma etching, for subsequent deposition of at least one layer necessary for the production of an optoelectronic or electronic component.

11. Manufacturing method according to the preceding claim, wherein the work hardening step c in step is carried out until the variation in truncation angle is less than 5x10 3° / mm of the largest dimension of said plate.

12. A manufacturing method according to claim 10, wherein the work hardening step c is carried out on the face of the crystal for which the curvature of the crystal lattice is concave.

13. Manufacturing method according to claim 10, in which the work hardening step c is carried out until a crystal deflection of between 0.2 and 2 pm / mm of the largest dimension of the wafer is obtained, in the case where the face N of the wafer is concave.

14. Manufacturing method according to claim 10, in which the work hardening step in step c is carried out until a crystal deflection of between 0.1 and 1.5 pm / mm in absolute value of the largest dimension of the wafer is obtained, in the case where the face N of the wafer is convex.

15. Manufacturing method according to one of claims 10 to 14, wherein the step of growing the element 13 or III nitride wafer comprises the following phases: a growth phase (10) of a first layer of element 13 nitride (5a) on a starting substrate (1) having a crystallization plane whose truncation angle is between 0.1 and 5 degrees, preferably between 0.2 and 0.8 degrees, and even more preferably between 0.3 and 0.6 degrees; - a phase of forming (20) a separation zone (4); - a growth resumption phase (30) to form a second layer of element 13 nitride (5”), - a separation phase (40) to obtain the crystal (5).

16. Use of a wafer of element 13 or III nitride according to one of claims 1 to 9 for the manufacture of optoelectronic or electronic components, such as light-emitting diodes, laser diodes, transistors with vertical or horizontal geometry, power electronics or electronic components for radio frequency transmission, reception or amplification, current rectifier diodes or sensors.

Images