III-N Material Growth Process

Abstract

Title: Method for Growing a III-N Material. The invention relates to a method for growing a III-N material. A first step consists of providing a primer layer (100) based on a first III-N material. A buffer layer (200) is then grown epitaxially from the primer layer, and subsequently a germination layer (300) is grown epitaxially from the buffer layer. The germination layer is then bonded to a receiving substrate (2000). After the bonding step, the buffer layer is subjected to laser radiation having a principal wavelength belonging to an emission range in which the primer layer is configured to transmit a portion of the laser radiation, the laser radiation being configured to alter the buffer layer to decouple the primer layer and the germination layer. A layer of interest (400) is then grown epitaxially from the germination layer. Figure for the abstract: Fig. 1F

Description

Title of the invention: Method for growing a III-N material Technical field

[0001] The present invention relates to the growth of III-N materials, such as GaN. It advantageously allows the growth of III-N materials to produce so-called vertical microelectronic devices. It finds, for example, particularly advantageous applications in the field of power electronics and the field of micro-display. STATE OF THE ART

[0002] Power electronic components based on a III-N material are often produced by epitaxy of this material, in particular by organometallic vapor phase epitaxy, from crystalline substrates. The most used substrates are GaN substrates and silicon substrates.

[0003] GaN substrates are best suited for GaN epitaxy. The production of GaN layers by homoepitaxy makes it possible to obtain very low dislocation densities. There is no problem of lattice mismatch or thermal expansion coefficient, the epitaxial layers are therefore of crystalline quality close to that of the substrate and are not very constrained. The document Hu, J. et al. Materials and Processing issues in vertical GaN power electronics, Materials Science in Semi-conductor Processing 78, 75-84 (2018) presents the advantages of this type of substrate. On the other hand, GaN substrates are small and very expensive. The devices produced on these substrates are therefore generally efficient but very expensive, which is a very significant obstacle to the use of these substrates on a large scale.

[0004] Conversely, silicon substrates are available at low cost and in large dimensions, which makes it very common in microelectronics. It is possible to carry out heteroepitaxial growth of GaN on silicon substrates because the lattice of (111) oriented silicon is compatible with the crystal lattice of (001) oriented GaN. However, there is a lattice mismatch of -17% between GaN and silicon. GaN layers produced by GaN epitaxy on silicon therefore have significant densities of crystal defects such as dislocations. In addition, the coefficient of thermal expansion (CTE) at room temperature of Si is 2.59.106 while that of GaN is 5.59.106. Since GaN epitaxy is carried out at high temperature (of the order of 1000°C), high stresses are generated during the temperature decrease that takes place after GaN epitaxy.For the thicknesses targeted for vertical power components based on GaN, of the order of several micrometers. or even tens of micrometers, the constraints are such that the GaN layer can crack when the temperature drops. It may be possible to structure the layers obtained and in particular to facet their surface (see in particular Tanaka, A., Choi, W., Chen, R. & Dayeh, SA Si Complies with GaN to Overcome Thermal Mismatches for the Heteroepitaxy of Thick GaN on Si. Advanced Materials 29, 1702557 (2017)) in order to limit the quantity of structural defects, but this is difficult to implement and very expensive.

[0005] Furthermore, it is common to grow a GaN layer from a sapphire substrate, then to detach it from this substrate using a laser lift-off step (see in particular Wong, W.S., Sands, T. & Cheung, N.W. Damage-free separation of GaN thin films from sapphire substrates. Appl. Phys. Lett. 72, 599-601 (1998)). However, since the laser is absorbed by the GaN, part of the GaN layer is volatilized by the radiation and damages it, which is not satisfactory from the perspective of improving the quality of the devices. Furthermore, there is, as between GaN and silicon, a lattice mismatch and a CTE mismatch between GaN and sapphire. The green layer therefore has significant defect densities. It should also be noted that sapphire substrates are more expensive.

[0006] Furthermore, it has been investigated to use an ion cutting technique to separate a GaN layer from its monocrystalline substrate. However, this method has the disadvantage of damaging the GaN layer during its ion implantation. The GaN layer has a high roughness after the annealing used for the detachment of the GaN layer, which is not satisfactory. Furthermore, the detachment step following the ion implantation step requires a high temperature annealing, which limits the substrates on which the transfer can be carried out. For example, with this technique it is impossible to transfer a GaN layer onto a CMOS device (from the English "Complementary Metal Oxide Semiconductor").

[0007] The present invention therefore aims to propose a solution making it possible to limit, or even eliminate at least some of the drawbacks of known solutions. Another object of the present invention aims to propose a solution for growing an IILN material, preferably in a less expensive manner than homoepitaxy, and preferably in a manner making it possible to obtain a quantity of structural defects lower than existing heteroepitaxy methods. SUMMARY

[0008] To achieve this objective, according to one embodiment, a method for growing an IILN material, such as GaN, is provided, comprising the following steps: a. Providing a seed layer based on a first IILN material, the seed layer having an upper face extending mainly in a horizontal plane defined by a first direction and a second direction, b. Growing a buffer layer by epitaxy of a buffer material from the upper face of the seed layer, the buffer layer having an upper face and a lower face opposite each other, the lower face of the buffer layer facing, and preferably in contact with, the upper face of the seed layer, c. Growing by epitaxy a seed layer, in a second III-N material, preferably identical to the first III-N material, from the upper face of the buffer layer, the seed layer having an upper face and a lower face opposite each other, the lower face of the seed layer being opposite, and preferably in contact with, the upper face of the buffer layer, d. Bonding the upper face of the seed layer to a receiving substrate, e. Subjecting the buffer layer to laser radiation having a main wavelength Xiaser, the seed layer being configured to transmit at least part of the laser radiation, the seed layer preferably having an absorption of less than or equal to 10%, preferably less than or equal to 5%, preferably less than or equal to 2% for the main wavelength Xiaser, the laser radiation being configured to alter the buffer layer so as to separate the seed layer and the seed layer, f. Growing by epitaxy of a third III-N material, preferably identical to the second III-N material, a layer of interest from the lower face of the seed layer.

[0009] In this method, the separation of the seed layer and the seed layer is enabled by the alteration of the buffer layer due to its subjection to laser radiation. The method involves little collateral damage to the seed layer and the seed layer, which in particular makes it possible to grow a layer of interest having a low density of structural defects.

[0010] Furthermore, once the seed layer and the primer layer have been separated by alteration of the buffer layer, the primer layer can be reused for new iterations of the process. It is therefore possible to opt for a primer layer of excellent quality because its cost will be amortized due to its reuse.

[0011] By ensuring that the primer layer is transparent to laser radiation, it will be possible to ensure that it is not altered during the step of subjecting the buffer layer to radiation. Its quality will thus be preserved for its future uses during new implementations of the method.

[0012] The proposed solution thus constitutes an alternative to existing methods of growth of a III-N material presenting numerous advantages, notably the fact that the layer of interest obtained presents a very good crystalline quality without the process being as expensive as a homoepitaxy of this same III-N material. BRIEF DESCRIPTION OF THE FIGURES

[0013] The aims, objects, as well as the characteristics and advantages of the invention will emerge more clearly from the detailed description of an embodiment thereof which is illustrated by the following accompanying drawings in which:

[0014] [Fig.1A] Figures 1A to 11I illustrate a first embodiment of the method according to the invention, in which the primer layer forms a plurality of disjointed islands.

[0015] [Fig.lB]

[0016] [Fig.lC]

[0017] [Fig.lD]

[0018] [Fig.lE]

[0019] [Fig.lF]

[0020] [Fig.lG]

[0021] [Fig.lH]

[0022] [Fig. II]

[0023] [Fig.2A] Figures 2A to 2D illustrate a second embodiment of the method according to the invention, in which the primer layer forms a single continuous layer.

[0024] [Fig.2B]

[0025] [Fig.2C]

[0026] [Fig.2D]

[0027] [Fig.3A] Figures 3A to 3C illustrate steps of a third embodiment of the method according to the invention, in which an adhesion layer surrounds the primer layer, the buffer layer and the seed layer.

[0028] [Fig.3B]

[0029] [Fig.3C]

[0030] [Fig.4A] Figures 4A to 4G illustrate an embodiment in which electrical contacts are deposited on the seed layer and on the receiving substrate before bonding these two elements.

[0031] [Fig.4B]

[0032] [Fig.4C]

[0033] [Fig.4D]

[0034] [Fig.4E]

[0035] [Fig.4F]

[0036] [Fig.4G]

[0037] [Fig.5A] Figures 5A to 5E illustrate an embodiment in which the receiving substrate is removed and the layer of interest is transferred to a secondary receiving substrate.

[0038] [Fig.5B]

[0039] [Fig.5C]

[0040] [Fig.5D]

[0041] [Fig.5E]

[0042] [Fig.6A] Figures 6A to 6G illustrate an embodiment in which the receiving substrate is electrically conductive.

[0043] [Fig.6B]

[0044] [Fig.6C]

[0045] [Fig.6D]

[0046] [Fig.6E]

[0047] [Fig.6F]

[0048] [Fig.6G]

[0049] The drawings are given as examples and are not limiting of the invention. They constitute schematic representations of principle intended to facilitate the understanding of the invention and are not necessarily on the scale of practical applications. In particular the dimensions, in particular the thicknesses, are not representative of reality. DETAILED DESCRIPTION

[0050] Before beginning a detailed review of embodiments of the invention, optional features which may possibly be used in combination or alternatively are set out below:

[0051] According to an advantageous example, in the horizontal plane, the first IILN material has a first mesh length alu.N,i and the buffer material has a buffer mesh length atampon, with | (am.N,i- atampon ) | / aul.N,i < 0.02. This condition makes it possible to limit the mesh mismatch between the first IILN material and the buffer material. This makes it possible to improve the quality of the epitaxy of the buffer material from the first IILN material, to reduce the quantity of defects in the buffer layer and thus to improve its crystalline quality. In this way, the crystalline quality of the seed layer which is grown by epitaxy from the buffer layer is also improved, as well as the layer of interest which is grown by epitaxy from the seed layer.

[0052] In the case where the second IILN material is distinct from the first IILN material, preferably, in the horizontal plane, the second IILN material has a second mesh length a111N 2 and the buffer material has a buffer mesh length atampon, with | (aIU_N>2- atampon ) | / alu_N>2< 0.02. This condition limits the mesh mismatch between the buffer material and the second III-N material. This limits or even prevents the generation of defects during growth. This improves the quality of the epitaxy of the second III-N material from the buffer material, reduces the quantity of defects in the seed layer and thus improves its crystalline quality compared to if this precaution were not taken. In this way, the crystalline quality of the layer of interest which is grown by epitaxy from the seed layer is also improved.

[0053] According to an advantageous example, in the horizontal plane, the mesh of the first III-N material has a first angle aul.N4 and the mesh of the buffer material has an angle atampon, with | a111Nji-atampon | < 5°. In this way, the mesh mismatch between the first III-N material and the buffer material is limited. Here again, this makes it possible to limit the quantity of defects in the buffer layer, and consequently in the seed layer and in the layer of interest.

[0054] In the case where the second III-N material is distinct from the first III-N material, preferably, in the horizontal plane, the mesh of the second III-N material has a second angle alu.N,2 and the mesh of the buffer material has an angle atampon, with I «ni N,2-atamPon | < 5°. In this way, the mesh mismatch between the buffer material and the second III-N material is limited. Here again, this makes it possible to limit the quantity of defects in the seed layer, and consequently in the layer of interest.

[0055] According to one embodiment, the primer layer forms a support substrate. In this case, it does not rest on a support substrate different from the primer layer.

[0056] According to an advantageous example, the buffer layer has, in a third direction perpendicular to the horizontal plane, a thickness e2Oo configured so that the buffer layer has a dislocation density p of less than 1010 cm2, preferably less than 108 cm2, preferably less than 107 cm2. If the buffer layer is too thick, it relaxes and dislocations appear. Limiting the thickness of the buffer layer allows it to be coherent with the substrate and to have good crystalline quality, which will also have the effect of guaranteeing good crystalline quality of the layer of interest. It should be noted that the crystalline quality of the buffer layer is evaluated with respect to that of the substrate. Thus, limiting the thickness of the buffer layer makes it possible more precisely to limit the quantity of defects added compared to the defects initially present in the substrate and transmitted to the buffer layer.Preferably, when the primer layer rests on a support substrate, the dislocation density of this support substrate is denoted pSubStrat, and the thickness e200 is configured so that the difference between the dislocation density p of the buffer layer and the dislocation density psubstrat of the support substrate, p-pSUbstrat, is less than or equal to . to 1010 cm2, preferably 109 cm2, preferably 108 cm2. Preferably, when the seed layer is a single continuous layer and does not rest on a support substrate, and thus constitutes a substrate, the dislocation density of the seed layer is denoted psubstrate, and the thickness e2Oo is configured so that the difference between the dislocation density p of the buffer layer and the dislocation density psubstrate of the seed layer, p-pSUbStrat, is less than or equal to 106 cm2. Generally, the thickness e2Oo is configured so that the dislocation density p of the buffer layer is less than or equal to 1.5 times the dislocation density pSUbStrat of the support substrate.For a GaN substrate, psubstrate is typically less than 107 cm2, for example between 106 cm 2 and 107 cm 2, or even of the order of 104 cm2 (for example if the substrate is formed by an ammonothermal method or by oxide vapor phase epitaxy (commonly referred to as OVPE, from the English Oxide Vapor Phase Epitaxy)).

[0057] Advantageously, e200<100 nm, preferably e200<20 nm.

[0058] Preferably, e2Oo>2 nm and advantageously e2Oo>5 nm. This makes it possible to guarantee good absorption of the laser radiation by the buffer layer. The minimum thickness of the buffer layer making it possible to guarantee its alteration depends, however, on the buffer material and may be established on a case-by-case basis.

[0059] Preferably, the first IILN material has a melting temperature Tfusion lll N,i, the epitaxy of the buffer material being carried out at a growth temperature Tgrowth,buffer, with Tgrowth2ampon<'Tfusion,iii-N,i* This makes it possible to guarantee that the seed layer is not altered during the epitaxy of the buffer material. In general, the growth is carried out under conditions making it possible to limit or even prevent the degradation of the seed layer. Preserving a good quality seed layer throughout the process makes it possible, among other things, to be able to reuse it during new implementations of the process.

[0060] Preferably, the buffer material has a melting temperature Tmelting>buffer, the epitaxy of the second IILN material being carried out at a growth temperature Tgrowth,iii-N,2, with Tgrowthîm.^N^^Tmelting,buffer*This ensures that the buffer layer is not altered during the epitaxy of the second IILN material.

[0061] According to an advantageous example, in a third direction perpendicular to the horizontal plane, the primer layer has a thickness eiOo and the layer of interest has a thickness e400, with e4oo>2*eioo, preferably e4OO>3*eiOo, preferably e 4oo^5*eioo. This is particularly desired when the primer layer is a layer resting on a support substrate. According to an advantageous example, in a third direction perpendicular to the horizontal plane, the buffer layer has a thickness e 2oo and the layer of interest has a thickness e4Oo, with e4Oo>2*e2Oo, preferably e4Oo >3*e2Oo, preferably e4Oo>5*e2Oo.

[0062] According to a preferred embodiment, the laser radiation is configured to pass through the primer layer before altering the buffer layer.

[0063] According to an advantageous example, after the separation of the primer layer and the seed layer, at least a portion of the buffer layer remains on the seed layer, and the method further comprises, after the step of subjecting the buffer layer to laser radiation and before the step of growing the layer of interest, a step of removing said at least portion of the buffer layer.

[0064] According to one embodiment, the primer layer forms a single continuous layer.

[0065] According to one embodiment, the primer layer forms germination substrates disjointed.

[0066] According to a preferred example, each of the disjointed germination substrates has, in the horizontal plane, a maximum dimension less than 150 mm, typically less than 100 mm. In the typical case of germination substrates having, in projection in the horizontal plane, a circular shape, their maximum dimension corresponds to their diameter.

[0067] According to one embodiment, during the step of providing the primer layer, the germination substrates are transferred onto the support substrate.

[0068] According to an advantageous example, the method further comprises, after the step of epitaxy of the second III-N material and before the step of bonding the upper face of the seed layer to the receiving substrate, the formation of an adhesion layer on the upper face of the seed layer. In this case, the bonding of the seed layer to the receiving substrate is done via the adhesion layer. The presence of the adhesion layer ensures the quality of the bonding. The formation of this adhesion layer typically comprises a step of planarization of this same layer. Thus, it is guaranteed that the bonding to the receiving substrate is done at the level of a flat layer, which improves the quality of the bonding.

[0069] According to one example, the adhesion layer may for example be based on SiO2, SiN, Si, or any other material identified by a person skilled in the art depending on the intended application. Generally, it may be, among other things, a metal, a dielectric, a semiconductor or even a polymer. If it is desired to carry out a new epitaxy from the stack obtained using the method according to the invention, a dielectric will typically be chosen for its good temperature resistance. A polymer material may be used to produce an adhesive bond.

[0070] When the primer layer forms a plurality of germination substrates from which islands of the buffer layer and islands of the germination layer are grown, it is advantageously provided that, in projection in the horizontal plane, the adhesion layer extends around each of the germination substrates, each of the islands of the buffer layer and each of the islands of the germination layer, preferably being in contact with the sides of the germination substrates, the sides of the islands of the buffer layer and the sides of the islands of the germination layer. This ensures good mechanical support of the whole.

[0071] According to one example, the III-N material is based on at least one of GaN, AlN, InN and any combination of these materials.

[0072] According to one example, the buffer material is one of the following materials: HfN, ZrN, NbN, TiN.

[0073] Generally, the buffer material may be a nitride material. In the context of GaN growth, HfN is a particularly advantageous material because it has a mesh mismatch of only 0.2% with GaN. In addition, it can be deposited over a thickness of up to 100 nm without relaxing and therefore without having a detrimental dislocation rate. Furthermore, a layer of approximately 100 nm of HfN has an absorption rate of 99% in the infrared, which is very sufficient to ensure good alteration of the buffer layer when it is subjected to laser radiation. Typically, if the buffer material is HfN, a thickness e200 of the buffer layer is provided of substantially between 5 and 20 mm.

[0074] ZrN is another nitride whose properties enable it to be a good buffer material in the context of the invention. It has a very low lattice mismatch (1.3%) with GaN, can be deposited up to a critical thickness of 10 nm and has, for this critical thickness, an absorption rate of 45% in the infrared. It should be noted that with regard to the degradation condition of the buffer layer to allow the separation of the seed layer and the seed layer, a material having an absorption rate as low as 10% or even 2% can be perfectly suitable for the buffer layer. Sufficient degradation can be obtained for these absorption values.

[0075] According to an advantageous example, the method further comprises the formation of electrical contacts on the upper face of the seed layer and secondary electrical contacts on the upper face of the layer of interest.

[0076] According to an advantageous example, the receiving substrate is electrically conductive.

[0077] According to one embodiment, the receiving substrate comprises at least one integrated electronic circuit.

[0078] According to an advantageous example, the buffer material is chosen so that the buffer layer induces a polarity within the seed layer.

[0079] It is specified that, in the context of the present invention, the terms “on”, “overcomes”, “covers”, “underlying”, “facing” and their equivalents do not necessarily mean “in contact with”. Thus, for example, the deposition, transfer, bonding, assembly or application of a first layer on a second layer does not necessarily mean that the two layers are directly in contact with each other. contact with each other, but means that the first layer at least partially covers the second layer by being either directly in contact with it or by being separated from it by at least one other layer or at least one other element.

[0080] A layer may also be composed of several sub-layers of the same material or of different materials.

[0081] A substrate, a layer, a device, “based” on a material M, is understood to mean a substrate, a layer, a device comprising this material M only or this material M and possibly other materials, for example alloying elements, impurities or doping elements. Thus a material based on a III-N material may comprise a III-N material with added dopants.

[0082] A reference frame, preferably orthonormal, comprising the axes X, Y, Z is represented in figures 1A, 2A, 3A, 4A and 5A. This reference frame is applicable by extension to the other figures.

[0083] In the present patent application, we will preferably speak of thickness for a layer and of height for a structure or a device. The height is taken perpendicular to the horizontal plane XY. The thickness is taken in a direction normal to the main extension plane of the layer. Thus, a layer typically has a thickness along Z, when it extends mainly along the horizontal plane XY, and a projecting element, for example an isolation trench, has a height along Z. The relative terms “on”, “under”, “underlying” preferably refer to positions taken in the Z direction.

[0084] The terms “substantially”, “approximately”, “of the order of” mean “to within 10%, preferably to within 5%”.

[0085] A first embodiment of the method according to the invention will be described with reference to Figures 1A to 111.

[0086] [Fig.1A] illustrates the provision of a support substrate 1000 that can be used in the method according to the invention. This support substrate 1000 can for example be a SiC substrate, a sapphire substrate, a silicon substrate, a monocrystalline GaN substrate or any other substrate based on an IILN material. It has an upper face 1001 extending mainly in a horizontal plane XY defined by a first direction X and a second direction Y.

[0087] [Fig.1B] represents the formation of a seed layer 100 on the upper face 1001 of the support substrate 1000. The seed layer 100 is based on a first IILN material such as GaN. The seed layer is crystalline.

[0088] The primer layer 100 may be discontinuous, as illustrated in [Fig. 1B], or continuous, as will be presented in the context of the second embodiment illustrated in FIGS. 2A to 2D. In the case of a discontinuous primer layer 100, the latter may for example be made up of a plurality of ger substrates mination 100a, 100b, 100c disjoint. These disjoint germination substrates 100a, 100b, 100c may also be designated disjoint islands 100a, 100b, 100c of the seed layer 100. This set of disjoint germination substrates 100a, 100b, 100c may be referred to as a tiling. For example, according to an advantageous embodiment, the assembly illustrated in [Fig.lB] is obtained by bonding to the support substrate 1000 a plurality of substrates with dimensions smaller than those of the support substrate 1000. These are typically substrates based on monocrystalline GaN, the diameter of which is currently most often between 50 mm and 150 mm, bonded to a support substrate 1000 whose dimensions allow a plurality of them to be carried. Such a support substrate 1000 may for example be a silicon substrate with a diameter of 200 or 300 mm.

[0089] The primer layer 100 has a lower face 102 facing, and preferably in contact with, the upper face 1001 of the support substrate 1000, and an upper face 101 opposite its lower face 102.

[0090] As illustrated in [Fig.lC], a buffer layer 200 is grown from the upper face 101 of the primer layer 100.

[0091] The buffer layer 200 is based on a material called buffer material whose structural properties must allow its epitaxy from the seed layer 100. It is expected in particular that the crystalline parameters of the buffer material and of the first IILN material are close.

[0092] The buffer layer 200 can be grown by different techniques, in particular by physical vapor deposition (PVD) such as pulsed laser ablation deposition (commonly referred to by the English acronym PLD, "Pulsed Laser Deposition"), cathode sputtering deposition (reactive or not) or molecular beam epitaxy (MBE, "Molecular beam epitaxy") or even by chemical vapor deposition (CVD, "Chemical Vapor Deposition") such as organometallic vapor phase epitaxy (MOCVD, "Metal Organic Chemical Vapor Deposition").

[0093] The buffer layer 200 has a lower face 202 facing, and preferably in contact with, the upper face 101 of the primer layer 100, and an upper face 201 opposite its lower face 202. In the case of a primer layer 100 consisting of disjointed germination substrates 100a, 100b, 100c, the buffer layer 200 is made up of islands 200a, 200b, 200c which are also disjointed.

[0094] In order not to alter the seed layer 100, the epitaxy of the buffer layer 200 is carried out at a temperature Tgrowthjbuffer lower than the melting temperature Tfusion >1U N>i of the first IILN material. This guarantees the stability of the first IILN material during the growth of the buffer layer 200.

[0095] Furthermore, the thickness e200 of the buffer layer 200 in the vertical direction Z is advantageously chosen so as to avoid the appearance of structural defects within the buffer layer 200 which could then propagate in the seed layer and the layer of interest which will be described later. To do this, it is possible to establish, for any material chosen to act as a buffer material, a critical thickness ecritique from which the dislocation density p within the buffer layer 200 is greater than a given threshold piimite when the buffer layer 200 is grown from the seed layer. Typically, piimite < 1010 cm2. For example, piimite = 108 cm2, preferably piimite = 107 cm2. To guarantee a dislocation density lower than piimite, a thickness e20o lower than ecritique will therefore be set for the buffer layer 200. These precautions make it possible to guarantee excellent crystalline quality of the layer of interest 400.For example, if the difference in lattice parameter between the substrate and the buffer layer is of the order of 1%, critical is of the order of 10 nm, if on the other hand it is of the order of 0.1%, critical is of the order of 100 nm.

[0096] [Fig. 1D] then illustrates the epitaxial growth of a seed layer 300 from the upper face 201 of the buffer layer 200. Advantageously, to withstand the conditions imposed by the growth techniques of the seed layer 300, the buffer layer 200 is refractory and inert. In particular, it preferably has a melting temperature greater than 2000°C. It preferably has a chemical composition that is stable in air. Thus, a material that is stable under the working atmosphere and at the working temperature (typically around 1000°C) will preferably be chosen for the buffer layer. The material constituting the buffer layer may be modified to a minor extent during the process, i.e. its stoichiometry may change. However, it must not decompose during the epitaxy of the seed layer.

[0097] The seed layer 300 is based on a second IILN material preferably being identical to the first IILN material. The seed layer 300 is also crystalline. The first IILN material of the crystalline seed layer 100 is chosen, in particular in terms of lattice parameters, so that the seed layer 300 is also crystalline, without there being a need for the materials of the seed layer and the seed layer to be identical. If the first IILN material and the second IILN material are identical, the precautions taken with regard to the feasibility of epitaxy of the buffer layer 200 from the seed layer 100 (similarity of the crystalline parameters) guarantee the feasibility of epitaxy of the seed layer 300 from the buffer layer 200.If the first IILN material and the second IILN material are distinct, similar precautions can be taken regarding the similarity of the crystal parameters of the buffer material and the second IILN material.

[0098] The seed layer 300 has a lower face 302 facing, and preferably in contact with, the upper face 201 of the buffer layer 200, and an upper face 301 opposite its lower face 302. In the case where the primer layer 100 and the buffer layer 200 are made up of disjoint islands 100a, 100b, 100c, 200a, 200b, 200c, the seed layer 300 is also made up of disjoint islands 300a, 300b, 300c.

[0099] As illustrated in [Fig.1E], the seed layer 300 is then bonded at its upper face 301 to the upper face 2001 of a receiving substrate 2000. The assembly consisting of the support substrate 1000, the primer layer 100, the buffer layer 200 and the seed layer 300 may have been previously turned over in order to carry out this bonding. However, it is not excluded to bond the receiving substrate 2000 to the seed layer 300 and then only turn over the assembly.

[0100] The concept of bonding here covers all types of transfers with adhesion. This term is not limited to the use of glue or an adhesion layer. It may, for example, be a direct bonding, usually referred to as "direct bonding" in English, or a surface-activated bonding, referred to as "surface activated bonding" (SAB) in English.

[0101] The buffer layer 200 is then subjected to laser radiation ([Fig.lF]). This step can be referred to as a “laser lift-off” step, an expression that can be translated into French as “laser detachment”. The purpose of exposing the buffer layer 200 to this laser radiation is to alter it in order to cause the separation of the primer layer 100 and the seed layer 300. More precisely, the energy of the laser radiation is absorbed by the buffer layer 200, which has the effect of heating it. Under the effect of this heating, the buffer layer 200 decomposes at least partially into liquid and gaseous forms. The gaseous part dissipates while at least a part of the liquid part can remain in contact with the upper face 101 of the primer layer 100 and / or the lower face 302 of the seed layer 300.Potential residues, in particular liquid, from the buffer layer 200 may be cleaned during a subsequent step not shown in the figures. For example, this cleaning may include at least one of: chemical cleaning such as soaking in a cleaning solution or chemical cleaning by spray, polishing, typically chemical mechanical polishing (CMP), dry etching. This makes it possible to completely remove the residues of the buffer layer 200 on the surface of the primer 100 and seed 300 layers, without altering the surface or the structure of the latter. It is also possible to remove a very thin portion of the seed layer 300 from its lower face 302, for example a few tens of nanometers. This portion may indeed have been damaged by the . previous steps; removing it ensures good quality of the 300 seed layer.

[0102] The laser radiation has a main wavelength Xiaser. In practice, the laser radiation does not correspond to a single wavelength but is spread over an emission range A^. Typically, the emission range Aa is between 0.8 * Xiaser and 1.2 * Xiaser. Preferably Aa is between 0.9 * Xiaser and 1.1 * Xiaser. Xiaser is defined by the equipment that generates the laser radiation. For the same equipment, Xiaser can take several values. The user selects one of these values. Typically, the main wavelength Xiaser is in the infrared (from approximately 780 nm to approximately 1 mm) and / or the visible range (from approximately 380 nm to approximately 780 nm). In order to ensure the separation of the seed layer 100 and the seed layer 300, as explained previously, the buffer layer 200 must absorb the laser radiation.The emission range Aa can therefore be defined in particular by the fact that the buffer layer 200 has an absorption greater than or equal to 5%, preferably greater than or equal to 20%.

[0103] As illustrated in [Fig. 1F], the laser radiation typically passes through the seed layer 100 before reaching the buffer layer 200. In order to limit the degradation of the seed layer 100, an emission range A, is preferably chosen in which the seed layer 100 has an absorption less than or equal to 50%, preferably less than or equal to 20%, preferably less than or equal to 5%. In this way, the quality of the seed layer 100 is little impacted or even not impacted by the passage of the laser radiation, and it can be reused for a new implementation of the method according to the invention.

[0104] The laser radiation also most often passes through a support substrate 1000 as described above. An emission range Aa will then also preferably be chosen in which the support substrate 1000 has an absorption less than or equal to 90%, preferably less than or equal to 50%.

[0105] As illustrated in [Fig.lG], after subjecting the buffer layer 200 to laser radiation, a stack is obtained on one side comprising the support substrate 1000 and the seed layer 100, and on the other side a stack comprising the receiving substrate 2000 and the seed layer 300.

[0106] [Fig. 1H] illustrates a step of growth by epitaxy from the lower face 302 of the seed layer 300 of a layer of interest 400. This layer of interest 400 typically has a dimension in the vertical direction greater than its dimensions in projection in the horizontal plane XY in order to be used in so-called vertical devices, for example for power components. In the example illustrated in [Fig. 1H], each island 300a, 300b, 300c of the seed layer 300 serves as a basis for the epitaxy of an island 400a, 400b, 400c of the layer of interest 400, the islands 400a, 400b, 400c of the layer of interest 400 being disjoint.

[0107] [Fig. II] illustrates a variant of this same growth step of the layer of interest 400. In this embodiment, the islands 400a, 400b, 400c of the layer of interest 400 grown from the islands 300a, 300b, 300c of the seed layer are coalesced so that the layer of interest 400 forms a single continuous layer.

[0108] It should be noted that the method may comprise, depending on the intended applications, one or more steps of doping the seed layer 300 and / or the layer of interest 400. The method may also comprise the deposition of one or more layers, possibly doped, on the upper face 301 of the seed layer 300 before its transfer to the receiving substrate 2000. These various optional steps may in particular allow the production of electronic or microelectronic devices. A P / N junction may in particular be formed within the seed layer 300 and the layer of interest 400.

[0109] The receiving substrate 2000 is advantageously based on a material having a coefficient of thermal expansion (CTE) close to that of the second IILN material and, if the third IILN material is distinct from the second IILN material, close to that of the third IILN material. Advantageously, the difference between the CTEs does not exceed 20%. In this way, the appearance of cracks due to differences in CTE is limited, in particular during the temperature reduction of the assembly after the epitaxy of the layer of interest 400. With this in mind, and in particular when the second IILN material and / or the third IILN material is GaN, a poly-AIN, poly-SiC or even silicon receiving substrate 2000 is advantageously chosen. The receiving substrate 2000 may, for example, be a silicon substrate integrating an integrated circuit such as a CMOS circuit. This is particularly the case when the seed layer 300 and / or the layer of interest 400 are doped during the process.

[0110] Furthermore, it is noted that the stack comprising the support substrate 1000 and the primer layer 100 ([Fig. 1G]) is identical to that shown in [Fig. 1B]. It can be reused for a new implementation of the method just described with reference to FIGS. 1B to 1F. The same donor substrate 1000 and the same primer layer 100 can be used around ten times or more for the growth of an IILN material. Thus, the present method allows a rationalization of manufacturing costs and is much less expensive than certain current techniques for growing IILN materials.

[0111] A second embodiment of the method according to the invention will now be described with reference to FIGS. 2A to 2D. It repeats in all respects the steps of the first embodiment but differs in that the primer layer 100 forms a single continuous layer, as illustrated in [Fig. 2A], and not disjointed islands. In this embodiment, the primer layer 100 can, as shown in figures, rest on the support substrate 1000. According to another example, the seed layer 100 is a substrate based on the first III-N material, typically a GaN substrate. Thus, the seed layer 100 itself forms the support substrate. In this case, it typically does not rest on a separate support substrate 1000 (case not shown). In this embodiment, the seed layer 100 may for example have a diameter of 200 or 300 mm.

[0112] As illustrated in [Fig.2B], the epitaxial growth of the buffer material and then of the second III-N material makes it possible to form a buffer layer and a seed layer, each forming a single continuous layer. [Fig.2C] illustrates the step of bonding the upper face 301 of the seed layer 300 to the receiving substrate 2000 and the step of subjecting the buffer layer 200 to laser radiation. The characteristics of these steps presented in the context of the first embodiment apply mutatis mutandis to this second embodiment.

[0113] The choice of a continuous or discontinuous primer layer 100 (and, consequently, of a continuous or discontinuous buffer layer 200 and seed layer 300) depends on the intended applications and in particular on the dimensions of the devices in which the layer of interest 400 will be used.

[0114] According to an example illustrated in Figures 3A to 3C, an adhesion layer 500 is deposited around the primer layer 100, the buffer layer 200 and the seed layer 300. This example is advantageously implemented in combination with the first embodiment because it allows good mechanical support between the different islands. As illustrated in [Fig.3A], the adhesion layer 500 is preferably in contact with the upper face 1001 of the support substrate 1000, the flank 103A, 103B, 103C of each of the islands 100a, 100b, 100c of the primer layer 100, the flank 203a, 203b, 203c of each of the islands 200a, 200b, 200c of the buffer layer 200 and the flank 303a, 303b, 303c of each of the islands 300a, 300b, 300c of the seed layer 300. It may also extend above, and preferably in contact with, the upper face 301 of the seed layer 300.In the latter case, it is the upper face 501 of the adhesion layer 501 which is directly in contact with the upper face 2001 of the receiving substrate 2000.

[0115] The step of exposing the buffer layer 200 to laser radiation ([Fig.3B]) is carried out in the same way as in the previous embodiments. The separation of the primer layer 100 and the seed layer 300 due to the alteration of the buffer layer 200 causes a weakening of the adhesion layer 500 level with the altered buffer layer 200. The adhesion layer 500 itself ends up being split into two parts, one surrounding the seed layer 300 and the other surrounding the primer layer 100. As illustrated in [Fig.3C], we obtain, as in [Fig.1G], a reusable assembly for other iterations of the method according to the invention, and an assembly making it possible to grow a third III-N material from the seed layer 300.

[0116] According to an advantageous example illustrated in FIGS. 4A to 4C, before bonding the seed layer 300 to the receiving substrate 2000, upper electrical contacts 610 are deposited on the upper face 301 of the seed layer 300 and / or lower electrical contacts 620 are deposited on the upper face 2001 of the receiving substrate 2000. FIGS. 4A to 4C illustrate the case where contacts have been deposited both on the upper face 301 of the seed layer 300 and on the upper face 2001 of the substrate, and are then brought into contact during bonding ([Fig.4B]). The assembly formed by an upper electrical contact 610 and a lower electrical contact 620 will then be designated electrical contact 600.It is however perfectly conceivable that only upper electrical contacts 610 are deposited and that they are brought into contact with the upper face 2001 of the receiving substrate 2000 during the bonding step, or that only lower electrical contacts 620 are deposited and that they are brought into contact with the upper face 301 of the seed layer 300 during the bonding step. The electrical continuity between the seed layer 300 and the receiving substrate 2000 will be ensured in each of these cases.

[0117] In the case where contacts 620 are deposited on the upper face 2001 of the receiving substrate 2000, the presence of a secondary adhesion layer 550 can be provided to ensure effective bonding of the two stacks shown in [Fig.4A] as well as good mechanical maintenance of the assembly.

[0118] In this example, the bonding of the seed layer 300 to the receiving substrate 200 is done indirectly, more precisely at the level of the electrical contacts and possibly the upper face 501 of the adhesion layer 500.

[0119] It is also possible to provide the presence, between the receiving substrate 2000 and the electrical contacts 610, 620, of a layer 700 comprising electrical contacts structured and configured to meet the electrical power supply needs of the devices which will be manufactured using the method.

[0120] Figures 4D to 4G illustrate an example of successions of steps that can be carried out from the assembly comprising the seed layer obtained in [Fig.4C]. As in the previously described embodiments, one step consists of growing by epitaxy from the lower face 302 of the seed layer a layer of interest 400. Secondary electrical contacts 650 can then be formed on the upper face 401 of the layer of interest 400. A new adhesion layer can be deposited on the adhesion layer 500, around the layer of interest 400 and the secondary electrical contacts 650, in order to finish encapsulating the assembly. The seed layer 300 and the layer of interest 400 can then be cut into the desired shapes and the receiving substrate 2000 can be removed ([Fig.4G]), thereby forming a plurality of devices, each formed by an electrical contact 600, a portion of the seed layer 300, a portion of the layer of interest 400, and a secondary electrical contact 650.

[0121] According to another embodiment illustrated in FIGS. 5A to 5E, after the step of growing the layer of interest 400 ([Fig.5B]), the layer of interest 400 is transferred onto a secondary receiving substrate 3000 and the receiving substrate 2000 is removed (transition from [Fig.5C] to [Fig.5D]).

[0122] More specifically, according to this particular embodiment, the seed layer 300 was bonded directly to the receiving substrate 2000 ([Fig.5A]) before the alteration of the buffer layer 200. As illustrated in [Fig.5B], in the same way as in the previous embodiments, the layer of interest 400 is grown from the seed layer 300. Secondary electrical contacts 650 are then formed on the upper face 401 of the layer of interest 400, and the assembly is then transferred to the secondary receiving substrate 3000 at the level of the secondary electrical contacts 650 ([Fig.5C]). The secondary receiving substrate 3000 may for example be a silicon substrate. Preferably, an additional adhesion layer is deposited on the adhesion layer 500 in order to surround the layer of interest 400 and the secondary electrical contacts 650.After removal of the receiving substrate 2000, electrical contacts 600 are formed on the underside 302 of the seed layer 300. The seed layer 300 and the layer of interest 400 can then be cut into the desired shapes and the secondary receiving substrate 3000 can be removed ([Fig.5E]), thereby forming a plurality of devices, each formed by an electrical contact 600, a portion of the seed layer 300, a portion of the layer of interest 400, and a secondary electrical contact 650.

[0123] According to another embodiment, the receiving substrate 2000 is electrically conductive. It may for example be a poly-SiC substrate, possibly doped. Advantageously, electrical contacts 2010 have been formed, typically by etching, from the upper face 2001 of the receiving substrate 2000. It is also conceivable to form electrical contacts on the upper face 2001 of the receiving substrate 2000. Figures 6B to 6F illustrate steps similar to those described with reference to Figures 4B to 4F, the only difference being that the receiving substrate 2000 is this time electrically conductive. Furthermore, advantageously, as illustrated in [Fig.6G], the receiving substrate 2000 is also cut during the step of cutting the devices and is not removed.

[0124] The use of a conductive substrate makes it easier to inject current into the device, which is particularly interesting in the case of a vertical device. It is possible to keep the receiving substrate 2000 in the final device, as illustrated in [Fig.6G], which saves the step of removing it.

[0125] It should be noted that the presence of the buffer layer 200 may have the effect of causing growth of the seed layer 300 under nitrogen polarity. This effect is for example obtained when the buffer material is HfN or TiN.

[0126] The polarization of the seed layer 300 during its formation by epitaxy is a very great advantage because it gives it a polarization orientation allowing new growth under nitrogen polarity by epitaxy after a single transfer. More precisely, the layer of interest 400 can be grown from the lower face 302 of the seed layer 300 and itself be metal polarized. In most current approaches, it is necessary to carry out a double transfer of the layer that it is desired to polarize in order to orient the polarization in the required direction, but such a double transfer has a high cost. The method therefore makes it possible to reduce the production costs of a polarized IILN layer.

[0127] Through the different embodiments described above, it clearly appears that the invention proposes a solution making it possible to obtain layers of IILN materials of good quality while limiting manufacturing costs, thus meeting the current needs of the industry.

[0128] As stated in the introduction, one of the main pitfalls of growing GaN on silicon lies in the very significant difference in CTE between GaN and silicon. Because of this difference, the thickness of GaN that can be grown on silicon is limited because the thicker the layers, the more fragile the plates become. It is therefore difficult to obtain layers, particularly thick layers, of good quality. By growing the layer of interest 400 on a receiving substrate 2000 having a suitable CTE, such as a poly-AIN substrate, layers with a low dislocation rate can be obtained. By implementing the present invention, these good quality layers can be transferred onto a silicon plate, and then the active layers of a light-emitting diode (LED) can be epitaxied, for example (quantum wells, pGaN), without going through thick layer epitaxy on silicon.These layers will then be available for the fabrication of micro-display devices or other components including LEDs.

[0129] The invention is not limited to the embodiments previously described and extends to all embodiments covered by the invention.

Claims

1. Claims A method of growing a III-N material, such as GaN, comprising the following steps: • Providing a primer layer (100) based on a first III-N material, the primer layer (100) having an upper face (101) extending mainly in a horizontal plane (XY) defined by a first direction (X) and a second direction (Y), • Growing a buffer layer (200) by epitaxy of a buffer material from the upper face (101) of the primer layer (100), the buffer layer (200) having an upper face (201) and a lower face (202) opposite each other, the lower face (202) of the buffer layer (200) being opposite, and preferably in contact with, the upper face (101) of the primer layer (100), • Growing by epitaxy a seed layer (300), in a second III-N material, preferably identical to the first III-N material, from the upper face (201) of the buffer layer (200), the seed layer (300) having an upper face (301) and a lower face (302) opposite each other, the lower face (302) of the seed layer (300) being opposite, and preferably in contact with, the upper face (201) of the buffer layer (200), • Glue the upper face (301) of the germination layer (300) onto a receiving substrate (2000), • Subjecting the buffer layer (200) to laser radiation having a main wavelength Xiaser, the seed layer (100) being configured to transmit at least part of the laser radiation, the seed layer (100) preferably having an absorption less than or equal to 10% for the main wavelength Xiaser, the laser radiation being configured to alter the buffer layer (200) so as to separate the seed layer (100) and the seed layer (300), • Growing by epitaxy of a third III-N material, preferably identical to the second III-N material, a layer of interest (400) from the lower face (302) of the seed layer (300).

2. Method according to the preceding claim in which, in the horizontal plane (XY), the first III-N material has a first mesh length alu.N,i and the buffer material has a buffer mesh length a,ampOn, with | (am \4- a,ampOn ) | / am \4< 0.

02.

3. Method according to any one of the preceding claims in which, in the horizontal plane (XY), the mesh of the first III-N material has a first angle alu.N,i and the mesh of the buffer material has an angle cttampOn, with | cttampOn — •

4. Method according to any one of the preceding claims in which the buffer layer (200) has, in a third direction (Z) perpendicular to the horizontal plane (XY), a thickness e2Oo configured so that the buffer layer (200) has a dislocation density p of less than 1010 cm2, preferably less than 108 cm2, preferably less than 107 cm2.

5. Method according to the preceding claim in which e200<100 nm, preferably e200<20 nm.

6. Method according to any one of the preceding claims in which the first III-N material has a melting temperature T melting,ni n,i, the epitaxy of the buffer material being carried out at a Growth temperature T growth, buffer, with T growth, buffer^'Tmelting, III-N, 1 •

7. Method according to any one of the preceding claims in which, in a third direction (Z) perpendicular to the horizontal plane (XY), the primer layer (100) has a thickness e»» and the layer of interest (400) has a thickness e400, with e4oo>2*eioo, preferably e4oo>3*eioo, preferably e4oo>5*eioo.

8. Method according to any one of the preceding claims in which, in a third direction (Z) perpendicular to the horizontal plane (XY), the buffer layer (200) has a thickness e200 and the layer of interest (400) has a thickness e400, with e4oo>2*e2oo, preferably e4oo>3*e2oo, preferably e4Oo>5*e2OO.

9. A method according to any preceding claim wherein the laser radiation is configured to pass through the seed layer (100) before altering the buffer layer (200).

10. A method according to any preceding claim in which, after the separation of the seed layer (100) and the seed layer (300), at least a portion of the buffer layer (200) remains on the seed layer (300), and further comprising, after the step of subjecting the buffer layer (200) to laser radiation and before the step of growing the layer of interest (400), a step of removing said at least portion of the buffer layer (200).

11. A method according to any preceding claim wherein the primer layer (100) forms a single continuous layer.

12. A method according to any one of claims 1 to 9 wherein the seed layer (100) forms disjointed seed substrates (100a, 100b, 100c).

13. Method according to the preceding claim in which each of the disjointed germination substrates (100a, 100b, 100c) has, in projection in the horizontal plane (XY), a maximum dimension less than 150 mm, preferably less than 100 mm.

14. Method according to any one of the two preceding claims in which, during the step of providing the primer layer (100), the germination substrates (100a, 100b, 100c) are transferred onto a support substrate (1000).

15. Method according to any one of the preceding claims further comprising, after the step of epitaxy of the second III-N material and before the step of bonding the upper face (301) of the seed layer (300) on the receiving substrate (2000), the formation of an adhesion layer (500) on the upper face (301) of the seed layer (300).

16. A method according to any preceding claim wherein the III-N material is based on at least one of GaN, PAIN, InN and any combination of these materials.

17. A method according to any preceding claim wherein the buffer material is one of the following materials: HfN, ZrN, NbN, TiN.

18. A method according to any preceding claim further comprising forming electrical contacts (610) on the upper face (301) of the seed layer (300) and secondary electrical contacts (650) on the upper face (401) of the layer of interest (400).

19. A method according to any preceding claim wherein the receiving substrate (2000) is electrically conductive.

20. A method according to any preceding claim wherein the receiving substrate (2000) comprises at least one integrated electronic circuit.

21. A method according to any preceding claim wherein the buffer material is chosen such that the buffer layer (200) induces nitrogen polarity within the seed layer (300).

Images