Method for detaching an epitaxial layer from its growth substrate by laser pre-ablation of the sacrificial buffer layer

Abstract

Method for detaching a semiconductor material from a substrate via a hybrid detachment approach. The invention describes a hybrid method for detaching epitaxial layers from a growth substrate coated with a sacrificial buffer layer. This method combines initial laser ablation, which weakens the bond between the buffer layer and the substrate, and preferential chemical attack (ELO) to dissolve the buffer layer. This approach accelerates detachment without damaging the back side of the epitaxial layers, allowing for intact transfer. The process is applicable to epitaxial overlays or stacks of semiconductor materials used in electronic components. These materials, once detached and transferred onto a conductive substrate, find applications in power electronics, particularly for devices such as light-emitting diodes (LEDs) and transistors.This innovative method ensures increased efficiency and preserves the integrity of semiconductor materials, meeting the requirements of high power density technologies. Figure for abbreviation FIG.8.

Description

Title of the invention: Method for detaching an epitaxial layer from its growth substrate by laser pre-ablation of the sacrificial buffer layer. TECHNICAL FIELD OF THE INVENTION

[0001] The field of the invention relates to materials in the form of epitaxial crystalline layers and their transfer from a growth substrate to a host substrate with properties and / or a cost more suitable for a given application than the growth substrate. This is the case for active layers in various electronic and optoelectronic devices designed for high power density per unit area (such as power electronics, transparent electronics, piezoelectrics, LEDs, and lasers).

[0002] Indeed, the growth substrate used in epitaxial deposition typically consists of a bulk single crystal chosen to favor a crystalline structure and orientation that is reproduced by the epitaxially deposited films on it. Often the choice of growth substrate is very limited and the substrate cannot offer all the properties that are desirable for the final application (for example, flexibility, transparency, conductivity (electrical and / or thermal), etc.).

[0003] For example, in power electronics (the application of semiconductor electronics for the control, switching, and conversion of electrical energy to high voltage or power), components (e.g., the field-effect transistor, the pn junction, or the Schottky Barrier Diode (SBD)) suffer from self-heating due to the relatively low thermal conductivity of the growth substrates typically used (e.g., Ga2O3 on sapphire). The detrimental effects of this low thermal conductivity on the performance and lifespan of these components are now recognized, and solutions must be found to address this problem.

[0004] Currently, power electronics suffer from relatively low energy efficiency and a significant limitation in managing the maximum power / voltage used compared to the intrinsically available power / voltage. Improving these two aspects is currently considered a technological hurdle on the path to efficient, low-carbon electrical energy transmission and conversion.

[0005] Silicon (Si) is currently the most widespread electronic material in power electronics. However, silicon electronics faces a challenge in power electronics applications: device performance Silicon-based materials degrade significantly at high temperatures, limiting their use in high-power, high-density (i.e., per unit area) electronics. Furthermore, because silicon has a narrow, indirect band gap, it exhibits a relatively low breakdown electric field. This means that its high-power / high-voltage handling capabilities are limited. As an alternative to silicon, a new generation of wide-bandgap semiconductors is emerging that can operate at higher voltages, temperatures, and switching frequencies, with efficiencies exceeding those of silicon devices. Their properties not only result in lower losses but also enable significantly smaller devices due to reduced cooling requirements, contributing to a lower overall system cost.

[0006] For the full realization of the potential of these components, however, it is necessary that the growth of wide bandgap semiconductor materials can be carried out on substrates capable of promoting the removal of heat generated by dissipated losses.

[0007] Indeed, components based on wide bandgap semiconductor materials epitaxially grown on insulating growth substrates are unsuitable for use in power electronics without a cooling device because of the heat that accumulates and degrades the component's performance. Therefore, it is desirable to transfer the semiconductor layers directly onto heat sinks, but currently available growth substrates do not offer the possibility of achieving both good epitaxial quality and good thermal conductivity. For this reason, it is necessary to transfer the active epitaxial layers from their insulating growth substrates to host substrates that have higher thermal conductivity. PREVIOUS STATE OF THE ART

[0008] Currently, several techniques exist for removing an epitaxial layer from its growth substrate (ref - Figure). All these techniques have their advantages and disadvantages.

[0009] Prior art documents show how to separate an epitaxial layer of GaN from its insulating growth substrate (sapphire) by laser ablation through the substrate (a process known as "Laser Lift-Off" (LLO)). A significant problem with this lifting method is that the resulting GaN back face is damaged and rough, with remnants / particles of metallic gallium on the surface. The properties of GaN in the damaged region are degraded, making it difficult (mechanically) to bond this rough face to a host substrate. a smooth surface. This hinders optical, thermal, and electronic coupling with the host substrate.

[0010] Another proposed solution is ELO (preferential chemical etching of a sacrificial sublayer) – see [Fig. 6]. The process consists of chemically dissolving an epitaxial sacrificial sublayer (buffer) on the sapphire substrate, onto which an epitaxial overlayer is deposited. The chemical lift-off completely dissolves the sacrificial buffer layer, and thus the epitaxial layer separates from the growth substrate. Unlike a laser lift-off, the back face of the epitaxial overlayer is not damaged, resulting in a perfect, smooth, and defect-free surface that can be transferred to a conductive substrate with good mechanical, optical, electrical, and thermal coupling.

[0011] Two significant problems with the ELO process: i) the chemical etching gradually slows down, and ii) the time required for chemical etching increases significantly as the wafer surface area increases. Indeed, the ELO peeling process can take several hours, or even several days. However, such a process speed is not well suited for industrial manufacturing.

[0012] It is therefore desirable to have a peeling process which can, at the same time, not damage the back face of the epitaxial layer and be fast enough for industrial use or manufacture.

[0013] The object of the present invention is to achieve a rapid detachment of an epitaxial material from its growth substrate, such that the surface of the detached material is not damaged and the cost of the detachment process is not high so that it can be suitable for industrial manufacturing. Description of the invention

[0014] The invention consists of a hybrid method combining laser lode lysis (LLO) and electrophoretic lode lysis (ELO) for accelerating delamination. To this end, the invention proposes a method for delaminating an epitaxial layer or stack of epitaxial layers from a growth substrate covered with a sacrificial buffer layer, this sacrificial layer being located between said growth substrate and the epitaxial layer. The method is characterized in that it comprises a first step of laser ablation of the base of the sacrificial buffer layer, followed by a second step of preferential chemical attack of said sacrificial buffer layer (referred to as ELO), such that the laser ablation weakens the bond between the sacrificial sublayer and the growth substrate before the chemical attack, thereby accelerating the subsequent dissolution of the sublayer and the delamination of the epitaxial layer(s) from the growth substrate without damaging the back face of the epitaxial layer.

[0015] Furthermore, in the above peeling process, the growth substrate is sapphire and the embrittlement of the base of the sacrificial layer by laser ablation is done through the back face, the non-deposited face, of the substrate.

[0016] Furthermore, depending on the nature of the sacrificial buffer layer, laser ablation can be achieved by accumulating laser shots covering the entire surface of the sapphire substrate, or by scanning a laser beam optically shaped to obtain the appropriate and homogeneous geometry required for laser ablation.

[0017] Furthermore, after optical shaping of a laser beam, laser ablation is performed at any angle of incidence between 0 and 90°, perpendicular to the substrate surface up to the transverse plane of the substrate on a slice, focusing the energy on the sacrificial buffer sublayer so as to weaken its base. In addition, laser ablation can be combined with circular and / or transverse movement of the substrate for greater excitation homogeneity on large substrates.

[0018] Preferably, the epitaxial overlayer is of GaN or SiC or Ga2O3 #a##[3##Y#ô#on#K#, and the sacrificial buffer sublayer can be of ZnO, or CrN or SiO2, the thickness of said sacrificial buffer sublayer is between 0.5 nm and 100 micron.

[0019] The invention also includes a semiconductor component comprising an epitaxial layer of semiconductor material such as GaN or Ga2O3 or SiC with a smooth surface detached from its sapphire substrate according to the above process. The detached layer of semiconductor material is transferred onto a conductive host substrate. Said host substrate is such that it has the highest possible heat dissipation factor and possesses optical, electrical, and thermal characteristics suitable for power electronics devices, such as SiC, diamond, or a metal.

[0020] The invention further relates to the use of the above semiconductor component in power electronics or in devices such as light-emitting diodes or transistors. BRIEF DESCRIPTION OF THE FIGURES

[0021] [Fig. O] shows a map of accumulation prospects on the main comparative advantages and disadvantages of the main semiconductors used as the basis for power electronics; [Fig. A] a graph contrasting the properties of key materials [3-Ga2O3] with those of current power electronics materials; [Fig. B] illustration of a flipped chip configuration used to connect a Ga2O3 power device to a heat sink

[191] ; [Fig. C] a map of technologies considered to be technically superior in the power / frequency space

[139] ; [Fig.D] a comparison of projected costs for manufacturing electrical devices based on sic and [3-Ga2O3

[192] ;

[0022] [Fig-1] represents the laser lift-off (LLO) process of a layer epitaxially grown from Ga2O3 on a sapphire growth substrate;

[0023] [Fig.2] illustrates the surface of the back face of the Ga2O3 layer obtained after the LLO takeoff;

[0024] [Fig.3] illustrates a layer of Ga2O3 epitaxially on a sapphire substrate covered with a sacrificial buffer layer (ZnO);

[0025] [Fig.4] illustrates the chemical lift-off (ELO) process of a layer epitaxially of Ga2O3 from a sapphire substrate covered with a sacrificial buffer layer (ZnO);

[0026] [Fig. 5A and Fig. 5B] present an axial section illustrating the continuity of the interface and the smooth surface of the Ga2O3 epitaxial overlayer and the ZnO underlayer before takeoff;

[0027] [Fig.6] illustrates the process of progressive dissolution by ELO of the sacrificial buffer sublayer in ZnO;

[0028] [Fig.7] illustrates a comparative table of the characteristics of each of the approaches of takeoff above (laser takeoff (LLO) and chemical takeoff (ELO));

[0029] [Fig.8] illustrates the hybrid peeling process of an epitaxially coated Ga2O3 overlayer of the sapphire substrate covered with a sacrificial buffer layer (ZnO) according to the present invention;

[0030] [Fig.9] illustrates the technique of peeling an active layer from its substrate by Transverse laser ablation on the sacrificial sublayer. DETAILED DESCRIPTION

[0031] Fig. O shows a mapping of a current perspective on the main comparative advantages and disadvantages of the main semiconductors used as the basis of power electronics.

[0032] Figure A is a graph contrasting the key properties of [3-Ga2O3] materials with those of current power electronics materials. This graph clearly shows that although Ga2O3 dominates in terms of band gap energy and decay field, it exhibits very significant drawbacks in terms of electron mobility and thermal conductivity.

[0033] The second problem is that thermal management poses a major challenge for Ga2O3, since the thermal conductivity of the bulk substrate is an order of magnitude lower than that of sapphire itself. Consequently, heat accumulates during high-voltage operation leads to performance drift and accelerated aging. Several potential solutions have been proposed, such as direct epitaxial growth on a substrate with higher thermal conductivity (such as SiC or diamond) or the integration of Ga2O3 with a heat sink either by a flip chip approach (see below), or by thinning the Ga2O3 substrate and wafer, bonded to a heat sink [184 - 191].

[0034] Figure B illustrates a flipped chip configuration used to connect a Ga2O3 power device to a heat sink

[191] . Assuming that these mobility and thermal management issues can be resolved, Mastro et al.

[139] predict that [3-Ga2O3-based power devices could find a competitive advantage in the higher power and lower frequency space of the global market:

[0035] Figure C is a mapping of technologies considered to be technically superior in the power / frequency space

[139] . This will nevertheless depend on the ability of [3-Ga2O3] to compete in terms of cost with SiC. Recent market studies suggest that [3-Ga2O3] may well offer a significant manufacturing cost advantage over SiC in the future

[192] (SiC power devices are currently cheaper than GaN-based power devices).

[0036] Figure D is a comparison of the projected costs for manufacturing SiC-based and [3-Ga2O3]-based electrical devices

[192] . This is largely based on the fact that single-crystal [3-Ga2O3] substrates can be manufactured in a relatively large format by inexpensive melt growth processes [136-138]. However, at present, [3-Ga2O3] substrates cost almost two orders of magnitude more than equivalent sapphire substrates and are not yet available in production volumes (the current manufacturing base is very small because it depends exclusively on a limited R&D market).

[0037] There is therefore significant interest and work devoted to the epitaxy of [3-Ga2O3] on sapphire substrates, and, as was the case in the development of GaN LEDs, the improvement in the epitaxial quality of [3-Ga2O3] on sapphire has exceeded expectations. It has been understood that sapphire offers several advantages (other than cost and industrial availability in large formats / volumes) that could allow it to find a place in future [3-Ga2O3] power electronics product portfolios. First, sapphire has an order of magnitude better thermal conductivity than [3-Ga2O3] substrates, which allows for better intrinsic heat management. Second, the epitaxial shift with sapphire allows for deformation engineering that can improve doping potential, stabilize different phases / poly-types / orientations / out of equilibrium and extend the range of band gap engineering.

[0038] Furthermore, [3-Ga2O3] can be easily transferred from sapphire to another substrate by means of laser detachment. This approach uses a laser beam fired through the sapphire substrate to enable the Ga2O3 / sapphire interface and thus release the substrate epilayer (impossible with Ga2O3 substrates).Such an approach is widely used for the fabrication of high-brightness GaN LEDs and brings several advantages, including better heat dissipation (the new host substrate can be a heat sink) and the possibility of back-contacting the [3-Ga2O3], which could allow vertical current flow in the device and thus facilitate higher currents (due to less current footprint and less local heating compared to side-flow devices), simpler device architectures and smaller device footprints (i.e. more devices per wafer and therefore less current).

[0039] As shown in [Fig. 1], according to the prior art, a layer of Ga2O3 was epitaxially deposited directly onto a sapphire substrate. The laser lift-off (LLO) process was used to detach the Ga2O3 semiconductor material from its sapphire substrate in order to transfer it onto a conductive substrate.

[0040] [Fig.2] represents the rear face of the Ga2O3 layer after LLO peeling of the Sapphire substrate. LLO delamination is relatively rapid, but the back side of such a delaminated Ga2O3 layer has a rough surface contaminated by gallium metal residues. This is unsuitable for the direct transfer / coupling of the semiconductor material to a conductive host substrate.

[0041] [Fig.3] represents an epitaxial layer of Ga2O3 deposited on a sapphire substrate covered with a buffer layer of ZnO. The representative image shows the continuity of the interface and the smooth surfaces of the two materials.

[0042] [Fig.4] illustrates the detachment process of the Ga2O3 layer from the substrate of Sapphire. First, chemical etching progressively debonds the sacrificial underlayer located between the Ga2O3 and the sapphire substrate. Once the ZnO (sacrificial) layer is completely removed, the Ga2O3 overlayer detaches from the substrate. The axial cross-sections in [Fig. 5A] and [Fig. 5B] illustrate the continuity of the interface and the smooth surface of the Ga2O3 and ZnO material before debonding ([Fig. 5A]) and after the active Ga2O3 layer is transferred to a host substrate ([Fig. 5B]). The advantage of ELO, or chemical etching, is that it is performed at room temperature and the back surface of the debonded layer is very smooth, allowing it to adhere more easily to a host substrate with good mechanical, optical, electrical, and thermal coupling, and without any adhesive or bonding material.

[0043] [Fig. 6] illustrates the progressive stages of ELO. It is noted that the ELO process is so slow that it can take several hours, or even several days, but such a process, despite its advantages, is not suitable for mass industrial production.

[0044] The comparative table in [Fig. 7] illustrates the advantages and disadvantages of the two LLO and ELO takeoff approaches. The table shows that the particular advantage of the laser approach is the takeoff speed, but it also shows that this approach has a very high cost and that after takeoff the base of the Ga2O3 layer is damaged / roughened and metallic residues, in gallium, remain on the Ga2O3 semiconductor material layer.

[0045] [Fig.8] represents the process of a hybrid takeoff (pre-ablation followed by ELO) of an epitaxial Ga2O3 overlay on its growth substrate (sapphire) covered with a sacrificial buffer layer (ZnO) according to the present invention. The hybrid process consists of a first step of weakening the base of the buffer layer by laser pre-ablation, followed in a second step by accelerated ELO (chemical etching) debonding. As shown in [Fig. 8], laser exposure is performed through the back face of the growth substrate (the undeposited face). The laser ablation weakens the bond between the sacrificial layer and the epitaxial overlay. Following the laser pre-ablation, a typical ELO ratio is achieved with a subsequent accelerated chemical etch that dissolves the sacrificial buffer layer more rapidly and allows the Ga2O3 layer to be completely debonded from the sapphire substrate in a reduced time.The layer thus detached is not damaged by the laser and does not have gallium metallic residues on its surface (laser ablation being absorbed by the ZnO buffer layer) and thus presents a smooth surface, which is better suited for bonding to a host substrate, i.e. for example a substrate with the highest possible heat dissipation index for the case of power electronics devices.

[0046] In this hybrid process for removing the active layer, the epitaxial overlayer, laser ablation is performed through the back face of the substrate (the undeposited face). This exposure can be achieved (depending on the nature of the epitaxial overlayer to be transferred and / or the sacrificial sublayer) by accumulating laser shots covering the entire surface of the substrate, or by scanning a laser beam optically shaped to obtain the appropriate and homogeneous geometry necessary for laser ablation of the base of the buffer layer. This process facilitates the dissolution of the sacrificial layer without a liquid or gaseous catalyst.

[0047] The hybrid peeling approach of the present invention consists of laser preablation of the base of the sacrificial buffer layer, followed by preferential chemical attack (PCE) of said buffer layer. Indeed, laser attack of the base of the sacrificial buffer sublayer weakens the bond between the substrate The sapphire and buffer layer are combined in this way, allowing subsequent chemical attack to dissolve the sacrificial layer more rapidly and thus separate the active semiconductor material layer from the sapphire substrate. This hybrid peeling approach does not require hours or days to dissolve the sacrificial layer. The backside surface of the layer thus applied is smooth and can adhere more easily to a host substrate. The speed of this peeling process makes it better suited for the industrial manufacturing of semiconductor materials.

[0048] The peeling approach according to the present invention is a combined or hybrid approach; it is faster than conventional ELO and provides a back face of the detached epitaxial layer without the damage and metallic residues characteristic of LLO. With this smooth surface, the peeled epitaxial overlayer(s) is / are better suited for direct bonding to a conductive substrate.

[0049] It should be noted that the hybrid peeling process is not limited to the epitaxial layer of semiconductors such as Ga2O3, GaN, or SiC. It can be applied to peel piezoelectric or other epitaxial layers. The sacrificial layer used in this process is also not limited to ZnO; thus, the sacrificial layer can be CrN or SiO2. The thickness of said sacrificial layer can range from 0.5 nm to 100 microns.

[0050] The invention also relates to semiconductor components comprising a layer of semiconductor material with a smooth surface, such as GaN, SiC, or Ga2O3, detached from their substrate, onto which they have been epitaxially grown and transferred after detachment to a thermally and / or electrically conductive substrate. The substrate onto which the layer(s) of semiconductor material are transferred has the highest possible heat dissipation factor.

[0051] The invention relates to the use of the above semiconductor component in power electronics, high-frequency electronics, piezo electronics and in devices such as light-emitting diodes, lasers or transistors.

[0052] DESCRIPTION OF SOME EMBODIMENT MODES

[0053] We will describe below an example of an implementation method

[0054] of the present invention with all of its steps.

[0055] 1) First step: Deposition of a sacrificial buffer layer of ZnO onto a substrate single crystal sapphire.

[0056] Several techniques can be considered for carrying out this step:

[0057] 1.1: Deposition by evaporation (for example, using molecular beam epitaxy) (Molecular Beam Epitaxy, MBE), a solid Zn source and an O3 flux and / or an O2 RF plasma source are used.

[0058] 1.2: Metal-organic vapor phase epitaxial deposition Chemical Vapor Deposition, MOCVD. This technique can use diethylzinc (DEZn) as organometallic precursors for Zn and isopropanol (z-PrOH) and / or butanol (t-BuOH) for O.

[0059] 1.3: Deposition by PLD, typically with an excimer KrF laser and a solid source ((or The target is a ZnO powder (compressed and sintered) combined with the introduction of molecular oxygen into the deposition chamber. Thanks to the high energy of the plasma, PLD can operate at lower temperatures than conventional MBE, MOCVD, MOVPE, HVPE, or spraying processes.

[0060] 1.4 Organometallic vapor phase epitaxial deposition or “Metal

[0061] Organic Vapor Phase Epitaxy”.

[0062] 1.5 Spray deposition.

[0063] Thus, with such techniques and from a single-crystal substrate maintained at a temperature of a few hundred degrees Celsius during deposition, a sacrificial buffer sublayer of ZnO is first deposited on said single-crystal sapphire substrate.

[0064] 2) Second step: Deposition of an epitaxial overlayer of Ga2O3 onto the sublayer in ZnO: Several techniques can be considered to carry out this step:

[0065] 2.1: Molecular beam epitaxy (MBE) deposition using solid or gaseous sources of elements III and a flow of O2 or O3 and / or a source and / or a source of water vapor and / or an RF plasma source of O2.

[0066] 2.2: Metal-organic vapor phase epitaxial deposition Chemical Vapor Deposition, MOCVD. This technique can use trimethylgallium (TMGa) as an organometallic precursor for Ga. Nitrogen, instead of hydrogen, can be used as the carrier gas (the conventional approach).

[0067] 2.3: Deposition by PLD, typically with a KrF excimer laser and a solid source (or target) of GaO3 powder (compressed and sintered) combined with an O2 source or an O2 RF plasma. Thanks to the high energy of the plasma, PLD allows operation under lower temperatures than conventional MBE or MOCVD or MOVPE or HVPE or spraying processes.

[0068] 2.4 Metal-Organic Vapor Deposition Epitaxial Phase, MOVPE » This technique can use trimethylgallium (TMGa) as organometallic precursors for Ga. O2, O3, H2O and / or an RF plasma O2 source can be used as oxygen sources.

[0069] 2.5 Spray deposition,

[0070] 2.6 Vapor phase epitaxial deposition of a hybrid or "HVPE,

[0071] 3) Release of the GaO3 epitaxial layer from the single-crystal sapphire substrate.

[0072] The base of the ZnO buffer sublayer is first weakened by laser pre-ablation (with an excimer laser, KrF (248nm)) and then the sacrificial buffer layer is completely desoldered by chemical attack with acid (e.g. HCl 0.1M) or alkali (e.g. NaOH 0.1M) without altering the reference substrate or the Ga2O3 epitaxial coating.

[0073] This is accomplished through preferential chemical etching (called Epitaxial Lift-Off or "ELO") of a sacrificial sublayer (buffer layer) interposed between the epitaxial overlayer and the growth substrate. A significant problem with the ELO process is that the rate of chemical etching of the sacrificial sublayer increases (and tends to slow down) as the wafer surface area increases. Indeed, the ELO lift-off process can take several hours, or even several days. Such a process speed is not well suited for industrial manufacturing. More specifically, the invention relates to a method for accelerating the ELO process by laser pre-ablation (through the substrate) of the base of the buffer layer and the use of this method in various technological fields.

Claims

Demands

1. A method for detaching an overlayer or a stack of epitaxial overlayers from a growth substrate covered with a sacrificial buffer layer, this sacrificial layer being located between said growth substrate and the epitaxial layer, characterized in that to detach the epitaxial layer from the growth substrate the method comprises a first step of laser ablation of the base of the sacrificial buffer layer, followed by a second step by a preferential chemical attack of said sacrificial buffer layer (called ELO) such that the laser ablation weakens the bond between the sacrificial sublayer and the growth substrate before the chemical attack, to accelerate the subsequent dissolution of the sublayer and the detachment of the epitaxial overlayer(s) from the growth substrate without damaging the back face of the epitaxial layer.

2. A method for removing an overlayer or a stack of epitaxial overlayers from a growth substrate covered with a sacrificial buffer layer according to claim 1 characterized in that the growth substrate is sapphire and the embrittlement of the base of the sacrificial layer by laser ablation is done through the back face, the undeposited face, of the growth substrate.

3. A method for detaching a layer or stack of epitaxial layers from a growth substrate covered with a sacrificial buffer sublayer according to claim 2, characterized in that, depending on the nature of the sacrificial buffer layer, laser preablation can be carried out by accumulation of laser shots covering the entire surface of the sapphire substrate, or by scanning a laser beam optically shaped to obtain the appropriate and homogeneous geometry necessary for laser ablation.

4. A method for detaching an overlayer or a stack of epitaxial overlayers from a growth substrate covered with a sacrificial buffer underlayer according to claim 3, characterized in that after optical shaping of a laser beam, laser ablation is performed at any angle of incidence between 0° and 90° perpendicular to the surface of the growth substrate up to the transverse of the growth substrate on a slice, focusing the energy on the underlayer. sacrificial buffer layer so as to weaken the base of the sacrificial sub-layer.

5. A method for removing an overlayer or a stack of epitaxial overlayers from a growth substrate covered with a sacrificial buffer layer according to any one of claims 1 to 4, characterized in that laser ablation can be associated with a circular and / or transverse movement of the growth substrate for greater homogeneity of excitation on large substrates.

6. A method for peeling an epitaxial overlayer or stack of overlayers from a growth substrate covered with a sacrificial buffer underlayer according to any one of the preceding claims wherein the epitaxial overlayer is GaN or SiC or Ga2O3, and wherein the sacrificial buffer underlayer is ZnO, CrN or SiO2, the thickness of said sacrificial buffer underlayer being between 0.5 nm and 100 micron.

7. Semiconductor component comprising an epitaxial layer or a stack of epitaxial layers of semiconductor material such as GaN or Ga2O3 or SiC with a smooth surface detached from its sapphire substrate according to the process of claims 1 to 6 characterized in that said detached layer(s) of semiconductor material being transferred onto a conductive host substrate.

8. Semiconductor component according to claim 7 characterized in that said host substrate is such that it has the highest possible heat dissipation factor having optical, electrical and thermal characteristics for power electronics devices such as SiC or a diamond or a metal.

9. Semiconductor component according to claims 7 and 8, of which it is used in power electronics.

10. Semiconductor component according to claim 7 and 8, of which it is used in devices such as light-emitting diodes or transistors.

Images