METHOD FOR PRODUCE AT LEAST ONE PARTIALLY ARCHED OR CURVED SOLID LAYER

Description

[0001] According to claim 1, the present invention relates to a method for producing at least one solid layer that is at least partially curved or bent.

[0002] Semiconductor materials are grown, for example, in large cylinders of crystalline material called ingots, while industrial processes often require wafer material with varying thicknesses and surface qualities. Wafering of brittle semiconductor materials is frequently performed using diamond- or slurry-based wire sawing processes. These sawing processes not only result in the loss of potentially valuable material due to the kerf, but also cause surface roughness and, below the surface, damage to the crystal structure. These aspects of wafering with sawing processes necessitate additional polishing and grinding steps in the wafer fabrication process, leading to further damage and increased process costs.

[0003] To address these problems in conventional wafering and thinning of semiconductors, so-called... kerf-lessTechnologies have been developed that promise to reduce—if not eliminate—cutting gap losses, as well as damage below the surface and in grinding process steps. In particular, externally applied so-called spalling processes use stresses—often temperature-related—to separate crystalline materials along crystal planes with well-defined thickness. Spalling can be performed with nickel-chromium alloys, silver-aluminum paste, epoxy resin, aluminum, and nickel. Kerf-free wafering techniques have the potential to drastically reduce damage in semiconductor manufacturing processes. Stress-based separation methods, such as spalling, use externally applied stresses to separate crystalline materials along their crystal planes with well-defined thickness. However, substrates exhibit so-called Wallner lines after spalling, which result from crack propagation within the crystal.

[0004] Spalling was achieved by exploiting differences in the coefficients of thermal expansion between a brittle material and a polymer adhering to its surface. Cooling the bonded materials below the polymer's glass transition temperature induces stresses that lead to material separation along a crack plane. An advantage of this particular method, compared to other types of spalling, is that the cooling process does not result in increased diffusion of unwanted chemical components through the material, unlike high-temperature methods also used for spalling.

[0005] However, spalling processes tend to be limited in their control over the achieved wafer thickness, and tuning the vertical location of crack propagation is complex. Furthermore, spalling produces very prominent Wallner line patterns on the surface. This pattern consists of stripe-like grooves and ridges originating from crack propagation in the semiconductor material and allows for the deduction of crack dynamics in the substrate. Typically, the crack starts at a certain point near the edge and then propagates rapidly from the edge of the substrate. Wallner lines on conventional spalling surfaces significantly increase the resulting surface roughness, often to the point where additional polishing or grinding steps are required before further processing and circuit fabrication on the substrate.

[0006] As a solution to eliminate Wallner patterns, a laser-assisted spalling process was introduced prior to the actual spalling process. Here, the horizontal plane of crack propagation is defined by the material's treatment plane using short laser pulses and a high numerical aperture of the optics. Since the laser photon energy is selected below the material's bandgap energy, the material is transparent to the laser radiation, allowing for deep penetration. By translating the focusing optics along the optical axis, the laser modification plane can be precisely defined by the focal plane of the focusing optics, which can be adjusted with micrometer precision.

[0007] The subsequently published WO 2016 / 113309 A1 describes a method for separating a solid layer, such as a wafer, from a solid substrate. The method involves creating modifications within the solid substrate that define a curved detachment zone and then detaching the solid layer from the substrate along this curved detachment zone. The modifications within the solid substrate are created, for example, by a laser.

[0008] JP 2014 161908 A (disclosing the preamble of claim 1) describes a method for separating a structured solid layer from a solid. The method involves using laser beams to create modifications in the solid that define a structured detachment zone, and then separating the solid layer along the structured detachment zone to obtain a solid layer with a structured surface.

[0009] The object of the present invention is to improve the known manufacturing process, in particular to make it more efficient or to accelerate it.

[0010] The aforementioned problem is solved according to the invention by a method according to claim 1.

[0011] Another embodiment of this invention is defined in dependent claim 2.

[0012] The invention is described below by way of example only, using the attached figures.

[0013] This shows, for example: Fig. 1 schematically shows the compensation of material properties by adjusting the laser exposure depending on local property differences of the material; Fig. 2a schematically shows the generation of a modification sufficient to induce a crack; Fig. 2b schematically shows the generation of a modification which, compared to the modification from Fig. 2aexhibits a significantly greater extent in the longitudinal direction of the solid and, after the separation of the solid layer, causes the solid layer to bulge or bend; Fig. 3 schematically the creation of a trench in a solid, wherein the trench is preferably spaced apart from the outer circumferential surface of the solid; Fig. 4a schematically an arrangement in which particles, such as dust, accumulate in the path of the laser radiation, in particular at the intersection point of the reflected beams; Fig. 4b schematically an arrangement in which a purge, in particular by means of gas, such as ionized gas, is provided to remove the particles from the intersection point of the reflected beams; Fig. 5 schematically an arrangement according to which one or more coating(s) are arranged on the solid, wherein the coating(s) preferably have at least one different optical property than the solid; Fig.Fig. 6a schematically the incident laser beams at the Brewster angle; Fig. 6b curves to illustrate the relationship between incident angle and reflection; Figs. 7-18 examples of the calculation of the optimal incident angle for a 1 / e² Gauss profile and various numerical apertures, taking into account the refractive index dependence of the surface reflection; Fig. 19a schematically the curve of the cold-split method; Fig. 19b schematically the curve of the laser-assisted spalling method; Fig. 19c a photograph of a surface of a solid layer exposed according to method 19a; Fig. 19b a photograph of a surface of a solid layer exposed according to method 19b; Fig. 20a-f SEM images of surfaces; Fig. 21a a microscopic image of a surface after spalling; Fig. 21b Raman spectra from three different sites in 6H silicon carbide; Fig. 22 an example of a laser beam profile; and Fig.Fig. 23a schematically shows an example of the bending of the produced solid layer resulting from a first number of modifications and / or modification layers; and Fig. 23b schematically shows another example of the bending of the produced solid layer resulting from a second number of modifications and / or modification layers, the second number being greater than the first number; and Fig. 24a a schematic setup of a Raman instrument as it is used; Fig. 24b various exemplary vibration states of the lattice vibrations of SiC; Figs. 25a and 25b two diagrams representing doping concentrations in a solid; Fig. 26a a feedforward process and Fig. 26 a feedback process.

[0014] In Fig. 1A schematic diagram shows a setup for carrying out a method for separating at least one solid layer 14 from a solid body 1. Modifications 2 define a crack guidance area 4 for guiding a crack to separate a solid portion 14, in particular a solid layer, from the solid body 1. The fact that the laser beams 10 and 11 are shown at different locations on the solid body 1 indicates that the solid body 1 is moved relative to a laser application device 8. The laser beams 10 and 11 thus represent situations at different times.In principle, laser beams 10 are generated by means of the laser application device 8 to generate at least one modification 2, wherein the laser application device 8 is adjusted for the defined modification generation depending on at least one parameter, namely the transmission of the solid 1, at defined locations and for a defined solid depth. In the representation according to . Fig. 1 Laser beams 10 and 11 thus represent laser beams with different properties, or at least with one different property. The change in property from laser beam 10 to laser beam 11 takes into account the altered material properties in the respective area of ​​the solid 1 being illuminated. In the case shown, for example, laser beam 11 illuminates an area of ​​the solid 1 that exhibits a changed transmission, which could, for example, result from a doping spot.

[0015] Preferably, after the production of the modifications 2, the solid layer 14 is separated from the solid 1 according to the illustration of Fig. 19b .

[0016] Fig. 2a Figure 1 shows a situation in which modifications 2 are generated with a first extension in the longitudinal direction L of the solid body 1. The longitudinal direction L preferably extends orthogonally or substantially orthogonally to the incoming radiation surface 17 of the solid body 1, wherein the incoming radiation surface 17 of the solid body 1 is part of the solid body layer 14 after the solid body layer 14 has been separated. This is particularly preferred for all embodiments described in this document. Fig. 2a The generated modifications 2 are sufficient to guide a crack for separating the solid layer 14.

[0017] Fig. 2b shows an embodiment according to which the generated modifications 2 offer a difference compared to Fig. 2aexhibit a greater extent in the longitudinal direction L. Additionally or alternatively, several, in particular two or three or more than two or three, layers of modifications 2 can be generated, at least section by section.

[0018] This is advantageous because by extending the laser layer in the beam direction (depth or solid length) to a greater extent than necessary, the stress that the laser layer generates in the unmodified material can be increased. Thus, more material can preferably undergo a phase transformation, amorphization, or other modification than is required for the polymer split (cf. Fig. 19bThis increased stress serves to promote spontaneous splitting of the material (without polymer). The laser parameters, laser beam parameters, or parameters with which the laser application device is configured—for spontaneous splitting and the polymerization process—can therefore differ considerably, e.g., in the required numerical aperture and / or pulse length and / or pulse energy. A larger laser layer leads to higher pressure in the solid, thus increasing the probability of spontaneous splitting. Furthermore, this embodiment is advantageous becauseSince a solid layer 14 can be produced as a curved or bent solid layer 14, this method can preferably also be used to produce at least one solid layer 14 that is at least partially curved or bent. To produce a curved or bent solid layer (or bent or curved wafer), the method preferably comprises at least the following steps: moving the solid 1 relative to a laser irradiation device 8, successively generating laser beams 10 by means of the laser irradiation device 8 to generate at least one modification 2 inside the solid, wherein the modifications 2 define a crack-guiding area 4 for guiding a crack to separate a solid portion 6, in particular a solid layer, from the solid 1, and wherein the modifications cause a pressure increase in the solid.wherein the solid layer 14 is separated from the solid body 1 by crack propagation as a result of the pressure increase along the crack propagation area 4, wherein at least a portion of the modifications 2 as a component of the solid layer 14 is separated from the solid body 1, wherein the solid layer 14 is transformed into a curved or convex shape due to the modifications 2, wherein the further surface portion 16 of the solid layer resulting from the crack propagation area 4 is thus at least sectionally convex.

[0019] Fig. 3Figure 1 shows an arrangement in which, preferably after the generation of a crack-guiding region 4 in the solid body 1, a trench 26 is generated at least sectionally and preferably circumferentially, starting from the input surface 17 in the longitudinal direction L of the solid body 1. After the generation of the trench 26, the solid body layer 14 can be separated from the solid body by generating further modifications 2 using laser beams 10, which are preferably also introduced via the input surface 17. Alternatively, a stress-generating layer 18 is preferably arranged or generated at least on the area enclosed, surrounded, or bounded by the trench 26, in particular on the surface of the subsequent solid body layer 14.

[0020] The stress-generating layer preferably consists of a polymer material, in particular PDMS, and is in a further step preferably at least partially and particularly preferably completely subjected to thermal stress, in particular cooled, and in particular cooled below its glass transition temperature. This preferably applies to all embodiments described herein in which a stress-generating layer is used or employed.

[0021] As a result of the stress generation, a crack separates the solid layer 14 from the remaining solid 1.

[0022] Preferably, in a further step, the solid body 1 undergoes a surface treatment. Preferably, the frame 28 formed between the groove 26 and the surrounding surface and / or the surface of the solid body 1 exposed by removing the solid layer 14 is smoothed, in particular by grinding, lapping, polishing, or etching. Preferably, the frame 28 and the exposed surface are treated, in particular by machining, such that their surfaces lie in the same plane.

[0023] A method for separating at least one solid layer or solid layer, in particular a solid disk, 14 from a solid or donor substrate 1 is thus provided, preferably comprising at least the following steps: providing a solid 1, generating modifications 2 inside the solid 1 by means of laser beams 10, wherein the modifications 2 define a detachment area or crack propagation area along which the solid layer 14 is separated from the solid 1, removing material from the solid 1, in particular to create a circumferential depression 12, wherein the material removal is carried out in the longitudinal direction of the solid, whereby the detachment area is exposed by the material removal, and separating the solid layer from the solid.

[0024] This is advantageous because laser processing to the edge is problematic, and thus the generated solid layer 14 exhibits very homogeneous properties even at its edge. The basic concept presented here therefore preferably comprises laser ablation / grinding / material removal from above to create a notch or groove, thereby opening or exposing the previously generated laser layer. Subsequently, the solid layer 14 or the target wafer with the stress-generating layer 18 is lifted out. The remaining edge or frame 28 can then be ground away again during further surface preparation. Thus, the laser layer can be exposed by ablation from above, in particular by waterjet cutting or laser ablation, and edge effects during polymer splitting can be avoided.

[0025] Fig. 4ashows an arrangement in which particles, such as dust, accumulate at the intersection point of the reflected radiation marked with reference numeral 30, thus negatively affecting the generation of modifications.

[0026] Fig. 4b Figure 1 shows an arrangement in which a flushing device 32 or flushing is provided. A fluid, in particular a gas and preferably an ionized gas, is thus preferably supplied to the intersection point 30 in order to flush away the particles accumulating in the intersection point 30 by means of the fluid flow.

[0027] Thus, a method for separating at least one solid layer or solid layer, in particular a solid disk, 14 from a solid or donor substrate 2 is provided, which preferably comprises at least the following steps: providing a solid 1, adjusting the flow behavior of a gas, in particular air, located between the solid and the laser irradiation device 8, in particular in the area of ​​the radiation path, to prevent dust accumulation in the area of ​​the laser radiation, generating modifications 2 inside the solid 1 by means of laser beams 10 of a laser irradiation device 8, wherein the modifications 2 define a detachment area or crack propagation area along which the solid layer 14 is separated from the solid 1, and separating the solid layer from the solid.This solution is advantageous because high laser intensities statically charge dust, and this dust can be flushed out of the area between the lens and the workpiece by purging, particularly with ionized gas. The gas purge thus drives the dust out of the space between the lens 9 of the laser application device 8 and the workpiece or solid 1. Additionally or alternatively, the fluid flow, particularly the gas flow, can be directed through the lens to generate a cooling effect against absorbed power. Therefore, the lens is preferably designed to guide a fluid, particularly the purge fluid.

[0028] Furthermore, or additionally, spherical aberrations can be compensated for at the lens. This alters the focus at the surface (incorporating the material with a different refractive index), which worsens the focus in air and thus results in lower intensity, leading in turn to reduced particle suction or vacuuming effect. Additionally or alternatively, reduced surface reflection can be achieved. This can be accomplished, for example, by applying specific layers or coatings, particularly spin coating, and / or by Brewster irradiation with polarized light.

[0029] Fig. 5Figure 1 shows a schematic arrangement in which the solid 1 is provided with at least one coating 34. The coating 34 can be single-layered or multi-layered. Preferably, the coating has a refractive index difference compared to the material of the solid 1; in particular, the refractive index of the material of the solid 1 is preferably higher than that of the coating 34. It is also conceivable that the coating 34 is composed of several layers, preferably at least two of which have a refractive index difference. Preferably, the refractive index of the layer closest to the solid 1 is greater than the refractive index of a layer located further away from the solid 1.

[0030] This schematic setup enables a method for separating at least one solid layer or solid layer, in particular a solid disk, 14 from a solid or donor substrate 1. Preferably, this method comprises at least the following steps: providing a solid 1, wherein the solid 1 has at least one coating 34 whose refractive index differs from that of the surface of the solid 1 on which the coating 34 is arranged, or wherein a coating 34 is produced on the solid 1 whose refractive index differs from that of the surface of the solid 1 on which the coating 34 is arranged; generating modifications 2 inside the solid 1 by means of laser beams 10 from a laser application device 8, wherein the modifications 2 create a crack propagation area 4 (cf. analogously). Fig. 1) is specified, along which a separation of the solid layer 14 from the solid 1 takes place.

[0031] The coating can be produced, for example, by spin coating. A solvent mixed with nanoparticles of materials with a high refractive index deposits one or more thin (sub-wavelength) layers with a slightly higher refractive index onto the solid or workpiece. This creates an intermediate surface with a reduced refractive index difference, resulting in lower surface reflection, less contamination, and improved material performance for more efficient processing. Spin coating is advantageous because it is fast and inexpensive. Possible nanoparticles include, among others, silicon (n=3.55), silicon carbide (n=2.6), titanium oxide (n=1.8), glass (n=1.5), and Al₂O₃ (n=1.72). In the case of multiple layers with gradually increasing refractive indices for even more efficient refractive index matching and antireflective properties, a multi-layer process is conceivable.As a purely exemplary example, a layer arrangement could then be produced consisting of the following layers: 1st layer: Si, 2nd layer: SiC, 3rd layer: TiO2, each layer preferably 50–400 nanometers thick. This method is further advantageous because spin-coating such layers can also compensate for minute roughness on the material surface (better material coupling due to less scattering at the interface, better wavefront overlap at the focus depth, and thus a lower laser power is required). This leads to more efficient processing because a higher multiphoton transition probability results. The spin-coating layer, or the creation of a coating 34, can be applied to the ingot or solid 1 during the surface conditioning and reconditioning step after splitting or separating the solid layer 14.Therefore, a grinding / lapping / etching or polishing step can be performed first, followed by, or in combination with, one of the preceding steps, the spin-coating step or the coating step that applies the thin layer or coating 34. Fig. 6Figure 1 schematically shows an arrangement for coupling laser beams 10, whereby reflection is reduced. Preferably, the laser beams 10 are coupled in at the Brewster angle. The Brewster angle is an angle of incidence for light of a specific polarization (E-vector pointing into the material, not along the surface) at which no reflection occurs. The prerequisites for this are that the light is injected at an angle dependent on the refractive index difference between air and material. Furthermore, the light must be polarized (usually the case with laser light, requiring single-mode lasers and not photonic crystal fibers). Brewster angle coupling thus serves to minimize back reflections. When the light is injected at the Brewster angle, the 30% surface reflection can be almost completely utilized for material processing at depth within the material.

[0032] Irradiation at the Brewster angle is complex because the different beam components travel different path lengths in the higher-refractive-index medium. The focus must therefore be adjusted by increasing the energy and / or by beam shaping. Beam shaping is preferably achieved, for example, using one or more diffractive optical elements (DOEs), which compensate for this difference depending on the laser beam profile. The Brewster angle is relatively large, which, at high numerical apertures, places demands on the optics, their dimensions, and the working distance. Nevertheless, this solution is advantageous because reduced surface reflections also contribute to reduced surface damage, as the light intensity couples better into the material. Laser beams 10 can also be irradiated at the Brewster angle, or substantially at the Brewster angle, in all other embodiments disclosed in this document.For Brewster angle coupling, reference is hereby made to the document "Optical Properties of Spin-Coated TiO2 Antireflection Films on Textured Single-Crystalline Silicon Substrates" (Hindawi Publishing Corporation International Journal of Photoenergy, Volume 2015, Article ID 147836, 8 pages, http: / / dx.doi.org / 10.1155 / 2015 / 147836). This document is hereby incorporated in its entirety into the present patent application by reference.

[0033] The aforementioned document, which is included in this document, reveals in particular calculations regarding the optimal angle of incidence for various materials and thus refractive indices. The energy of the laser or the laser application device 8 is adjusted not so much depending on the material, but rather on the possible transmission at a specific angle. Therefore, if the optimal transmission is, for example, 93%, these losses must be taken into account compared to tests with perpendicular incidence and losses of, for example, 17%, and the laser power must be adjusted accordingly.

[0034] For example, 83% transmission perpendicularly compared to 93% at an angle means that to achieve the same energy at depth, only 89% of the laser power used with perpendicular illumination is required (0.83 / 0.93=0.89). In accordance with the invention, the portion of the beam at an angle thus preferably serves to reduce light loss through surface reflection and to transmit more light into the depth. A possible subsequent problem that can arise in certain configurations is that the focus at depth may acquire a "skew" profile, and therefore the achieved intensities—the key parameter for multiphoton processing—may again be lower, possibly even lower than with perpendicular illumination, where all beam components follow the same optical path through the material.This can then preferably be achieved by one or more diffractive optical elements, a continuous wedge, or multiple continuous wedges—and / or other optical elements—in the beam path, which compensate for these additional paths and / or the influence on the individual beams—especially different spherical aberrations across the beam profile. These DOEs can be calculated numerically using suitable software solutions (e.g., Virtuallab from Lighttrans, Jena) and then manufactured or provided.

[0035] Thus, it provides a method for separating at least one solid layer or solid layer, in particular a solid disk, 14 from a solid or donor substrate 1. The method preferably comprises at least the following steps: providing a solid 1, generating modifications 2 inside the solid 1 by means of laser beams 10 from a laser application device 8, wherein the modifications 10 define a crack propagation area along which the solid layer 14 is separated from the solid 1, the laser radiation being directed onto the solid 1 at the Brewster angle or with a deviation in the range of -10° to +10° from the Brewster angle. Furthermore, the method preferably comprises the step of separating the solid layer 14 from the solid 1.

[0036] It was thus recognized that a high refractive index difference between air and material results in power losses of up to 30% with perpendicular irradiation. With a 100W laser, therefore, 30W are unavailable for material processing or have other effects. For example, it was also recognized that contamination can occur on the optics, similar to what happens with so-called "optical tweezers." In this case, minute dust particles in both air and liquids always migrate towards the focus of the laser beam (highest intensity) at high laser intensities – reflected power at the surface is focused in the air or near the optics, and the dust is driven / drawn towards the optics. Furthermore, it was recognized that at 100W laser power and 97% transmission at the lens, significant heat (3W) can also enter the lens, which must be dissipated / compensated to prevent thermal damage / changes in the process.Furthermore, it was recognized that high power levels carry the risk of surface damage. This is because the absorption of the material can be increased due to surface conditions; therefore, even the smallest dust particles can first burn up in the laser beam and then form absorption nuclei, which can lead to further damage through absorption. It was also recognized that high power levels are distributed across multiple foci in the focal plane using diffractive optical elements (DOEs). DOEs exhibit interference phenomena even before reaching the focal plane. It was found that interference at the surface, in front of the focal plane, can create local intensity maxima that can damage the surface and lead to reduced transmissivity for laser radiation for processing at depth. Finally, it was recognized that some materials (e.g., SiC) exhibit local differences in refractive index and other material properties (e.g.,Absorption, transmission, scattering), e.g., due to material doping (frequent occurrence: doping spot). Furthermore, it was recognized that, depending on the surface roughness of the material at the laser coupling surface, the laser wavefront can be significantly impaired at depth within the material, resulting in reduced focus intensity (lower multiphoton transition probability), which in turn would lead to higher intensities with the aforementioned problems.

[0037] One, several, or all of these problems can be addressed by one or more of the methods disclosed herein. It can therefore preferably be understood as a method for separating at least one solid layer from a solid body. In this method, modifications 2 preferably define a crack-guiding region 4 for guiding a crack to separate a solid component 6, in particular a solid layer, from the solid body 1. The method preferably comprises at least the following steps: moving the solid body 1 relative to a laser application device 8, successively generating laser beams 10 by means of the laser application device 8 to generate at least one modification 2 each time, and separating the solid layer from the solid body.

[0038] This describes a spalling process that scales particularly well for large-area semiconductor substrates with diameters up to 300 mm or greater. To eliminate Wallner line patterns, a laser conditioning process, preferably with a high numerical aperture, is used at photon energies below the material's bandgap energy. This process leads to multiphoton interactions within the material and, after spalling, yields a surface roughness of preferably Ra < 1 µm.

[0039] Preferably, one or more of the aforementioned solutions can be combined, as this can result in even better solid layer production or separation. According to the method, the laser application device 8 is set to generate defined modifications depending on at least one parameter, namely the transmission of the solid, at defined locations and for a defined solid depth, and / or the modifications cause a pressure increase in the solid, whereby the solid layer is separated from the solid as a result of the pressure increase along the crack propagation area by crack propagation, wherein preferably at least a portion of the modifications is separated from the solid as a component of the solid layer, and wherein the solid layer is preferably transformed into a curved or convex shape due to the modifications.wherein the further surface area of ​​the solid layer resulting from the crack propagation region is thus at least partially convex and / or the solid has at least one coating whose refractive index differs from the refractive index of the surface of the solid on which the coating is applied, or a coating is produced on the solid whose refractive index differs from the refractive index of the surface of the solid on which the coating is applied, and / or the laser radiation is directed onto the solid at the Brewster angle or with a deviation in the range of -5° to +5°, in particular with a deviation in the range of -4° to +4° or with a deviation in the range of -3° to +3° or with a deviation in the range of -2° to +2° or with a deviation in the range of -1° to +1° from the Brewster angle,and / or the method additionally or alternatively comprises one or more of the following steps: removal of material from the solid body 1, in particular to create a circumferential depression 12, wherein the material removal takes place in the longitudinal direction of the solid body, whereby the crack propagation area is exposed by the material removal, or adjusting the flow behavior of a gas, in particular air, located between the solid body and the laser application device 8, in particular in the area of ​​the radiation path, to prevent dust accumulation in the area of ​​the laser radiation.

[0040] The Figures 7-18 show examples of the calculation of the optimal angle of incidence for a 1 / e2 Gauss profile and various numerical apertures, taking into account the refractive index dependence of the surface reflection.

[0041] The Figures 7-10 The results are shown when using silicon carbide (n=2.7)

[0042] Objective: To maximize the laser power coupled into the sample, ideally by utilizing the Brewster angle for minimal surface reflection with p-polarized light. Result: For a NA of 0.8, Brewster coupling is not worthwhile (the beam cone couples almost directly under Brewster at the outer edge); smaller NAs can benefit from it, especially NAs of 0.2; higher NAs have an ideal angle somewhere in between.

[0043] Fig. 7 The upper figure shows: relative coupled power over incidence angle for Gaussian profile and the lower figure shows: reflection coefficient for p-(red) and s-(blue) polarized light and angle Gaussian profile for NA=0.2 (green), optimal incidence angle here 63.8°.

[0044] Fig. 8The upper figure shows: relative coupled power over incidence angle for Gaussian profile and the lower figure shows: reflection coefficient for p-(red) and s-(blue) polarized light and angle Gaussian profile for NA=0.4 (green), optimal incidence angle here 52.5°.

[0045] Fig. 9 The upper figure shows: relative coupled power over incidence angle for Gaussian profile and the lower figure shows: reflection coefficient for p-(red) and s-(blue) polarized light and angle Gaussian profile for NA=0.6 (green), optimal incidence angle here 35.6°.

[0046] Fig. 10 The upper figure shows: relative coupled power over incidence angle for Gaussian profile and the lower figure shows: reflection coefficient for p-(red) and s-(blue) polarized light and angle Gaussian profile for NA=0.8 (green), optimal incidence angle here 0°.

[0047] The Figures 11-14The results are shown when using silicon (n=3.6).

[0048] The goal here is to maximize the laser power coupled into the sample, ideally by utilizing the Brewster angle for minimal surface reflection with p-polarized light. The result: For a NA of 0.8, Brewster coupling is not worthwhile (the beam cone couples almost at the Brewster angle on the outside), smaller NAs can benefit from it, especially NAs of 0.2, while higher NAs have an ideal angle somewhere in between.

[0049] Fig. 11 The upper figure shows: relative coupled power over angle of incidence for Gaussian profile for silicon and the lower figure shows: reflection coefficient for p-(red) and s-(blue) polarized light and angle Gaussian profile for NA=0.2 (green), optimal angle of incidence here 67.9°.

[0050] Fig. 12The upper figure shows: relative coupled power over angle of incidence for Gaussian profile for silicon and the lower figure shows: reflection coefficient for p-(red) and s-(blue) polarized light and angle Gaussian profile for NA=0.4 (green), optimal angle of incidence here 57.4°.

[0051] Fig. 13 The upper figure shows: relative coupled power over angle of incidence for Gaussian profile for silicon and the lower figure shows: reflection coefficient for p-(red) and s-(blue) polarized light and angle Gaussian profile for NA=0.6 (green), optimal angle of incidence here 43.6°.

[0052] Fig. 14 The upper figure shows: relative coupled power over incidence angle for Gaussian profile for silicon and the lower figure shows: reflection coefficient for p-(red) and s-(blue) polarized light and angle Gaussian profile for NA=0.8 (green), optimal incidence angle here 0°.

[0053] The Figures 15-18 The results are shown when using Sapphire / ALO (n=1.72)

[0054] The goal here is to maximize the laser power coupled into the sample, ideally by utilizing the Brewster angle for minimal surface reflection with p-polarized light. The result: For an NA of 0.8, Brewster coupling is not worthwhile (the beam cone couples almost directly under Brewster at the outer edge), smaller NAs can benefit from it, especially NAs of 0.2, higher NAs have an ideal angle in between, with NAs of 0.6 requiring almost no angle for illumination.

[0055] Fig. 15 The upper figure shows: relative coupled power over incidence angle for Gaussian profile for ALO and the lower figure shows: reflection coefficient for p-(red) and s-(blue) polarized light and angle Gaussian profile for NA=0.2 (green), optimal incidence angle here 54.9°.

[0056] Fig. 16The upper figure shows: relative coupled power over incidence angle for Gaussian profile for ALO and the lower figure shows: reflection coefficient for p-(red) and s-(blue) polarized light and angle Gaussian profile for NA=0.4 (green), optimal incidence angle here 41.2°.

[0057] Fig. 17 The upper figure shows: relative coupled power over angle of incidence for Gaussian profile for ALO and the lower figure shows: reflection coefficient for p-(red) and s-(blue) polarized light and angle Gaussian profile for NA=0.6 (green), optimal angle of incidence here 14.5°.

[0058] Fig. 18 The upper figure shows: relative coupled power over incidence angle for Gaussian profile for ALO and the lower figure shows: reflection coefficient for p-(red) and s-(blue) polarized light and angle Gaussian profile for NA=0.8 (green), optimal incidence angle here 0°.

[0059] Figure 19aoutlines a standard procedure for laser-free spalling of large-area substrates. The wafer samples,

[0060] The wafers typically used have sharp edges to avoid complications caused by rounded edges. Rounded edges are used on conventional wafers to prevent the formation of cracks at the wafer edges, which can propagate inwards and disrupt the substrate and manufacturing processes on it.

[0061] The process preferably proceeds as follows: after a standard cleaning procedure, the wafers are coated with a primer layer to improve surface adhesion and a sacrificial layer to improve polymer-wafer separation at the end. The wafer is then coated with a PDMS (polydimethylsiloxane) polymer film of varying thickness and a PDMS adhesive containing a platinum polymerization catalyst.

[0062] The samples are then pre-cooled to a temperature just above the polymer glass transition temperature before being immersed in liquid nitrogen. Depending on the sample size, it will reach the temperature of liquid nitrogen up to 20 seconds later. At this point, the system is in thermal equilibrium. The semiconductor layers then separate in a spontaneous spalling event. The spalling process is induced by the polymer glass transition, and as a consequence, the Young's modulus in the polymer increases significantly. The additional difference in coefficients of thermal expansion (CTE) between the semiconductor and the polymer then induces sufficient stress to separate the crystal horizontally. It is important that the process requires a relative contraction of the polymer with respect to the semiconductor.The next process step is the immersion of the semiconductor parts with attached PDMS films into a separation bath, which ultimately dissolves the sacrificial layer and thus enables the recycling of the polymer as well as the provision of the semiconductor wafers for further processing steps.

[0063] Fig. 19a Figure 1 shows a conventional spalling process for substrates with an ensemble of preferably two stress layers, i.e., polymer films. The films are attached to both sides of the substrate, followed by a rapid cooling step to induce thermal cycling, followed by cracking and separation of the substrate.

[0064] The laser-assisted spalling process shown in 19b is remarkably similar. The main difference is an additional laser processing step in which a laser beam is focused on a defined plane or path within the sample and then scans the sample. The laser layer created in this way then defines the plane of crack formation and thus also the separation in the subsequent spalling process.

[0065] Fig. 19b This illustrates the laser-assisted spalling process. In addition to the process steps shown in 19a, a structurally weakened layer is created in the material in a further process step using a modification layer or laser system generated by laser beams, which defines a preferred plane for spalling crack propagation.

[0066] A typical resulting wafer surface from laser-free spalling is in Figure 19cA pattern of Wallner lines results from crack propagation within the material. Grooves can be identified on the material's surface, tracing the crack behavior along the separation plane. Cracks originate at the sample's edges and propagate inwards, creating a detailed pattern, as shown in Figure 19cThe fourfold symmetry of the pattern is a consequence of the fourfold crystal symmetry in silicon, with a singularity or center point of the crack waves in the middle of the sample. However, to compete technologically with wire sawing processes, surface quality after separation is crucial for any spalling process. The resulting total thickness variation (TTV) of spalling surfaces without a laser process is typically far above industry requirements. The typical TTV of spalling processes is on the order of 50 µm, which would require grinding steps prior to further processing, driving costs up too high. Using the LAS process, however, results in surface roughness values ​​of Sa < 1 µm. Sa is the arithmetic mean of the absolute values ​​of the surface coordinates z(x, y).

[0067] Fig. 19cThis image shows one half of a 300 mm silicon wafer after separation using conventional spalling. Wallner lines are clearly visible as crack grooves and are indicative of high surface height variation (TTV).

[0068] Fig. 19d This image shows half of a 300 mm silicon wafer after laser-assisted spalling. The surface is homogeneous with a surface roughness of less than 1 µm and no visible crack propagation fringes. The vertical line on the left side of the substrate is due to a limitation in the travel distance of the laser system's substrate holding stage.

[0069] The Figures 20a-f show an overview of the material surfaces after laser-assisted spalling. Fig. 20a shows a silicon surface from the in Figure 19d The sample shown has a surface roughness of Sa = 0.79µm. Fig. 20bshows a sapphire (Al 2 O 3 ) substrate surface (C-plane) after laser-assisted spalling, with a surface roughness of Sa = 1.96µm. Fig. 20c and Fig. 20d Show surfaces of silicon carbide polymorphs 4H and 6H (both with N doping) after laser-assisted spalling with surface roughnesses Sa = 1.85µm and Sa = 1.29µm. Fig. 20e shows an example of spalling of non-crystalline material, surface of polycrystalline Al 2 O 3 , with a surface roughness Sa = 3.89µm. Fig. 20f Shows quartz glass, principle study, experiment of laser-assisted spalling, with surface roughness of Sa = 6.89µm.

[0070] Fig. 21a shows a microscopic image of a surface after spalling. Fig. 21b shows Raman spectra from three different sites in 6H silicon carbide. The Raman spectra from the darker region (on the right side of Fig. 21a) are curves K1 and K2, with curve K3 being the Raman spectrum from the brighter area on the left side in Fig. 21a The peak height is reduced for almost all peaks in the darker areas, and no Raman peaks are visible at position 2 in the darker area.

[0071] Fig. 22 shows another additional or alternative beam profile. When subjected to Brewster irradiance with high numerical apertures, the laser beam profile can be adjusted. Thus, at high numerical apertures (NA), higher intensity can result in the flanks of the incident laser beam profile. In the extreme case, this is a kind of donut profile with a clear intensity minimum in the center. However, it is also conceivable that the laser beam profile is designed as a Gaussian profile flattened in the middle. This preferably exploits the fact that at high NA, the edge regions of the laser profile can already approach the Brewster angle. The example profile shown in Fig. 22 could therefore preferably be generated with a relatively higher intensity component (compared to the other embodiments) in the flanks.

[0072] Fig. 23a describes a first solid-state generation configuration. According to this configuration, it is possible to separate the solid layer 14 from the solid 1. According to this configuration, a first number of modifications 2 are generated in the solid. The modifications 2 preferably already cause a deflection of the separated solid layer 14.

[0073] In Fig. 23bA second solid-state generation configuration is shown. According to this configuration, it is also possible to separate the solid layer 14 from the solid body. However, according to this configuration, a second number of modifications 2 are generated in the solid body 1. The second number of modifications 2 is preferably larger than the first number of modifications. Additionally or alternatively, it is possible that the configuration has several modification layers 2.1, 2.2 or has more modification layers 2.1, 2.2 than the first configuration, which can also have several modification layers. Additionally or alternatively, it is also conceivable that individual or the majority of the modifications 2 according to the second configuration are more pronounced than in the first configuration. More pronounced preferably means that the individual modifications each extend over a larger volume than in the first configuration.The laser beams 10 preferably penetrate the solid 1 longitudinally or at an angle of up to 60° to the longitudinal direction L of the solid 1 via a surface, preferably a planar one, which is preferably part of the solid layer, and the crack propagation region 4 is then preferably formed from several layers of modifications 2. The layers are preferably spaced apart from each other in the longitudinal direction L or offset. Preferably, at least several of the modifications 2 have a longitudinal extent L of between 1 and 50 µm, and / or the laser beams 10 are preferably introduced into the solid 1 to generate the modifications 2 such that the numerical aperture is less than 1, preferably less than 0.9, less than 0.8, less than 0.7, less than 0.6, or less than 0.5.

[0074] Thus, a method for separating at least one solid layer from a solid is described, wherein the modifications define a crack guidance area for guiding a crack to separate a solid component, in particular a solid layer, from the solid, comprising at least the following steps: moving the solid relative to a laser application device, successively generating laser beams using the laser application device to generate at least one modification each time, wherein the laser application device is adjusted at defined locations and for a defined solid depth depending on at least one parameter, in particular the transmission of the solid.wherein inhomogeneities of the solid in the area of ​​the treated surface and / or in the area of ​​the treated volume of the solid are compensated for by adjusting the laser application device, separating the solid layer from the solid.

[0075] Fig. 24aFigure 58 shows a Raman instrument 58. The Raman instrument 58 shown here has a laser 60 for emitting radiation. The radiation is preferably supplied to an optical system for excitation by means of at least one optical fiber 61 and is preferably focused by this optical system, in particular a lens 64, preferably into the solid. This radiation is at least partially scattered, whereby light components having the same wavelength as the radiation emitted by the laser are preferably filtered out by means of a filter device or excitation filter 62. The remaining radiation components are then supplied to a spectrograph 68 and detected by means of a camera device, in particular a CCD detector 70, and evaluated or processed by a control device 14, 72, in particular a computer. Thus, atomic vibrations in the crystal are preferably excited by a preferably external or, more preferably, a further laser.These vibrations are generated by light scattering at crystal atoms, resulting in observable scattered light with a photon energy altered by the vibrational energy. If multiple excitable vibrations are present, multiple peaks appear in the spectrum of the scattered light. The resulting Raman scattering spectrum can then be examined in more detail using a spectrometer (grating spectrometer) (so-called Raman spectroscopy). In this method, the shape of the individual Raman lines reflects the local conditions within the crystal, and the degree of doping can be deduced from the shape of the Raman line.

[0076] Fig. 24bThis shows what possible lattice vibrations in SiC look like, where these modes are determined by crystal symmetry and directions and can also be simultaneously excited. The views shown extend along the crystal axis A. Here, vibrations of the atoms are only possible in certain directions, which are determined by the symmetry of the crystal.

[0077] Fig. 25a This shows a section of a Raman curve for a nitrogen-doped 4H silicon carbide solid (example spectrum for Raman on doped SiC). The shape of the LO(PC) mode is used to measure the doping concentration and is fitted. Lower panel: Fitting residual.

[0078] Fig. 25b shows a smaller section of the Raman curve.

[0079] As shown, a direct method to determine the dopant concentration using Raman measurements results from a measurement of the shape and subsequent fit to the LO(PC) mode.

[0080] In general, the goal is therefore to adjust the laser parameters to achieve the optimal (smallest possible, shortest possible) crack propagation in the material, which still leads to successful separation as a result of crack propagation, but also minimizes or reduces all other material losses (including in grinding steps).

[0081] Fig. 26a and Fig. 26b Two ways are shown to design the lifting of individual wafers from the boule / ingot.

[0082] According to Fig. 26a This is called a feedforward loop and according to Fig. 26b designed as a feedback loop.

[0083] In feedforward processing, the distribution is characterized prior to the laser process, and a map, treatment instructions, and parameter adjustments (particularly location-dependent) are calculated for the laser process, especially for modification generation. Feedforward processing is preferably performed on the ingot / boule.

[0084] Alternatively, as in Fig. 26b As shown, a feedback loop is implemented, according to which the resulting wafer is characterized after each separation step and serves as a template for the next one.

[0085] Depending on the material and doping, different adjustments can be made during the laser process: With SiC, different adjustments to the laser parameters can be made at different depths, depending on the doping concentration. Under the following boundary conditions, this can lead to the functions also listed below: depth 180 µm, pulse duration 3 ns, numerical aperture 0.4 Low doping: 7µJ - 21mOhmcm high doping: 8µJ 16mOhmcm Depth 350µm, pulse duration 3ns, numerical aperture 0.4 Low doping: 9.5µJ - 21mOhmcm high doping: 12µJ - 16mOhmcm Formula for a depth of 180µm: Energy in µJ E0 Offset energy at lowest doping level K Factor Energy scaling R Measured doping level B Base doping level (21 mOhmcm) E=E0+B−R*K

[0086] Here K = 1 / 21 − 16 μJ / mOhmcm = 0 , 2 μJ / mOhmcm E 0 = 7 μJ B = 21 mOhmcm Example: measured doping level of 19 mOhmcm: E = 7.4µJ Formula for a depth of 350µm:

[0087] Energy in µJ E0 Offset energy at lowest doping level K Factor Energy scaling R Measured doping level B Base doping level (21 mOhmcm) E=E0+B−R*K

[0088] Here K = 2 , 5 / 21 − 16 μJ / mOhmcm = 0 , 5 μJ / mOhmcm E 0 = 9 , 5 μJ B = 21 mOhmcm Example: measured doping level of 19 mOhmcm: E = 10.5µJ

[0089] The Figures 27a to 27i show various arrangements that can be provided for initiating the crack after the production of the further material layers or components 150.

[0090] The Figures 27a-27i show various solid arrangements 176, which are advantageous for introducing crack guiding and / or crack initiation stresses.

[0091] Fig. 27a Figure 1 shows a processed solid 1 or wafer with structures or components 150.

[0092] Compared to the in Fig. 27a The solid body 1 shown is in the Fig. 27bIn the solid body 1 shown, a receiving layer 140 is arranged or created on the component side, in particular on the components 150 or the further material layers 150. The receiving layer 140 is preferably arranged on the solid body layer to be separated. The receiving layer 140 can also be referred to as a splitting film and is thus preferably laminated onto the structural side. In the subsequent step, the entire assembly is cooled, which causes the splitting or the crack initiation and / or crack propagation.

[0093] In contrast to the presentation of Fig. 27b is according to the Fig. 27cA holding layer / bonded wafer is arranged on the underside of the solid or on the exposed surface of the solid. The holding layer can also be a tool carrier or Chuck 300. In the subsequent step, the entire assembly is cooled, which causes the splitting or crack initiation and / or crack propagation.

[0094] Fig. 27d shows in relation to the Fig. 27bAn arrangement in which the solid body is provided on both sides with receiving layers 140, 146. The additional receiving layer 146 is arranged on a surface of the remaining solid body, wherein an adhesion-promoting layer 148 and / or sacrificial layer 149 and / or protective layer 142 may be arranged or formed between the additional receiving layer 146 and the solid body 1. The two receiving layers 140 and 146 are preferably laminated. In the subsequent step, the entire arrangement is cooled, thereby causing the splitting or crack initiation and / or crack propagation.

[0095] Fig. 27e shows an arrangement according to which, in contrast to the from Fig. 27dIn the known arrangement, no bonding agent layer 148 and / or sacrificial layer 149 and / or protective layer 142 is arranged or generated between the further receiving layer 146 and the solid body 1. In the subsequent step, the entire arrangement is cooled, which causes the splitting or the crack initiation and / or crack propagation.

[0096] Fig. 27f shows an arrangement that is inverse to the one from Fig. 27dThe known arrangement is constructed, i.e., the bonding agent layer 148 and / or sacrificial layer 149 and / or protective layer 142 is not arranged or produced between the further receiving layer 146 and the solid body 1, but rather between the receiving layer 140 and the solid body 1, and thus is produced or arranged on the solid body layer to be separated. One or more layers can be produced on the components 150 or the structures, for example, by means of spin coating. As a subsequent step, the entire arrangement is then cooled, which causes the split or the crack initiation and / or crack propagation.

[0097] Fig. 27g shows an arrangement or form that is a combination of the arrangements of Figures 27d and 27fThe solid is preferably laminated on both sides with split film; a protective layer and / or adhesion promoter layer and / or sacrificial layer can also be provided on both sides beneath the split film; spin coating is also possible on the structures. As a subsequent step, the entire assembly is cooled, which causes the splitting or crack initiation and / or crack propagation.

[0098] Fig. 27h shows an arrangement similar to the one in Fig. 27b The arrangement shown is such that the receiving layer is not located on a surface of the solid layer to be separated, but rather is arranged or laminated on one side of the residual solid remaining after separation. The separation then occurs as a result of cooling, analogous to the separation from an ingot or as in an ingot process.

[0099] Fig. 27i shows an arrangement similar to the one from Fig. 27cThe known arrangement is wherein one or more of the following layers or devices are arranged or produced on the component side of the solid or on or above the components 150. These layers or devices are preferably: at least or exactly one adhesion promoter layer 148 and / or at least or exactly one sacrificial layer 149 and / or at least or exactly one protective layer 142 and / or at least or exactly one stabilizing device 3, in particular a tool carrier or chuck 300 (preferably a cooling device) or another wafer. As a subsequent step, the entire arrangement is then cooled, which causes the split or the crack initiation and / or crack propagation.

[0100] Fig. 28Figure 1 illustrates an example of a writing pattern during XY machining: Arrows 170 and 172 represent the laser feed direction, and the black circles represent the different laser shots or modifications 9, which do not overlap in their damaging effect on the material. It is preferred that the laser first moves in one direction and creates modifications 9 before reversing and writing modifications 9 in the second (lower) direction.

[0101] The Figures 29a to 29d show various cooling devices 174. The solid-state arrangements 176 processed in these cooling devices 174 result from the various in the Figures 27a to 27iThe shown and described embodiments or designs of the solid bodies 1 provided with one or more receiving layers 140, 146. The cooling devices 174 shown herein all use a liquefied gas 178 as the initial cooling medium. Depending on the embodiment, this initial cooling medium is either atomized or vaporized. Preferably, the initial cooling medium is liquid nitrogen. Alternative cooling methods, e.g., using piezoelectric elements, are also conceivable and possible.

[0102] The cooling device 174 is preferably used to cool the receiving layer 140, 146 to a temperature between -85°C and -10°C, in particular to a temperature between -80°C and -50°C.

[0103] According to Fig. 29aThe cooling device 174 comprises a nitrogen bath, wherein the receiving layer is positioned at a distance from the liquid nitrogen contained in the nitrogen bath, in particular by means of an adjustable positioning device 180. Thus, the solid assembly is preferably arranged on a positioning device or on a support above a nitrogen bath. This results in a temperature gradient across the chamber height, and the temperature at the solid assembly can be adjusted via the fill level with the initial cooling medium or the position of the solid assembly 176 (distance to the bottom of the chamber).

[0104] According to the embodiments of the Figures 29b to 29dThe cooling device may preferably comprise a nebulizing agent, in particular at least or exactly one perforated pipe, for nebulizing liquid nitrogen or a nebulizing agent for nebulizing liquid nitrogen, and the cooling effect may be generated by nebulized or vaporized nitrogen.

[0105] According to Fig. 29b A homogeneous sprayer / fogger is preferably provided for spraying or atomizing. The spraying or atomizing preferably takes place above the solid assembly 176. Furthermore, temperature measurements are preferably taken for temperature control, which provide output data for controlling a valve, in particular a nitrogen valve. The temperature measurements are preferably taken on the substrate or on the solid 1 or on the receiving layer 140.

[0106] The substrate or solid 1 or solid arrangement 176 preferably rests above the chamber floor to avoid nitrogen settling at the bottom of the chamber.

[0107] According to Fig. 29c A perforated pipeline is preferably used as a homogeneous spraying device. Furthermore, temperature measurements are preferably taken for temperature control, providing output data for controlling a valve, in particular a nitrogen valve. The temperature measurements are preferably taken on the substrate or on the solid 1 or on the receiving layer 140.

[0108] The substrate or solid 1 or solid arrangement 176 preferably rests above the chamber floor to avoid nitrogen settling at the bottom of the chamber.

[0109] According to Fig. 29dFigure 176 shows a cooling device 176 comprising a homogeneous sprayer / fogger 182 for cooling preferably several or each side. Furthermore, temperature measurements are preferably taken for temperature control, which provide output data for controlling a valve, in particular a nitrogen valve. The temperature measurements are preferably taken on the substrate or on the solid 1 or on the receiving layer 140.

[0110] The substrate or solid 1 or solid arrangement 176 preferably rests above the chamber floor to avoid nitrogen settling at the bottom of the chamber.

[0111] The chamber 184 of the cooling device 174 is preferably closed in order to reduce a temperature gradient as much as possible by insulation.

[0112] Fig. 30Figure 1 shows three examples of preferred relationships between crystal lattice orientation and modification generation. This method is particularly useful for separating solid layers from a solid consisting of or containing SiC. These relationships lead to a further method. This further method is preferably used for separating at least one solid layer from at least one solid 1, in particular from a wafer from an ingot, or for thinning a wafer. The further method according to the invention preferably comprises at least the following steps: generating a plurality of modifications 2 inside the solid 1 using laser beams to form a detachment plane 4, and introducing an external force into the solid 1 to generate stresses in the solid 1, wherein the external force is strong enough that the stresses cause crack propagation along the detachment plane 4.

[0113] The modifications are generated sequentially in at least one row or line, wherein the modifications 2 generated in a row or line are preferably produced at a distance X and with a height H, so that a crack propagating between two successive modifications, in particular a crack propagating in the crystal lattice direction, whose crack propagation direction is oriented at an angle W to the detachment plane, connects the two modifications. The angle W is preferably between 2° and 6°, particularly 4°. Preferably, the crack propagates from a region below the center of a first modification to a region above the center of a second modification. The essential relationship here is therefore that the size of the modification can or must be changed depending on the distance between the modifications and the angle W.

[0114] Furthermore, this method can also include the step of generating a composite structure by arranging or generating layers and / or components 150 on or above an initially exposed surface of the solid 1, wherein the exposed surface is preferably part of the solid layer to be separated. The modifications for forming the separation plane 4 are particularly preferably generated before the composite structure is created.

[0115] To introduce the external force, for example, an receiving layer 140 can be arranged on an exposed surface of the composite structure or the solid body, analogous to the methods described above.

[0116] Figures 30a to 30c illustrate how the size of the laser-amorphized / phase-transformed damage / modification zone influences the height traversed by the sawtooth crack pattern. Generally, the crack propagates along the crystal planes, i.e., between individual atoms of the crystal. In the modified zone, these distinct planes no longer exist, and the crack therefore comes to a halt.

[0117] By using a preferably high numerical aperture, the damaged zone can be reduced along the beam direction as well as laterally in the focal plane. Since only the threshold intensity needs to be reached, a lower pulse energy is sufficient. If the damaged zone is suitably smaller, the laser modifications can be placed more densely, resulting in a shorter sawtooth pattern and an overall smaller height extent of the modified plane (first image). Conversely, if the damaged zone is larger (higher energy and / or lower numerical aperture – Fig. 30b ) - the increased pressure of the amorphized zone also triggers a larger microcrack, which can be captured (i.e., stopped in a controlled manner) by a larger damage zone at a greater distance.

[0118] Fig. 30cFinally, the danger becomes apparent if the damaged area is not sufficiently large and the laser modification triggers excessively long cracks. Firstly, the cracks will extend too far – meaning the height difference caused by the cracks will be greater than desired – and secondly, the cracks will propagate beneath the other damaged areas and will not be stopped by the amorphized material. This then leads to further material loss, as all the cracked material layers must be removed for the final product or for further laser processing.

[0119] Fig. 31Figure 1 shows a schematic snapshot from a further process. This further process is preferably used for separating at least one solid layer from at least one solid body 1, in particular from a wafer to an ingot, or for thinning a wafer. The further process preferably comprises at least the following steps: generating a plurality of modifications 2 inside the solid body 1 using laser beams to form a detachment plane 4, and introducing an external force into the solid body 1 to generate stresses in the solid body 1, wherein the external force is strong enough that the stresses cause crack propagation along the detachment plane 4.

[0120] In a first step, the modifications are generated on a line 103, preferably at equal intervals. It is also conceivable that a plurality of these lines generated in the first step are produced. These first lines are particularly preferably generated parallel to the crack propagation direction and preferably straight or arc-shaped, especially in the same plane. After the generation of these first lines, second lines 105 are preferably generated to trigger and / or drive preferably subcritical cracks. These second lines are also preferably generated straight. The second lines are particularly preferably inclined relative to the first lines, especially orthogonally oriented. The second lines preferably extend in the same plane as the first lines or particularly preferably in a plane parallel to the plane in which the first lines extend.Subsequently, third lines are preferably created to connect the subcritical cracks.

[0121] This method is particularly useful for separating solid layers from a solid consisting of or containing SiC.

[0122] Furthermore, the modifications can be generated successively in at least one row or line, wherein the modifications 2 generated in a row or line are preferably produced at a distance X and with a height H, so that a crack propagating between two successive modifications, in particular a crack propagating in the crystal lattice direction, whose crack propagation direction is oriented at an angle W to the detachment plane, connects the two modifications. The angle W is preferably between 2° and 6°, particularly 4°. Preferably, the crack propagates from a region below the center of a first modification to a region above the center of a second modification. The essential relationship here is therefore that the size of the modification can or must be changed depending on the distance between the modifications and the angle W.

[0123] Furthermore, this method can also include the step of generating a composite structure by arranging or generating layers and / or components 150 on or above an initially exposed surface of the solid 1, wherein the exposed surface is preferably part of the solid layer to be separated. The modifications for forming the separation plane are particularly preferably generated before the composite structure is created.

[0124] To introduce the external force, for example, an receiving layer 140 can be arranged on an exposed surface of the composite structure or the solid body, analogous to the methods described above.

[0125] Thus, in the further laser process according to the invention, lines parallel to the crack propagation direction (preferably called transverse lines) are preferably generated on SiC (but also other materials) in order to first define a plane for the preferred crack initiation (crack initialization) before longitudinal lines drive the cracks. Here, the cracks are first initialized transversely, then longitudinally, before a final step places lines between the longitudinal lines of the second step to initiate the cracks across the entire surface. This enables shorter crack propagation paths, which minimizes the final surface roughness.

[0126] Example image for transverse lines (with the sawtooth) and crack initiation lines (on the wave crests of the sawtooth). Reference symbol list

[0127] 1 Solid 2 Modification 3 Location of modification generation 4 Crack propagation area 6 Solid fraction 8 Laser impact device 9 Lens 10 Laser beams 11 Modified laser beams 12 Depression 14 Solid layer 16 Surface of the solid layer 17 Beam surface 18 Stress generation layer or receiving layer 140 19 Adhesive or further stress generation layer 24 Local property change of the solid (e.g., transmission) 26 Trench 28 Frame 30 Crossover point of the reflected beams 32 Flushing 34 Coating 300 Chuck Longitudinal direction of the solid

Claims

1. Method for forming at least one solid-body layer (14) which is curved or bent at least in sections, wherein the method at least comprises: moving the solid body (1) relative to a laser application device (8), successively forming laser beams (10) by means of the laser application device (8) for forming at least one layer with modifications (2) inside the solid body (1), wherein the at least one layer with the modifications (2) is spaced apart from an irradiation surface (17) via which the laser beams (10) penetrate into the solid body (1), wherein a crack guiding region (4) for guiding a crack for separating the solid-body layer (14) from the solid body (1) is predetermined by the modifications (2), characterized in that laser parameters are selected such that an extension of the at least one layer with the modifications in a beam direction of the laser beams (10) is so large that the modifications (2) cause a pressure increase in the solid body (1) and the solid-body layer (14) is separated from the solid body (1) by a crack propagation as a result of the pressure increase along the crack guiding region (4), wherein the laser parameters comprise a numerical aperture, a pulse length and a pulse energy of the laser beams, wherein the irradiation surface (17) of the solid body (1) is part of the solid-body layer (14) after the separation of the solid-body layer (14), wherein at least a portion of the modifications (2) is separated from the solid body (1) and forms a part of the solid-body layer (14), and wherein the solid-body layer (14) is converted into a bent or curved shape as a result of the modifications (2) forming a part of the solid-body layer (14), so that a surface portion of the solid-body layer (14) resulting from the crack guiding region (4) is thus convexly shaped at least in sections.

2. Method according to claim 1, characterized in that the laser beams (10) penetrate into the solid body (1) in a longitudinal direction of the solid body (1) or at an angle of up to 60° inclined to the longitudinal direction (L) of the solid body (1) via the, in particular planar, irradiation surface (17), which is preferably part of the solid-body layer (14), and the crack guiding region (4) is formed from a plurality of layers of modifications (2), wherein the layers are formed spaced apart or offset from one another in the longitudinal direction (L), and / or at least a plurality of the modifications (2) have an extent in the longitudinal direction (L) which is between 1 and 50 µm, and / or the laser beams (10) for forming the modifications (2) are introduced into the solid body (1) such that the numerical aperture is less than 0.8.