Method for temporarily connecting and detaching a product substrate and a carrier substrate, and device for such a method

Abstract

The invention relates to a method for temporarily connecting a product substrate (1) and a carrier substrate (2), comprising: - providing the product substrate (1) and the carrier substrate (2), - connecting the product substrate (1) and the carrier substrate (2) to one another to form a bonding layer (4) between the product substrate (1) and the carrier substrate (2) in a connected state, - separating the product substrate (1) from the carrier substrate (2), the product substrate (1) and the carrier substrate (2) being separated: - in a primary region (11) with gas development in a first separation step, and - in a secondary region (12), surrounding the primary region (11), in particular in a closed and / or gastight manner, in a second separation step.

Description

[0001] ■□I

[0002] EV Group E. Thallner GmbH MSP Ref: 46100 PT-WO PM / SK

[0003] Method for temporarily joining and separating a product substrate and a support substrate and a device for such a method

[0004] The present invention relates to a method for temporarily bonding a product substrate and a support substrate, and to a device for carrying out such a method. The product substrate can be bonded to the support substrate for various processing methods and subsequently separated from it, for example during debonding or layer transfer.

[0005] In the semiconductor industry, it is becoming increasingly important to develop processes and devices that can both generate high adhesion between surfaces and, if necessary, prevent this adhesion. Such substrate processing processes and devices are primarily used for two different techniques: layer transfer and debonding of substrates after temporary bonding.

[0006] Layer transfer involves transferring, in particular, inorganic, usually extremely thin layers that are mechanically unstable on their own between self-supporting substrates. The layers to be transferred are created, for example, on special growth substrates, but then must be transferred to another substrate, especially a product substrate, in order to fulfill their functional properties. In the rarest of cases are the growth substrate and the product substrate identical.

[0007] In the state of the art, so-called release layers are used.

[0008] Separation layers are used to selectively and locally separate the wear layer from a support substrate. To date, an organic polymer layer is used for this purpose, acting as a bonding layer. In the temporary bonding and debonding of substrates, two substrates—a product substrate and a support substrate—are first joined together so that the product substrate, which often lacks sufficient internal stability for the process due to its lower thickness or strength, can be processed. It is supported by the second substrate, i.e., the support substrate. Polymers are used to create a so-called temporary bond, i.e., a bond that can be removed without damage. For this, at least one polymer is applied to a substrate using a process, particularly a centrifugal coating process.The carrier substrate is primarily coated with the polymer (polymeric release layer), and the carrier substrate is bonded to the product substrate. After bonding the product substrate to the carrier substrate, the product substrate is further processed so that it can be separated from the carrier substrate in a later process step. Various methods for separating the carrier substrate exist in the prior art.

[0009] In the early years, the substrate was usually coated with the polymer across its entire surface. This resulted in very good adhesion between the substrate and the product substrate. If the product substrate has raised areas, such as solder balls (bumps), dies, or chips, these can be embedded in the polymer, provided the polymer layer acting as a separating layer has a sufficient thickness. The disadvantage of this method becomes apparent when separating the substrate from the product substrate. Separation requires action on the polymer layer across its entire surface, making complete separation more complex and often requiring multiple processing steps. The adhesive strength of the bonding layer can be reduced by a chemical that penetrates the layer laterally or through the substrate to allow for mechanical separation. This chemical action on the bonding layer preferably occurs at elevated temperatures.Separating a full-surface polymer layer also requires more energy (e.g., heat, ultrasound) and / or solvents for removing the adhesive materials, resulting in higher costs. In this process, the separation itself is preferably controlled by detachment forces perpendicular to the substrate surface.

[0010] An alternative development is the so-called "slide-off debonding." This process uses a device that fixes both the product substrate and the carrier substrate across their entire surface with individual substrate holders. The two substrate holders are then moved parallel to each other, causing the substrates to shear apart along the bonding surface. This requires heating the intervening polymer layer. The polymer layer softens as a result and loses its adhesive strength. The advantage of this method is that it eliminates the need for solvents and ultrasound. The disadvantage is that the debonding process still requires an elevated temperature.

[0011] Alternative debonding methods do not use elevated temperatures and rely solely on solvents that attack the polymer from the periphery of both substrates. The action of the solvent can be accelerated by ultrasound. The dissolved polymer is preferably transported away from the substrate stack by circulating the solvent and is preferably continuously removed from the solvent bath. Alternatively, the solvent can be sprayed laterally onto the polymer and removed by a solvent jet.

[0012] A further development of the solvent process is the so-called ZoneBond™ method, described in publication WO 2009 094 558 A2. This is a process in which a support substrate must be specially prepared. The support substrate consists of two distinct zones. The central zone, which covers the largest area, is coated with an anti-stick layer and exhibits low adhesion to all types of polymers. The second zone surrounds the first zone in a circular fashion, usually as a closed ring, and extends to the edge of the support substrate.

[0013] The ring thickness is only a few millimeters. This very small area is sufficient to fully fix the polymer and thus the product substrate bonded to the carrier substrate. The advantage is that, for debonding, only the polymer from the second, more easily accessible zone needs to be removed to allow the carrier substrate to be separated from the product substrate.

[0014] Another possibility is the use of photosensitive release layers. These are usually applied to a support substrate, particularly a transparent one. The support substrate is then coated with a polymer layer and bonded to form a product substrate. After the product substrate has been fully processed, the release layer can be separated from the product substrate by irradiating the support substrate. Such processes are described, for example, in US 2015 0 035 554 A1, US 2016 0 133 486, and US 2016 0 133495 A1. A special embodiment of this process, in which the release layer is applied not to the support substrate but to the product substrate, can be found in WO 2017 076 682 A1.

[0015] The orientation of the product substrate can also be changed. Since the carrier substrates are changed in this process, it is also referred to as a carrier flip. For example, WO 2011 120 537 A1 describes a process in which a product substrate is fixed to a first carrier by means of a polymer layer and then processed. After processing, in particular re-thinning, the product substrate is very thin and is bonded to a second carrier substrate for further use of the other side. The transfer of the product substrate takes place across two polymer layers. The first carrier substrate must be separated from the product substrate after the product substrate has been fixed to the second carrier substrate.

[0016] Another possibility is the use of thin, especially inorganic, separation layers. These are applied, particularly as a layer sequence, to a support substrate, which is transparent, especially to separation radiation. The support substrate is then bonded to a product substrate. After the product substrate has been fully processed, the separation layer can be irradiated through the support substrate, causing it to detach from the product substrate. Such processes are described, for example, in publication WO 2023 179 868 A1.

[0017] All the methods mentioned are mostly based on the fact that some part, but especially a layer, is influenced in such a way that the adhesion to other parts is reduced, in particular completely eliminated. In other words: The adhesive properties of the separating layer are reduced.

[0018] The devices and methods described in the prior art thus describe a separation process that uses layers as separating layers or combines organic bonding layers with inorganic separating layers. However, the use of polymers as bonding layers and / or separating layers that temporarily connect two substrates has several disadvantages. Polymers are long-chain molecules whose main component is usually carbon. Organic materials are often disadvantageous and undesirable in the semiconductor industry because they can contaminate a cleanroom environment, especially the devices in which the substrates are processed.Furthermore, polymers have the disadvantage that they only maintain their adhesive properties up to a relatively low temperature, which is an advantage for debonding but a disadvantage if the product substrate needs to be processed on the support substrate at high temperatures. In addition to the low temperature resistance of organic layers, organic release layers often need to be applied relatively thickly to provide adequate adhesion.

[0019] It is therefore an object of the present invention to disclose a method and an apparatus for processing a substrate which eliminate at least some, and preferably completely, the disadvantages listed in the prior art. In particular, it is an object of the invention to disclose an improved method and an improved apparatus for separating substrates or for separating layers on a substrate. Furthermore, it is an object of the invention to disclose an alternative method and an alternative apparatus with which a temporarily bonded arrangement of product substrate and support substrate can be processed, in particular separated (debonded), simply and cleanly. Alternatively, a layer should be transferable using the apparatus and the method, in particular without generating additional particles.

[0020] It is further an object of the present invention to provide a method for processing a substrate that can be carried out with minimal contamination. It is also an object of the present invention to provide an arrangement of a product substrate and a support substrate which, as a substrate stack, withstands high processing temperatures without spontaneous separation or delamination and remains separable without particles. Furthermore, it is an object of the present invention to provide an arrangement of a product substrate and a support substrate from which, in particular, a support substrate or a product substrate can be separated particularly easily and without contaminating the apparatus.

[0021] The present invention solves the problem with a method according to claim 1 and a device according to claim 15. Advantageous embodiments of the invention are specified in the dependent claims. The scope of the invention also includes all combinations of at least two features specified in the description, the claims, and / or the drawings. Where specified value ranges are given, values ​​lying within the stated limits are also considered limit values ​​and may be claimed in any combination. According to a first aspect of the present invention, a method for temporarily bonding a product substrate and a carrier substrate is provided, comprising:

[0022] Provision of the product substrate and the carrier substrate,

[0023] Connecting the product substrate and the support substrate together, forming a bonding layer between the product substrate and the support substrate in a connected state,

[0024] Separating the product substrate from the carrier substrate,

[0025] wherein the product substrate and the carrier substrate

[0026] in a primary area with gas evolution in a first separation step and in a secondary area that surrounds the primary area, in particular closed and / or gas-tight, in a second separation step

[0027] are separated, in particular in such a way that a gas produced during gas evolution is stored in the primary area between the product substrate and the support substrate.

[0028] In contrast to the prior art, this method provides for a separation step in the primary region involving gas evolution, followed by a second separation step in the secondary region. The second separation step preferably occurs after the first. Preferably, the second separation step follows the first, and in particular, immediately. The secondary region preferably encloses the primary region, most preferably in a closed or gas-tight manner, particularly to facilitate pressure build-up between the support substrate and the product substrate. This pressure build-up keeps the support substrate and the product substrate separated after the first separation step in the primary region. This allows the gas generated during gas evolution in the first separation step to be trapped between the product substrate and the support substrate.The gases released during the dissolution of the bonding layer can be stored between the product substrate and the support substrate. This approach has proven to enable essentially contamination-free debonding, preventing material components of the bonding layer released during separation from reacting again with the support substrate or the product substrate, and / or from entering and contaminating the temporary bonding device. Separation can occur within the bonding layer itself or in a separate separation layer. Primary and secondary areas define planar sections between the product substrate and the support substrate that correspond to local sections of the bonding layer. These sections do not necessarily have to be part of the bonding layer itself.

[0029] Preferably, the bonding layer is formed and provided essentially from an inorganic material, wherein an inorganic gas, preferably with gas evolution, more preferably sublimation, enables at least local separation of the bonding layer or a separating layer. In other words, the bonding layer does not contain any carbon chains and / or carbon groups. It has been shown that such bonding layers are significantly more temperature-resistant than those containing a polymer. This allows, for example, treatments to be carried out on the product substrate in the bonded state, even if high temperatures occur or are required during the treatment. Furthermore, the use of an inorganic, i.e., organic-carbon-free, bonding layer allows for contamination-free operation. Preferably, the bonding layers are inorganic.Especially for such temporary bonding, it is advantageous to divide the separation process into a first and a second step. The product substrate and the carrier substrate preferably form a substrate stack together with the bonding layer.

[0030] Contamination-free operation is preferably achieved by separating the substrate stack in two steps, in particular with the two different separation steps.

[0031] Preferably, the bonding layer is essentially polymer-free. In a further embodiment of the bonding layers, it is possible for the bonding layers to be free of organic carbon. It is particularly preferred that the bonding layer is part of a layer system comprising at least one inorganic layer. The inorganic layer can comprise or consist of TiN, SiCN, and / or SiC. The layer system can also include metal layers, nitride layers, and / or ceramic layers.

[0032] The term "essentially polymer-free" means that the polymer content or the organic carbon content in the bonding layer can be less than 5 wt.%, preferably less than 2.5 wt.%, and particularly preferably less than 1.0 wt.%. Furthermore, the product substrate and the support substrate are preferably indirectly bonded to each other. Thus, it is possible to achieve a carbon content of 20% to 30% in the bonding layer, particularly as SiCN, but with little to no organic carbon content.

[0033] Preferably, the primary region is treated with a first means for the first separation step, and the secondary region is treated with a second means different from the first. In particular, the first means is electromagnetic radiation and / or the second means is a mechanical force. This proves to be a particularly effective approach for achieving localized separation in the primary and secondary regions.

[0034] In the primary region, the bonding layer is separated using laser radiation. This process sublimates the bonding layer, producing gas evolution and separating the product substrate from the support substrate, at least locally. It has been found that the local high temperature exposure not only generates gases but also produces liquid, molten particles or droplets. However, since the substrate stack is only separated in the primary region, the solidifying particles are deposited either on the product substrate or on the support substrate, as they cannot escape from the resulting cavity. This reduces the particle load on the device.

[0035] Alternatively, the secondary layer can be separated by means of cold cracking, so that the differences in thermal expansion of the materials of the respective substrates and the bonding layer under the influence of cold can be exploited.

[0036] Alternatively, the secondary area is separated at a later time, in particular in a different area, chamber or module, using the same method as in the area, chamber or module in which the primary area is or was treated.

[0037] Preferably, the primary region is more than twice, preferably more than five times, and most preferably more than ten times larger than the secondary region. This allows, for example, the comparatively demanding separation of the secondary region to be limited to the smallest possible area. In particular, the energy expenditure for the secondary region can be reduced, so that separation in the second step also requires less energy. At the same time, it is advantageously possible to provide sufficient space in the primary region so that a gas can expand to the desired or required extent.

[0038] In other words, the primary area can extend over a predominant area of ​​the substrate surface, and the secondary area for mechanical separation can be less than one millimeter wide, preferably less than 0.5 mm wide, particularly preferably less than 0.2 mm wide, and most preferably less than 0.1 mm wide at the edge of the substrate stack, especially in the so-called edge exclusion zone. The increased alignment accuracy of the substrate stack relative to the laser beam or the substrate holder is correspondingly necessary to form a mechanically stable, complete secondary area that can be separated mechanically without substrate breakage. This is because the alignment accuracy of a purely mechanical alignment with the higher tolerance would produce a secondary area that is too wide, which would not necessarily guarantee damage-free separation of the substrate stack.Furthermore, increased alignment accuracy is necessary to preferably design the secondary area to be a few tenths of a millimeter wide. This width is preferably less than the tolerance of the mechanical alignment accuracy, which is a few hundred micrometers.

[0039] Preferably, the primary area is circular and the secondary area is annular. This advantageously allows for radially symmetrical separation, which proves to be a particularly homogeneous separation, especially for wafers.

[0040] The primary area is defined within a diameter. Preferably, the diameter is set at the boundary of the so-called "edge exclusion zone," but it can deviate from this technologically, so that the secondary area can be separated, particularly mechanically.

[0041] The secondary area extends directly outside the primary area, as a closed circular ring, preferably to the substrate edge or the edge of the substrate stack.

[0042] Preferably, a separating layer, in particular a separating layer decoupled from the bonding layer, is provided between the carrier substrate and the product substrate in the bonded state, wherein the product substrate is separated from the carrier substrate by separation along the separating layer. This separation advantageously also creates a functional separation between the bonding layer and the separating layer, allowing for optimization such that the separation does not occur under undesired conditions, but is exclusively controlled and preferably solely initiated by an intentional separation process. Furthermore, the separation can be carried out in such a way that neither the bonding layer nor the product substrate is affected by the separation process.

[0043] Analogous to the definitions of the primary region of the bonding layer and the secondary region of the bonding layer, the definitions of the primary / secondary regions apply to the separation layer, so that a temporarily joined substrate stack can be separated at the separation layer.

[0044] The partial separation of the substrate stack in the separation layer in the primary region using laser radiation and a mechanical separation of the secondary region, possibly carried out separately in time and space, is preferably provided. This enables a reduction in the particle load of the device.

[0045] According to a preferred embodiment, the separating layer is provided as part of a layer system arranged between the support substrate and the product substrate. This allows, for example, a portion of the layer system to be specifically adapted to create a controlled decoupling between the separating layer and the bonding layer. For example, at least two layers, preferably at least three layers, are arranged between the separating layer and the bonding layer. This makes it possible, for example, to create a separation between the separating layer and the bonding layer by means of the layer(s) arranged between them, such that a separation process acting on the separating layer does not affect the bonding layer.In other words, the one or more layers in the layer system, arranged between the bonding layer and the separation layer, protect the bonding layer during the separation process, thus preventing separation until the separation layer is initiated. Through the first stage of the two-stage separation, the primary region in the substrate stack may already have been separated into a product and a support substrate; however, the primary region between the support and product substrates remains hermetically sealed from the external environment. Accordingly, the molten particles generated during separation in the primary region between the support and product substrates can resolidify and do not increase the external particle load. A further advantage is the lack of gas exchange: during sublimation with nitrogen generation in the separation layer, the bonding layer, or rather the product substrate, can be exposed to nitrogen.The product substrate or carrier substrate in the primary area does not oxidize and remains in an inert gas atmosphere until complete separation.

[0046] A further advantage is that the unseparated secondary area of ​​the substrate stack allows for joint handling and, in particular, damage-free transport of the substrate stack before separation.

[0047] Furthermore, the use of the separate separating layer does not modify or affect the product substrate or the bonding layer. Additionally, by arranging one or more layers between the separating layer and the bonding layer, it is possible for one or more layers to become part of the separated product substrate after separation. This allows, for example, the targeted transfer of a layer formed between the separating layer and the bonding layer from the support substrate to the product substrate. In other words, it is particularly preferred that a layer system be implemented between a product substrate and a support substrate, wherein the layer system, especially with a separate separating layer, serves to ensure separability in a subsequent process.

[0048] Preferably, the product substrate and the carrier substrate are transported in a connected state after the first separation step, particularly over a distance greater than 1 m, preferably over more than 10 m, and most preferably over more than 100 m. This two-stage separation allows for better utilization of substrate processing facilities worldwide, enabling the transport of partially separated substrates between production sites or fabs. By delaying the separation of the substrate stack at the processing location, manufacturers can optimize their resource utilization and, in particular, avoid oxidation or contamination of the separated substrate stacks.

[0049] The layer system on the support substrate is particularly preferably entirely inorganic, thus enabling higher temperatures and contamination-free processing than with organic, especially polymer-containing, layers. Higher temperatures, in particular, allow the particles of the separating layer to melt and solidify during separation without leaving the primary area.

[0050] In particular, the arrangement consisting of product substrate, support substrate, and layer system is entirely inorganic, i.e., essentially free of organic carbon. In an alternative embodiment of the layer system on the support substrate, a transfer layer, i.e., a layer to be transferred, of the layer system—preferably encapsulated—can contain carbon-containing layers, so that the transfer layer does not come into contact with the external atmosphere or with the plant.

[0051] A substrate, and in particular a product substrate, is any type of transferable component, especially a wafer or plate, but also small, individual components such as chips, dies, and chiplets of varying complexity, shape, and functionality, such as LEDs, MEMS, and optical applications, especially optical layers, etc. For clarity and simplicity, substrates are always depicted as wafers in the following text and figures. Support substrates are almost always wafers, whereas product substrates can be either wafers or, very often, individual components such as chips, LEDs, dies, or MEMs. Panels, especially rectangular panels, are also considered substrates.

[0052] The substrates, whether product substrates or support substrates, can have any shape, but are preferably circular. The diameter of the substrates is standardized, particularly in industry. For wafers, the industry-standard diameters are 1 inch, 2 inches, 3 inches, 4 inches, 5 inches, 6 inches, 8 inches, 12 inches, and 18 inches. If one of the substrates needs to be irradiated by an electromagnetic wave, the substrate must have a sufficiently high transparency for the corresponding electromagnetic wavelength.

[0053] The materials of the substrates, in particular the support substrate and the product substrate, preferably comprise at least one of the following materials or material classes:

[0054] - a semiconductor, especially silicon,

[0055] - a nitride, especially GaN,

[0056] - an oxide, especially sapphire, @ - a carbide, especially SiC,

[0057] - a metal and / or

[0058] - a composite material.

[0059] In a further preferred embodiment, the distance between the bonding layer and the separating layer is greater than 250 nm, preferably greater than 500 nm, and particularly preferably greater than 800 nm when connected. This spatial separation provides a sufficiently large distance to allow, or at least simplify, separate processing of the bonding layer and the separating layer. This allows, for example, manipulation of only the separating layer without affecting the bonding layer.

[0060] Preferably, the bonding layer, in particular the dielectric bonding layer, is formed during the joining process by a dielectric bonding surface on the product substrate and / or a further dielectric bonding surface on the layer system, wherein the dielectric bonding surface and / or the further dielectric bonding surface comprises a ceramic, in particular a technical or functional ceramic, especially a silicon-based ceramic. Here, ceramics are defined as inorganic, non-metallic, amorphous layers. For this purpose, the product substrate has an inorganic, dielectric bonding surface prior to the formation of the bonding layer, which is suitable for fusion bonding or hybrid bonding. This surface can be realized by a corresponding coating. Preferably, this dielectric coating, the layer thickness of the dielectric, or the bonding surface, etc., has a specific thickness.The further bonding surface has a layer thickness between 0 nm and 500 nm. Preferably, the dielectric bonding surface or the further bonding surface comprises silicon oxide (Si₂O₃) and / or silicon carbonitride (Si₄N₄) and / or silicon nitride (Si₄N₄) and / or aluminum nitride (Al₂O₃) and / or silicon carbide (Si₂O₃) and / or silicon carbide (SiC), generally, i.e., a ceramic bonding layer. The bonding surface and / or the further bonding surface is thus designed as an inorganic dielectric, in particular as a ceramic bonding surface. In particularly preferred embodiments, the ceramic bonding surface can have an amorphous or at least partially amorphous or amorphized bonding surface. Preferably, the surface quality of the bonding surface or the further bonding surface is designed for fusion bonding, although other bonding methods are also possible.

[0061] The following numerical ranges for roughness are to be understood as values ​​for the mean roughness, the quadratic roughness, or the mean roughness depth. The roughness is less than 100 µm, preferably less than 10 µm, even more preferably less than 1 µm, most preferably less than 100 nm, and most preferably less than 10 nm. The bonding surface of the support substrate should have a good conformity with the product substrate surface. Preferably, the bonding surfaces are planar, i.e., flat. For the fabrication of the substrate stack, the product substrate requires a bonding surface which, for a fusion bond, should preferably be made of the same material and preferably have the same surface quality as the other bonding surfaces of the layer system. However, it is possible to successfully perform fusion bonds with the corresponding surface quality using selected material pairings.

[0062] In other words, an inorganic, dielectric layer with a thickness between 0 nm and 500 nm is applied to the product substrate or formed from the substrate material itself. This layer is composed of a nitride or, preferably, an oxide, most preferably silicon oxide. The additional bonding surfaces listed for the support substrate or layer system can also be used on the product substrate, allowing the bonding surface and the additional bonding surface to fuse together during fusion bonding via heat treatment.

[0063] In particular, it is provided that the product substrate and the support substrate, in their bonded state, are exposed in a treatment step to a temperature greater than 600 °C, preferably above 800 °C, or most preferably above 1100 °C. The temperature-resistant bonding layer advantageously allows treatments to be carried out on the product substrate without risking spontaneous and unintended separation between the support substrate and the product substrate across the bonding layer. For the successful processing of the arrangement, which comprises at least the product substrate and the support substrate with the described layer sequence or layer system before the substrate stack is separated, it is advantageous that the substrate stack or the arrangement exhibits high thermal stability with respect to maintaining separability. For this purpose, it is provided that the substrate stack or the arrangement...The arrangement withstands process temperatures above 400°C, preferably above 500°C, particularly preferably above 600°C, most preferably above 700°C, in the optimal case above 750°C, in the ideal case above 900°C without damage and can be separated in a subsequent separation process at the separation layer without particle generation, without spontaneous delamination or separation or without spontaneous diffusion processes or undesirable chemical changes.

[0064] In preferred embodiments, an additional diffusion barrier layer can be used, particularly at higher process temperatures. Accordingly, high temperatures of over 400°C, preferably over 500°C, particularly preferably over 600°C, most preferably over 700°C, optimally over 750°C, ideally over 900°C, and in extreme cases over 1100°C can be used in the processing of the product substrate in its bonded state. The arrangement, i.e., the substrate stack, remains stable, so that, in particular, no outgassing of the layer sequence occurs, preventing spontaneous separation of the substrate stack. These high process temperatures, especially with thinned product substrates, open up new possibilities for the design of FEOL (front-end of line) processes. In the following, the terms arrangement and substrate stack are used synonymously.

[0065] A particularly advantageous feature is the targeted, two-stage separation of the substrate stack. After separating the substrate stack within a diameter, i.e., in the primary region, gas formation causes the product and the support substrate to separate. This alters the thermal properties of the substrate stack, which remains attached at the edge, i.e., in the secondary region. Consequently, higher temperatures can be achieved locally without heat dissipation, especially with a thinned product substrate. However, the different thermal expansions after the first separation step must be taken into account.

[0066] In other words, the production of the substrate stack using the layer system and a two-stage, low-particle separation process enables the use of previously unusable processing technologies for the substrate stack, particularly re-thinning and / or processing steps of the thinned product substrate, thus allowing high process temperatures to be achieved. At the same time, the present technology offers all the advantages of temporary bonding and hybrid bonding, which allow the processing of thinned product substrates without the technological limitation of the low temperatures of temporary bonding with polymeric interlayers.@ As a further advantage, the substrate stack with the two-stage separable bonding layer or separation layer enables particle-free separation even at the increased process temperatures, because the heated particles are immediately trapped in the separated space between product substrate and carrier substrate and can condense and solidify locally.

[0067] The physical limit to the stability of the substrate stack is determined on the one hand by outgassing within the layer system and on the other hand by the diffusion processes that are unavoidable at high temperatures. Since these temperature limits are characteristic of the different layer sequences, such limits are particularly well known to materials scientists and physicists.

[0068] A support substrate is generally understood to be any substrate suitable for the deposition of the layer system, in particular an inorganic layer sequence. Preferably, the support substrate is an inorganic substrate, and most preferably a silicon wafer. Silicon wafers are particularly preferred because they are readily available in extremely high quality, at a suitable size, and are well-established in semiconductor fabrication. Furthermore, the infrared transparency of the material allows for laser irradiation of the interface using widely available infrared laser systems. The support substrate serves, in particular, to mechanically stabilize the product substrate when it is bonded to the support substrate, i.e., in the bonded state.

[0069] The use of silicon as a substrate has proven particularly advantageous, as the chemical fingerprint of a silicon substrate is indistinguishable from that of a silicon-based product substrate. Furthermore, the surface finish and flatness of silicon substrates are orders of magnitude better than those of glass substrates. Additionally, glass substrates require organic layers and / or coloring for handling, which negatively impacts the chemical fingerprint.

[0070] Preferably, the separating layer is made of an inorganic material, in particular a dielectric. Inorganic separating and bonding layers typically have a tailored thermal conductivity, thus enabling better separation of the substrate stack. In the case of silicon oxide as a separating or bonding layer, the thermal conductivity is low, so that heat can be concentrated in the layer, which can promote separation, especially local separation of the substrate stack. Furthermore, polymer-based organic separating and bonding layers can adversely contaminate the equipment and negatively affect its quality. With the closed edge, i.e., the secondary area, the substrate stacks can store all particles generated during separation without contaminating the device.Using inorganic materials for the separation and bonding layers reduces contamination of the equipment. The high adhesive properties of these inorganic separation and / or bonding layers also allow for a thinner separation layer and thus a thinner overall layer structure in the substrate stack being processed. This results in improved flatness of the substrate stack and reduces the amount of material to be processed, thereby saving on potential contamination. A further advantage of this method for separating carrier substrates is the lower energy input required for processing the substrate stack and the correspondingly lower heat load. This allows even temperature-sensitive product substrates within the substrate stack to be processed without damage.

[0071] In other words, an inorganic, dielectric layer with a thickness between 0 nm and 500 nm is deposited on the support substrate as a separating layer. This layer is formed from an oxide, particularly preferably titanium oxide, or preferably a nitride, e.g., titanium nitride. In the two-stage separation process, the separating layer or bonding layer can first be locally exposed to electromagnetic radiation, particularly laser radiation, within a certain diameter across the support substrate. This causes the bonding layer or separating layer to sublimate, generating gas, and locally separate the substrate stack in the primary region without completely separating the substrate stack from one another. Thus, the generated particles remain within the separated interface, preventing contamination of the device.

[0072] Preferably, after the product substrate is detached from the support substrate, the bonding layer is at least temporarily or permanently part of the detached product substrate. This makes it possible, for example, to transfer a layer integrated into the layer system, the transfer layer, onto the product substrate.

[0073] In particular, it is intended that at least part of the layer system is removed from the detached product substrate. This can, for example, serve a cleaning purpose or a targeted thinning process to expose the back side of the product substrate or another layer on the product substrate. Preferably, residues of the bonding layer are removed.

[0074] Preferably, the separating layer is dissolved by irradiation, preferably laser radiation, in special cases with laser light, by means of a solvent, and / or by mechanical force. Two-stage separation by laser radiation followed by mechanical force is particularly preferred, as this allows for localized processing. In particular, the method for separating the substrate stack is used such that the holding force between the support substrate and the product substrate is reduced within a certain diameter, but the substrate stack remains self-supporting, i.e., no immediate separation occurs. Separation only takes place in a further process step, in particular with the aid of an applied mechanical force.

[0075] Another aspect of the inorganic layer sequence on the support substrate is the use of a separate separating layer and at least one absorption and / or reflection layer, which, during the separation of the substrate stack, shields the product substrate completely from the laser radiation and thus protects it from damage by the laser radiation. Preferably, the absorption layer and reflection layer are arranged on opposite sides of the separating layer on the support substrate.

[0076] In particular, the process involves separating a product substrate from the carrier substrate, with at least the following steps:

[0077] i) Providing an arrangement in which the product substrate and the support substrate are connected to each other via a bonding layer and a separating layer,

[0078] ii) Alignment of the substrate stack, in particular the edge of the substrate stack on the substrate holder, into a predetermined position in order to position the irradiation unit within a diameter of the substrate stack,

[0079] iii) Irradiating the separating layer with electromagnetic radiation, in particular with a local laser beam, and

[0080] iv) Separation of the product substrate from the support substrate at the separation layer within a diameter for separation in the primary area, @ v) Cooling of the partially sealed substrate stack separated within a diameter,

[0081] vi) Complete separation of the product substrate from the carrier substrate by mechanical means, in particular with at least one blade outside a diameter, in the secondary area

[0082] wherein the separating layer is preferably a dielectric,

[0083] wherein the layer system includes at least one separating layer and at least one further inorganic layer to protect the product substrate and

[0084] The arrangement exhibits high thermal stability up to over 700°C. In a very specific, less preferred embodiment, the separating layer and the protective layer can be made of the same material.

[0085] In other words, an arrangement, i.e., a substrate stack, comprising at least the support substrate and the product substrate, is treated at the interface by targeted laser treatment such that the bond between the product substrate and the support substrate is severed within a diameter, i.e., in the primary region, and then completely separated mechanically in a further separation step, i.e., in the secondary region. Preferably, in contrast to the prior art, an interface made of inorganic material, preferably a dielectric, is subjected to laser beams using a laser, thus reducing the adhesive properties of this inorganic interface without completely separating the substrate stack and allowing hot gases and / or particles to enter the device. The laser beams reach the interface through the support substrate.By releasing the carrier substrate through action on the inorganic release layer, a completely inorganic layer structure of the substrate stack is possible. This allows for the advantageous elimination of the use of organic release or bonding layers in various substrate processing methods, particularly in debonding and substrate transfer. The higher temperature resistance of inorganic materials enables more flexible process design, and the near-complete absence of particles ensures the purity of the device and the product substrate, allowing critical processing steps to be carried out without intermediate cleaning. In other words, the substrate stack is arranged so that the substrate stack is irradiated for separation up to the substrate stack contour, or an edge, i.e.,up to the secondary area, preferably in the form of a full-circumference border, which is thick enough to avoid contamination (to be kept within a diameter in the separated substrate stack area, i.e. in the primary area) and thin enough to be overcome during the separation in the second step without damaging the substrate stack.

[0086] In another, less preferred embodiment, the substrate stack is arranged such that the substrate stack is irradiated for separation up to the substrate stack contour, and the primary area extends to the substrate stack contour at least at one or a few points. In this way, the secondary area cannot form a complete border, but in special cases, a controlled release of excess pressure can be achieved so that the substrate stack is not damaged.

[0087] To perform the first separation step of the substrate stack using electromagnetic radiation, in particular laser radiation, within a certain diameter, the radiation source and the substrate must be aligned with sufficient accuracy relative to each other.

[0088] This alignment is achieved indirectly, such that the substrate stack, preferably the substrate stack edge, or possibly the substrate stack edge with a notch or a flat, is positioned relative to positioning pins on the substrate holder, particularly by means of a mechanical robot arm, and thus placed into position. This establishes a relative positioning between the radiation source and the substrate stack, although the positioning between the positioning pins and the substrate on the substrate holder is established. Details of the flat or notch adjustment are known to those skilled in the art.

[0089] In another embodiment, an edge detection of the substrate stack can be recorded using a camera system and, in a control loop with the robot as a movement device of the substrate stack, a specific, fixed position, which is known to the person skilled in the art, can be aligned.

[0090] In a further embodiment, alignment marks on or in the substrate stack can be used for positioning. This is the usual alignment task between substrates and substrate holder, which is known to those skilled in the art. The alignment accuracy between the substrate stack and the radiation source is less than 100 micrometers, preferably less than 50 micrometers, particularly preferably less than 20 micrometers, and most preferably less than 5 micrometers.

[0091] The laser unit preferably acts in a targeted manner and with precisely calibrated parameters on the separating layer or bonding layer in different, preferably regularly defective, areas. The inorganic separating layer, which is preferably opaque to the laser radiation from the laser unit, absorbs the laser radiation, thus locally reducing the adhesive properties and / or the stability of the separating layer, preferably with gas formation within the separating layer to further reduce the adhesive properties. However, the laser unit only acts on the separating layer within a specific diameter, i.e., in the primary region, so that the inherent stability of the substrate stack is maintained and, in particular, a mechanical opening is created in a subsequent separation step. Specifically, due to the high local energy concentration from the laser radiation, lateral cracks are formed in the separating layer, allowing the support substrate to be locally separated from the product substrate.Since the layer sequence on the support substrate is essentially freely configurable, the layer transfer process can also be used to produce multiple layers in successive bonding and transfer onto the product substrate. The inorganic separating layer then allows the support substrate to be selectively separated and, in particular, reused after cleaning and layer build-up.

[0092] A continuous, large-area laser irradiation is also conceivable, preferably with shadow masking for the edge region of the substrate stack, so that the laser radiation does not strike the substrate stack in this edge region. In this case, the laser beam does not move relative to the substrate stack, but is spread out so that it strikes the entire substrate stack. In particular, in such a specific embodiment, the parameters of the laser irradiation, such as the exposure time, must be adjusted accordingly.

[0093] In a preferred embodiment of the method, laser beams emitted by the laser unit during irradiation in step iii) first penetrate the substrate and then strike the separation layer. In other words, the laser beams first pass through the substrate and are subsequently absorbed by the separation layer. The substrate is thus at least partially transparent to the laser radiation. The laser unit can therefore be advantageously arranged on the back side of the substrate, and separation can be carried out flexibly from the back side without placing special demands on the product substrate.

[0094] Furthermore, it is particularly preferred if the layer system comprises a confinement layer and / or a defocusing layer and / or at least one protective layer, in particular a dielectric protective layer, and / or a barrier layer. This allows further advantageous properties to be integrated into the layer system between the product substrate and the support substrate. An antireflective layer is also preferably bonded to the support substrate on the side opposite the layer system.

[0095] Preferably, a protective layer is formed on the further bonding surface in the layer system, i.e., on a side facing away from the product substrate in the bonded state, and particularly preferably directly adjacent to the further bonding surface. The protective layer preferably has a thickness between 0 nm and 500 nm. It is conceivable that the protective layer comprises a metal or metal alloy and / or nitride, particularly preferably titanium nitride. This protective layer or sequence of protective layers protects the product substrate from the laser radiation.

[0096] In a preferred embodiment of the protective layer, the protective layer comprises at least two layers:

[0097] A first protective layer, consisting of a dielectric, preferably an inorganic dielectric, and particularly preferably the same material as the boundary layer, is formed on the separating layer. The thickness of the protective layer correlates with the wavelength of the laser radiation such that destructive interference prevents the laser radiation from passing through the first layer. Layer thicknesses between 0 nm and 800 nm are thus achieved. A second layer, i.e., a second protective layer, is formed as a barrier layer on the first protective layer. Preferably, the second protective layer is formed from a metal, in particular copper or aluminum, which acts as a barrier to the laser beams from a layer thickness of 50 nm. Alternatively, and particularly preferably, the second protective layer is formed from the separating layer material as an absorber.This second protective layer can have a thickness of 0 nm to 500 nm and serves to shield the product substrate from the laser radiation. Furthermore, it is conceivable that the defocusing layer is arranged opposite the protective layer and the subsequent bonding surface in the stacking direction. The defocusing layer is preferably designed as a dielectric layer. The defocusing layer can have a thickness between 0 nm and 800 nm. The thickness of the defocusing layer is preferably designed such that no standing waves of electromagnetic radiation are generated for separating the substrate stack or the arrangement. Adjacent to the defocusing layer, the separating layer is preferably formed, and the separating layer can preferably have a thickness of 0 nm to 200 nm.For example, the separating layer preferably comprises titanium nitride and / or is designed such that the separating layer dissociates with gas evolution in reaction, in particular to infrared irradiation.

[0098] Preferably, a boundary layer or light-focusing layer is formed on the side of the separating layer facing away from the product substrate. This boundary layer or light-focusing layer is also designed as a dielectric layer. The boundary layer is designed such that, in particular by adjusting its thickness, the generation of standing optical waves of electromagnetic radiation for separating the substrate stack is promoted and enabled, allowing an energy maximum to be directed into the separating layer. The thickness of the boundary layer can be between 0 nm and 800 nm. The boundary layer, designed as a dielectric layer, can in particular comprise silicon carbonitride and / or silicon oxide and / or silicon nitride and / or aluminum nitride. Thus, the boundary layer is preferably inorganic.

[0099] Furthermore, it is provided that the carrier substrate for the electromagnetic radiation for separating the substrate stack is at least largely, preferably over 50%, transparent.

[0100] Furthermore, it is conceivable that the support substrate can have an antireflective layer on the surface facing away from the bonding side to separate the substrate stack from electromagnetic radiation. This antireflective layer preferably comprises a dielectric and / or its thickness is between 0 nm and 1000 nm. The support substrate is preferably coated with a dielectric on the side facing away from the bonding side to create the antireflective layer. The support substrate preferably comprises silicon, particularly undoped silicon, to ensure the appropriate infrared transparency. An inorganic dielectric layer, the boundary layer, is formed on the support substrate, preferably consisting of SiC₂ and / or SiCN and / or Al₂O₃ and / or SiC and / or SiCO₃.The layer thickness is preferably set between 0 nm and 800 nm such that the thickness of the boundary layer correlates with the wavelength of the laser beam in constructive interference. This preferably amplifies the laser radiation because a standing wave of the laser radiation can form in the boundary layer. A separating layer, in particular made of TiN and / or TaN and of the materials listed below, is produced on the boundary layer. The layer thickness is between 0 nm and 200 nm, and the thermal resistance is preferably designed to exceed 700 °C, particularly preferably 1000 °C, and most preferably 1200 °C. The separating layer dissociates upon infrared radiation, forming a gas. A defocusing layer, preferably made of an inorganic dielectric, particularly preferably of the same material as the boundary layer, is produced on the separating layer.The thickness of the defocusing layer correlates with the wavelength of the laser radiation such that destructive interference prevents the laser radiation from passing through the layer. Layer thicknesses between 0 nm and 800 nm are thus produced.

[0101] The diffusion barrier layer prevents or at least reduces diffusion between the separation layer and the protective layer, thus enabling successful separation after processing of the substrate stack at high temperatures.

[0102] In other words, a stacked layer arrangement will preferably have the following layer sequence:

[0103] -bonding layer,

[0104] -protective layer,

[0105] -Separation layer, especially for two-stage separation,

[0106] -boundary layer or light-bundling layer,

[0107] - Carrier substrate (bulk),

[0108] -rear antireflective layer. @ Another, particularly preferred layer sequence of a stacked layer arrangement (including substrates) can have the following layers:

[0109] -product substrate,

[0110] -protective layer,

[0111] -Binding layer (originally formed on the product substrate),

[0112] -Binding layer (originally formed on the carrier substrate),

[0113] -Separation layer, especially for two-stage separation,

[0114] -boundary layer or light-bundling layer,

[0115] - Carrier substrate (bulk),

[0116] -rear anti-reflective coating.

[0117] This embodiment is particularly well suited for applications where a protective layer has already been produced on the product substrate.

[0118] Another particularly preferred layer sequence of a stacked layer arrangement (including substrates) can have the following layers:

[0119] -product substrate,

[0120] -bonding layer,

[0121] -Protective layer, especially for two-stage separation,

[0122] -Diffusion barrier layer,

[0123] -Separating layer,

[0124] -boundary layer or light-bundling layer,

[0125] - Carrier substrate (bulk),

[0126] -reverse antireflective layer. @ In further embodiments, the layer sequences described here can be combined; for example, a protective layer can be applied to the product substrate and / or bonding layers can be applied.

[0127] Further objects of the present invention are

[0128] - a product substrate for the inventive method, wherein the product substrate has a dielectric bonding surface for forming the bonding layer,

[0129] - a carrier substrate with a layer system for the inventive method, wherein the layer system has a further dielectric bonding surface for forming the bonding layer as well as

[0130] - a layer system for a method according to the invention, wherein the layer system has a dielectric bonding surface, in particular a polymer-free dielectric bonding layer, and a separating layer for two-stage separation.

[0131] All properties and advantages described for the process can be applied analogously to the product substrate, the carrier substrate and the layer system, and vice versa.

[0132] The invention further relates to a method for producing a layer sequence or layer system on a support substrate for transfer onto a product substrate in a subsequent process. In addition to the method in which the product substrate and support substrate are separated from each other in a two-stage separation process, the transfer method provides a layer to be transferred, i.e., the transfer layer, within the layer system. In the layer system, the layer to be transferred is arranged between the further bonding surface and the separation layer, preferably between the further bonding surface and the protective layer, and particularly preferably directly adjacent to the bonding surface. In the arrangement in which the product substrate is bonded to the support substrate via the bonding layer, the layer to be transferred, or the transfer layer, is then arranged between the bonding layer and the protective layer or the separation layer.

[0133] It is also conceivable to stack and transfer multiple substrates or layers. This can advantageously be achieved via a layer sequence in the layer system on the support substrate during the construction of the transfer layer on the support substrate. In other words, the at least one transfer layer can comprise multiple layers and / or several stacked and / or bonded substrates.

[0134] A further object of the present invention is an arrangement of product substrate and support substrate, i.e., a substrate stack, for a method according to the invention, wherein an inorganic, in particular dielectric, bonding layer is formed between the product substrate and the support substrate. All advantages and properties described for the method can be transferred analogously to the arrangement and vice versa.

[0135] Another object of the present invention is a device for carrying out a method according to the invention. All advantages and properties described for the method can be transferred analogously to the device and vice versa.

[0136] In particular, the device is provided for separating an arrangement according to the invention and has, for example, the following features:

[0137] a) a provisioning unit for providing the substrate stack,

[0138] b) a laser unit for irradiating a separating layer arranged in the substrate stack within a diameter or in the primary region, wherein the substrate stack can be separated into a support substrate and a product substrate by irradiating the separating layer, wherein the separating layer is inorganic, wherein the separating layer is a dielectric, wherein the separating layer can be separated with gas evolution, wherein the substrate stack, in particular the secondary region, remains completely gas-tight and does not release any particles generated during separation into the environment, and a layer system in the arrangement includes further layers for concentrating the radiation and for shielding the radiation from the product substrate, i.e., at least one boundary layer and / or at least one defocusing layer.

[0139] c) a mechanical separation unit for the complete separation of the substrate stack into a carrier substrate and a product substrate, whereby few additional particles are generated by the mechanical separation.

[0140] In particular, the device is designed for arrangements or substrate stacks that can be subjected to high-temperature processes without spontaneous separation. Consequently, the device itself is designed to withstand temperatures above 700 °C, although these high temperatures primarily occur during the processing of the substrate stack and not during the separation of the substrate stack into a support substrate and a product substrate. This applies especially to the corresponding holding devices used to hold and / or transport the arrangements.

[0141] The device is also preferably suited to reducing the adhesive properties of the separating layer and thus at least partially separating the support substrate and the product substrate from each other. The laser unit and the materials of the support substrate and the separating layer are specifically matched. In particular, the use of infrared radiation in conjunction with the preferred material combinations has proven advantageous.

[0142] In a further embodiment, the first separation of the substrate stack within the primary area is carried out in a first device, and the second separation in the secondary area is carried out in a second device, so that the substrate stack is stable and transportable without external contamination of the separated substrate sides of the product substrate and the support substrate. In particular, no further external substrate holders are required to stabilize the product substrate.

[0143] The advantages of the method for separating substrate stacks apply accordingly to the device for separating substrate stacks or removing layers, and vice versa.

[0144] Furthermore, the invention comprises an arrangement or a substrate stack, in particular for the production of semiconductor components, at least comprising,

[0145] A) a carrier substrate,

[0146] B) a separating layer arranged on the support substrate and

[0147] C) at least one further bonding surface for fusion bonding on the separating layer or the layer system into which the separating layer is integrated,

[0148] D) at least one product substrate and

[0149] E) at least one bonding surface on the product substrate for fusion bonding,

[0150] wherein at least one further bonding surface for fusion bonding is arranged on the side of the separating layer facing away from the support substrate on the layer system, and wherein the support substrate can be locally separated by irradiating the separating layer with a laser unit, wherein the layers in the layer system are preferably dielectric in the arrangement. It is also conceivable that the arrangement can be processed at temperatures above 400°C prior to separation.

[0151] Another aspect is that the inorganic layer system, with its dielectric and inorganic layers matched in material and thickness, and a reflective layer, improves the efficiency of the separation between the product substrate and the support substrate. This ensures that the laser's radiation energy is preferentially absorbed within a certain diameter in the separation layer with over 90% efficiency. Gas formation during sublimation or dissociation of the separation layer leads to local separation, particularly in the primary region. The resulting particles are captured and condense locally at the cold spots of the opened cavity, preventing the generation of additional particles during complete separation.

[0152] The gas formation locally exceeds the inherent strength of the separating layer, thus initiating crack formation in a lateral direction along the separating layer.

[0153] Another aspect is that the separation layer is locally separated within the primary region using electromagnetic radiation, particularly a laser beam, so that the product substrate and the support substrate are at least partially kept apart by the generated gas cushion, preventing spontaneous adhesion of the separated substrate sides of the product substrate and the support substrate. The particles generated during the separation, which are particularly hot, recondense in the closed interface without generating any particles within the device.

[0154] In particular, the method and apparatus for processing a substrate can be used to separate a product substrate from a support substrate or a support substrate from a product substrate, especially in a two-stage separation process. The term "individual substrate" also includes, in particular, composite, multilayer substrate stacks. A key aspect of the method and apparatus for separating a substrate stack is the deposition of exclusively inorganic layers onto at least one support substrate, at least one of which is a release layer. The thickness of the release layer is preferably less than 1 µm, typically less than 100 nm, preferably less than 50 nm, more preferably less than 40 nm, most preferably less than 30 nm, and most preferably less than 20 nm.

[0155] Preferably, the effectiveness of the separation decreases below a minimum thickness of the separating layer, so that the minimum thickness of the separating layer should be 1 nm, preferably 3 nm, particularly preferably 5 nm, and most preferably 10 nm. During the separation of the substrates, this separating layer is bombarded, in particular, by a beam, especially a laser beam or a comparable high-intensity electromagnetic radiation source, or by a particle beam. Laser beams are discussed as an example. Correspondingly adapted statements apply to particle beams. The electromagnetic beams are used within a certain diameter to partially separate the substrate stack, thereby creating a gas-tight sealed edge zone, i.e., the secondary region. The edge zone or secondary region is opened, in particular, by mechanical means such as at least one blade or a full-circumference blade, and the substrate stack is separated.Since the blade exerts a mechanical force on the edge zone or on the secondary area, it is understood here as a form of the mechanical force.

[0156] The laser parameters preferably meet certain conditions to cleanly separate the product substrate from the support substrate, or the layers arranged on the side of the separating layer facing away from the support substrate, by influencing the inorganic separating layer, preferably optically, with minimal stress, preferably without stress on the product substrate and the support substrate. In this way, the use of organic layers in the substrate to be processed can advantageously be completely dispensed with. The use of the high-temperature-resistant, inorganic separating layer or inorganic layer sequence thus enables processing steps that include temperature ranges which would not be achievable for organic and therefore less temperature-resistant layers, especially polymer layers, without negatively affecting the holding force.

[0157] In particular, it is intended that the energy of a laser is focused onto the separation layer within a certain diameter to induce a reduction in adhesion and / or even a desirable sublimation of the separation layer. This, through additional gas pressure, forces the layers apart beyond the lateral irradiation area, thus enabling high efficiency in the separation process, provided that the cohesion of the top and bottom bounded layers is higher than the adhesion of the adjacent material. This allows cracking to occur along the layers and not perpendicular to them. Preferably, the substrate material, the surface properties of the substrate, the laser wavelength, the laser energy, and / or, most importantly, the exposure time—primarily defined by the laser pulse duration in the case of pulsed lasers—and the point of application are coordinated.Since very few efficient, specific transfer processes exist in the infrared and visible spectral ranges (in contrast to the ultraviolet range) that lead to molecular cleavage through direct interaction with chemical bonds (photochemical dissociation), a "cold chemical" separation is not easily achieved. However, inorganic separation layers tuned to the laser wavelength with photochemical dissociation are preferred. Therefore, nonlinear optical effects or short thermal pulses are employed in this low-photon-energy, long-wavelength spectral range. The latter are intentionally kept so short that heat propagation within the pulse's interaction duration is limited as much as possible to the separation layer. This, along with the lateral limitation, also prevents the heat from acting on the product substrate in high concentrations, instead directing it to conversion processes (e.g.,...)in the gas phase) or is distributed over a significantly larger volume and cross-sectional area before extraction, thus reducing temperatures by several orders of magnitude and preventing undesirable damage to the substrates. In particular, the pulse duration should be in the 1-2 picosecond range because the heat concentration can be temporally confined within the layer at a thickness of <1 pm. Furthermore, this allows undesirable nonlinear processes in the support substrate to be suppressed. Later in the text, tables with optimal parameter sets are presented, illustrating the most specific and preferred combinations.

[0158] A further development of the process and the device describes the use of a purely inorganic bonding layer and additional layers for connecting two substrates or layers, in addition to the inorganic separating layer. This allows for clean layer or substrate transfer even at higher temperatures, since neither the separating layer nor the bonding layer consists of organic material, particularly organic polymers prone to carbonization. The inorganic layers are also characterized by their very high absorption (linear or nonlinear), enabling the production of particularly thin layers or more complex layer systems. A further advantage of the thin layers is the low thickness variation of the substrate stack, which allows for more precise processing, especially precise lithography or deposition processes of the product substrate.

[0159] The resulting substrate stack or layer system is preferably exclusively inorganic. Due to its inorganic structure, the substrate stack, and in particular the product substrate, can be processed at very high temperatures. A further advantage is the relatively high adhesive strength that exists between the two substrates or the inorganic layers. Therefore, the inorganic separating layers and, if applicable, inorganic bonding layers can be made thinner. The inorganic material of the separating layer can thus be advantageously tailored to the different parameters of the laser unit. This allows the laser unit to target the separating layer precisely, without affecting other layers and / or the substrate, or only to a minimal extent.

[0160] The method for processing a substrate stack aims at least partially removing or destroying, or reducing the adhesion of, the inorganic, particularly dielectric, separating layer. This enables the transfer of other layers and / or the separation of the substrate stack, especially the support substrate, from another substrate in two separation steps.

[0161] In one embodiment of the method, an inorganic bonding layer is used in addition to the inorganic separating layer. The inorganic bonding surface facilitates the connection of two substrates or several layers.

[0162] The following section describes the parameter ranges with which the procedure can be carried out, for example, or with which the device operates.

[0163] The method is based on focusing the laser beam of a laser, preferably an infrared laser, onto a separating layer within a specific diameter. For this purpose, the substrate stack is aligned, particularly at the substrate edge, using a process called edge alignment and mounted on a substrate holder.

[0164] The laser parameters meet some of the following criteria in particular:

[0165] The following lasers or laser units are preferred in the method and apparatus for separating carrier substrates: @ -In the ultraviolet spectrum

[0166] -F2, ArF, Nd:YAG, He-Ag, KrF, XeCl, He-Cd, XeF

[0167] - In the visible spectrum

[0168] -He-Cd, Ar, copper vapor He-Ne, Kr, ruby

[0169] -In the near-infrared spectrum

[0170] -Nd:YAG, He-Ne, ErGlass, Tm:YAG, Ho:YAG, ErYSGG. He:YAG

[0171] -In the far-infrared spectrum

[0172] Methanol, methylamine, methyl fluoride

[0173] The wavelength of the laser beam emitted by the laser unit is between 0.1 pm and 500 pm, preferably between 0.2 pm and 100 pm, more preferably between 0.3 pm and 50 pm, most preferably between 0.5 pm and 10 pm, and most preferably between 1 pm and 2.5 pm.

[0174] The following material classes and materials are preferably used for the separating layer.

[0175] • Semiconductors, especially Ge, other low-bandgap semiconductors,

[0176] • Metals, in particular

[0177] Ti, W, AI, Ta, Cu

[0178] • Nitrides, in particular

[0179] TiN, TaN, CrN, WN, W2N, WN2.

[0180] The laser or laser unit is preferably operated in pulsed mode. A short pulse duration is of particular interest. The short pulse duration ensures localized heat input in the interface and largely prevents heat conduction to other layers. The laser pulse duration is between 10,000 ps and 1 ps, preferably between 1,000 ps and 1 ps, more preferably between 500 ps and 1 ps, most preferably between 100 ps and 1 ps, and most preferably between 50 ps and 1 ps.

[0181] The laser spot size is the effective cross-sectional area of ​​the laser beam in the interface. If the laser spot is circular, it is preferably specified by a laser spot diameter. The laser spot diameter is less than 150 pm, preferably less than 25 pm, even more preferably less than 10 pm, most preferably between 5 pm, and most preferably less than 1 pm. If the laser spot is square, it is specified by a laser spot side length. The laser spot side length is less than 150 pm, preferably less than 100 pm, even more preferably less than 50 pm, most preferably between 25 pm, and most preferably less than 15 pm.

[0182] If the laser area is generally rectangular, it is specified by a first and a second laser area side length. The first and / or the second laser area side length is less than 150 pm, preferably less than 100 pm, even more preferably less than 50 pm, most preferably between 25 pm, and most preferably less than 15 pm. The average laser area is less than 20,000 pm², preferably less than 500 pm², even more preferably less than 80 pm², most preferably less than 20 pm², and most preferably less than 1 pm².

[0183] Neglecting convergence, the laser area corresponds approximately to the laser beam diameter along the path of the laser beam. The energy input per pulse is between 0.01 pJ and 128 pJ, preferably between 0.125 pJ and 64 pJ, more preferably between 0.25 pJ and 32 pJ, most preferably between 0.5 pJ and 16 pJ, and most preferably between 1 pJ and 8 pJ. The corresponding laser area energy density per pulse is calculated as the quotient of the energy per pulse and the laser area.

[0184] The roughness of the substrate surface influences the scattering of the laser radiation. Preferably, the roughness is adjusted so that a maximum number of photons penetrate the substrate. The roughness is specified either as mean roughness, quadratic roughness, or mean roughness depth. The determined values ​​for mean roughness, quadratic roughness, and mean roughness depth generally differ for the same measurement length or area, but are within the same order of magnitude. Therefore, the following numerical ranges for roughness should be understood as values ​​for mean roughness, quadratic roughness, or mean roughness depth. The roughness of the substrate surface is less than 100 pm, preferably less than 10 pm, even more preferably less than 1 pm, most preferably less than 100 nm, and most preferably less than 10 nm.The distribution of the laser surface energy along the position is not necessarily homogeneous. The laser surface energy is characterized in particular by one of the following distribution functions.

[0185] • Gaussian distribution

[0186] • Cauchy distribution

[0187] • Lorentz distribution

[0188] • Pearson distribution

[0189] • Equal distribution.

[0190] Furthermore, when separating a substrate layer, particularly within a specific diameter, the nonlinear optical effects occurring during laser irradiation of the separation layer can be considered and utilized. By combining the appropriate physical parameters, especially the layer sequence, the substrate material, the pulse length, the laser wavelength, and the laser energy, the behavior of the electromagnetic wave or photons within the substrate can be adjusted to focus the laser beam at the separation layer. This is particularly enhanced by the boundary layer and further amplified by the reflective layer. The electro-optical Kerr effect, which describes the changes in a material's optical properties, especially its refractive index, as a function of the electric field strength, is primarily responsible for this effect.

[0191] The selected laser parameters, and the resulting spatially focused, short-term, high-energy energy input, cause at least one of the following physical effects. The extreme temperature increase leads to thermal expansion in and / or between the interface layer and at least one other layer or substrate adjacent to the interface layer.

[0192] This effect is more efficient the greater the difference in the coefficients of thermal expansion between the interface layer and the adjacent layers or substrates. Coefficients of expansion are temperature-dependent, but are on the order of 10⁻⁶. 6 K-1. A ratio is therefore appropriate. The absolute value of the difference between the coefficient of expansion of the separation layer and the coefficient of expansion of at least one adjacent layer or substrate is greater than 0.1*10-6 K-1 , preferably larger than 1.0*10 -6 K-1, preferably larger than 2.5*10-6; K-1, most preferably larger than 5.0*10' 6 K-1, preferably larger than 10.0*10' 6 K-1. If the thermal expansion exceeds a critical value, the separating layer or the support substrate arranged on the other side of the separating layer can be locally separated or detached from at least one further layer or substrate. Since the laser radiation is only applied within a certain diameter, the particles are trapped in the separated interior of the substrate stack, and the substrate stack is completely separated, particularly mechanically, in the second separation step.

[0193] By selecting a very short pulse duration, more heat per unit of time can be introduced into the interface than is dissipated into the surroundings. This leads to sublimation and, in some cases, the formation of a plasma. If the pulse duration were too long, the inorganic interface would melt. Due to the faster heat dissipation into the surroundings, the melt would solidify very quickly, resulting in the interface re-welding to the surroundings – undesirable effects. However, in this way, the particles condense in the respective cold zones of the separated interface and do not burden the system.

[0194] Local melting of the inorganic interface is also conceivable. As already mentioned, when melting, it must be ensured that resolidification of the melt does not weld the interface back to its surroundings. This can be achieved, in particular, by additional gas evolution of the melt, and / or, in a particularly preferred approach, the laser spots in the interface do not overlap. In this case, the step size between two generated laser spots must be larger than the laser beam area. Each laser spot in the interface or area of ​​influence is then illuminated by a corresponding laser beam from the laser unit. The following parameter sets are given as examples.For a preferred circular laser area with a laser area diameter of 100 pm, the step size is between 10 pm and 300 pm, preferably between 10 pm and 250 pm, more preferably between 10 pm and 200 pm, most preferably between 10 pm and 150 pm, and most preferably between 10 pm and 120 pm. The step size should be chosen to be larger than the laser area diameter but still large enough to effectively weaken the separation layer or its adhesive properties. For example, with a laser area diameter of 10 pm, it can be ensured that the weakening or destruction of the separation layer occurs even with a step size of 30 pm. In particular, it is not necessary to reduce the step size to, for example, 15 pm or even 12 pm.

[0195] If the laser surfaces were to overlap, the sublimated inorganic material of the separating layer of one laser surface could condense or resublimate in the adjacent laser surface, leading to renewed welding.

[0196] However, it is desirable that sublimated material from the separating layer of a laser surface, which forms particles, condenses or resublimates in the adjacent laser surface without leading to renewed welding.

[0197] In the prior art, the polymer would absorb the sublimated separating layer material. Therefore, a crucial aspect of all embodiments of the method and the device for separating substrates is that, by correctly selecting the step size between two laser planes for a given laser area, re-welding can be prevented. Furthermore, it should be noted that a laser with a sufficiently short pulse duration will not function with systems containing organic layers.

[0198] Particularly preferably, photochemical dissociation and / or splitting of the interface can be achieved by the effect of the laser radiation with gas formation, such that the interface is broken up under gas formation and overpressure and / or due to the difference in thermal expansion, preferably over a larger area than the irradiation spot. In the less preferred case, the photochemical dissociation of the interface can occur within the size of the laser spot. In this case, the separation process may even require an overlap of the individual laser spots to separate the substrate stack within a single diameter.

[0199] In a first method, the separation layer is used to debond a substrate stack into two substrates. For clarity and to provide a comprehensive overview, the method is described using two substrates: a support substrate and a product substrate. However, the method can also be used to create a substrate stack consisting of several substrates, particularly several thinned substrates. In this case, either at least one of the product substrates should be thick enough to provide sufficient mechanical stabilization of the substrate stack, or the entire substrate stack should be thick enough to ensure mechanical stabilization. In this way, a person skilled in the art can successively create a substrate stack with multiple layer or substrate transfers, each using a two-stage debonding process between the resulting product substrate stack and the support substrate.

[0200] In the first process step, a sequence of layers with a release layer is applied to a support substrate. A bonding layer is then applied to the release layer. A bonding layer is also applied to a product substrate. The product substrate is then bonded to this bonding layer on the support substrate.

[0201] In a second process step, the product substrate is processed. This processing may include, in particular, re-dilution.

[0202] In a third, optional process step, the product substrate with its processed product substrate surface is bonded to another substrate, in particular a transfer substrate.

[0203] In a fourth process step, the substrate stack is aligned, particularly mechanically, on the substrate holder.

[0204] In a fifth process step, the separation layer through the support substrate is bombarded with a laser beam of a laser within a diameter and the first step of the separation is carried out.

[0205] In a sixth process step, the substrate stack is mechanically separated completely outside of a diameter. This separates the carrier substrate from the product substrate.

[0206] A layer sequence is produced on a support substrate to form a layer system with a separating layer. In particular, an oxide layer is produced on the support substrate. An absorber layer for adjusting the optical parameters of the support substrate is produced on the oxide layer. Another oxide layer is produced on the absorber layer. A separating layer is produced on the oxide layer. Another oxide layer is produced on the separating layer, which serves as a protective layer, for example, as a defocusing layer for the product substrate. A (further) bonding surface is produced on the protective layer. A product substrate, preferably also provided with a bonding surface, is bonded to the bonding layer of the support substrate. The bond is preferably a fusion bond. The product substrate, in particular, already possesses functional units.

[0207] In further process steps, the second, unbonded product substrate surface is processed. In particular, it is thinned to less than 100 pm, preferably less than 50 pm, more preferably less than 25 pm, most preferably less than 10 pm, and most preferably less than 5 pm. Further process steps, especially at high temperatures, can be carried out on the thinned product substrate. Preferably, the second product substrate surface is oxidized, and TSVs are generated, so that the second product substrate surface becomes a hybrid bonded surface. A further fusion bond between a second product substrate and the second product substrate surface of the first product substrate is also conceivable. This bond is then preferably a hybrid bond, i.e.,The electrical contacts of the first product substrate are directly connected to the electrical contacts of the second product substrate, while the dielectric environment of the electrical contacts is bonded together by a fusion bond. The surface of the second product substrate preferably has a bonding layer. If the substrate stack consisting of the first and second product substrates is mechanically stable enough, the process of interface weakening can be applied to the interface. The laser beam is preferably focused through the support substrate, which is transparent or permeable to the specific wavelength of the laser, onto the interface within a certain diameter. This causes the interface to locally lose its adhesive strength or at least be partially removed. The two bonded product substrates can then be mechanically separated from the support substrate in a second step.

[0208] Only inorganic layers are used, meaning the layer sequence and all bonding layers are inorganic.

[0209] In a second process, the separation layer is used to transfer a transfer layer. This process is called layer transfer.

[0210] In the first process step, a carrier substrate is provided. A sequence of layers is deposited on the carrier substrate, including a bonding layer,

[0211] -at least one transmission layer,

[0212] - a protective layer,

[0213] - a separating layer, especially for two-stage separation,

[0214] - a boundary layer,

[0215] -and optionally a reverse-side antireflective layer. Preferably, an optional diffusion barrier layer can be used in all methods.

[0216] In other embodiments, the transfer layer and the bonding layer can be combined. In such cases, the transfer layer is also used as the bonding layer and is bonded directly to the product substrate.

[0217] It is also conceivable, and preferred, to first apply a growth layer to the protective layer and then the transmission layer to the growth layer. It is also conceivable that the transmission layer is an overgrowth layer, which must grow through a mask. The creation of such an overgrowth layer is described in detail in publication WO 2016 184 523 A1. It is also conceivable that a diffusion layer (diffusion barrier) must be deposited between the protective layer and the transmission layer to prevent them from mixing during subsequent process steps. A diffusion barrier layer can, in particular, be created as a dielectric within the layer sequence.

[0218] In a second process step, the transfer layer, with its free transfer layer surface, is aligned with a product substrate. The product substrate may already contain functional units and / or other layers. It is also conceivable that the substrate is a transfer substrate, which only temporarily accepts the transfer layer and transfers it to a product substrate in subsequent process steps. In this case, a corresponding inorganic separation layer can also be created on the transfer substrate, particularly for two-stage separation.

[0219] In a third process step, the transfer layer is bonded to the product substrate. In a fourth process step, the separation layer, particularly within a certain diameter, is subjected to the laser beam, causing it to either lose its adhesive strength and / or be at least partially destroyed. However, the substrate stack is not completely separated.

[0220] In a fifth process step, the substrate stack is completely separated mechanically.

[0221] In a sixth process step, the support substrate is removed, leaving the transfer layer attached to the product substrate. In particular, the growth layer also remains on the transfer layer.

[0222] In a seventh process step, any remaining growth layer is removed from the transfer layer.

[0223] The main difference between the exemplary separation methods lies in the fact that, on the one hand, two substrates are separated from each other, and on the other hand, a layer is transferred. For all the methods mentioned, an inorganic separation layer is preferably used. This two-stage separation reduces the particle load both on the product substrate and in the device, because particles generated by the laser beam during separation are not released.

[0224] Table 1

[0225]

[0226] Further advantages, features, and details of the invention will become apparent from the following description of preferred embodiments and from the drawings. These show schematically in:

[0227] Fig. 1a shows a temporarily bonded substrate stack before separation, as a first process step.

[0228] Fig. 1b shows a process step of separating a substrate stack within a diameter d,

[0229] Fig 1c shows an intermediate result of the process of separating the substrate stack.

[0230] In the figures, identical components or components with the same function are marked with the same reference symbols.

[0231] Fig. 1a schematically shows a substrate stack 3 comprising a support substrate 1 and a product substrate 2, which are connected to each other via an inorganic layer sequence 4 with a separating layer.

[0232] Fig. 1b schematically shows the separation of the substrate stack 3, which comprises a support substrate 2 and a product substrate 1, which are connected to each other via the bonding layer 4, in particular in the form of a layer system. Diameter D shows the outer diameter at which the substrates 1 and 2 are connected to each other. The separation is carried out by means of a laser 5, which directs a laser beam 6 through the support substrate 1 to the separation layer 4 for sublimation of the separation layer.

[0233] Detail A shows that within a diameter d, substrates 1 and 2 are separated from each other in a primary region 11. Outside the diameter d, the connection between the substrates has not been severed. This is the secondary region 12. Detail B shows that within the diameter d, complete separation has occurred, and substrates 1 and 2 are no longer connected by a gap 7. The separation gas generates an overpressure that creates and maintains the gap 7 between substrates 1 and 2. This prevents spontaneous bonding. Furthermore, the generated particles are retained in the gap 7, thus preventing contamination of the device. Fig. 1c shows an intermediate stage of the process for separating the substrate stack 3.

[0234] Reference symbol list:

[0235] 1 Product substrate

[0236] 2 Carrier substrate

[0237] 3 substrate stacks

[0238] 4 Bonding layer

[0239] 5 lasers

[0240] 6 Laser beam

[0241] 7 gaps

[0242] 11 Primary area

[0243] 12 Secondary area

[0244] d inner diameter

[0245] D outer diameter

Claims

Claims 1. Method for temporarily joining a product substrate (1) and a support substrate (2), comprising: Providing the product substrate (1) and the support substrate (2), connecting the product substrate (1) and the support substrate (2) to each other, forming a bonding layer (4) between the product substrate (1) and the support substrate (2) in a connected state, separating the product substrate (1) from the support substrate (2), wherein the product substrate (1) and the support substrate (2) in a primary area (11) under a gas evolution in a first separation step and in a secondary area (12) which surrounds the primary area (11), in particular in a closed and / or gas-tight manner, are separated in a second separation step.

2. Method according to claim 1, wherein the bonding layer (13) is essentially inorganic.

3. Method according to one of the preceding claims, wherein the primary area (11) for the first separation step is processed with a first means and the secondary area (12) for the second separation step is processed with a second means different from the first means.

4. Method according to one of the preceding claims, wherein the first means is electromagnetic radiation, in particular laser radiation (6), and / or the second means is a mechanical force.

5. Method according to one of the preceding claims, wherein the product substrate (1) and the carrier substrate (2) are transported in the connected state after the first separation step.

6. Method according to one of the preceding claims, wherein the secondary area (12) is more than twice, preferably more than 5 times and particularly preferably more than 10 times as large as the primary area (11).

7. Method according to one of the preceding claims, wherein the primary region (11) is circular and the secondary region (12) is annular.

8. Method according to one of the preceding claims, wherein in the connected state between the support substrate (2) and the product substrate (1) a separating layer separate from the bonding layer (4), in particular a separating layer decoupled from the bonding layer (4), wherein the product substrate (1) is separated from the support substrate (2) by separation along the separating layer, wherein the separating layer (11) is in particular designed to be substantially inorganic.

9. Method according to claim 8, wherein the separating layer is provided as part of a layer system arranged between the support substrate (2) and the product substrate (1).

10. Method according to claim 8 or 9, wherein in the connected state the distance between the bonding layer (4) and the separating layer is greater than 250 nm, preferably greater than 500 nm and particularly preferably greater than 800 nm.

11. Method according to one of the preceding claims, wherein the product substrate (1) and the support substrate (2) in the combined state are exposed in a treatment step to a temperature greater than 600°C, preferably greater than 800°C or particularly preferably greater than 1100°C.

12. Method according to one of the preceding claims, wherein the layer system comprises a light focusing layer, in particular a focusing layer, and / or a defocusing layer and / or at least one protective layer, in particular a dielectric protective layer, and / or a barrier layer.

13. Method according to one of the preceding claims, wherein a layer sequence with - a bonding layer, - a protective layer, - a separating layer, @ - a light-bundling layer, -and preferably manufactured with a rear anti-reflective coating.

14. Method according to one of the preceding claims, wherein after the product substrate (1) is detached from the carrier substrate (2) the bonding layer (4) is at least temporarily or permanently part of the detached product substrate (2).

15. Device for carrying out a method according to any one of claims 1 to 14.

Images