WO2024104550A1 - Method for separating a first substrate layer, device for carrying out such separating, and substrate comprising a first substrate layer - Google Patents

Abstract

The invention relates to a method for separating a first substrate layer (1) along at least one separating line, comprising: – providing the first substrate layer (1), and – separating the first substrate layer (1) along the separating line produced by irradiating the first substrate layer (1) by means of laser light (2), characterized in that in order to form the separating line, a power of the laser light (2) is set in such a way that the power of the laser light (2) assumes a value above a critical power value for forming self-focusing of the laser light (2) in the first substrate layer (1).

Description

Method for separating a first substrate layer, device for carrying out such a separation and substrate with a first substrate layer

The present invention relates to a method for separating a first substrate layer, a device for carrying out such a separation and a substrate with a first substrate layer.

There are several methods in the state of the art for separating substrates along a plane. One of the most important and best known methods is the so-called SmartCut® method. In the SmartCut® method, atoms, particularly hydrogen atoms, are implanted into a substrate. In a first step, the atoms are ionized and accelerated onto the substrate using an electric field. The kinetic energy of the ions is sufficient to penetrate the substrate. The penetration depth can be adjusted very precisely using the acceleration voltage. In a second step, the substrate processed in this way is bonded to another substrate. It would also be conceivable to form an oxide layer on the substrate before the bonding process. In a third step, the substrate is heated. When the substrate is heated, the hydrogen atoms recombine to form hydrogen molecules. The hydrogen atoms are not incorporated directly into the crystal lattice of the substrate, so the substrate is only hydrogenated. The hydrogen molecules, on the other hand, form a gas. The phase transition from atomic hydrogen to hydrogen gas causes a volume expansion, which leads to a fracture of the structure along the implantation plane. This process allows very thin layers to be transferred from one substrate to another. Even more importantly, however, it enables the production of a substrate stack that has a very thin crystalline layer on its top, under which there is an oxide layer. Such substrate stacks are called silicon-on-insulator (SOI).

Other methods of substrate separation include chemical and/or mechanical aids. Some methods use a laser beam to pre-damage the substrate. along a plane in the substrate, but this is not enough to break the substrate along that plane. The substrate must then be etched with chemicals. The chemicals then react with the atoms along the plane in which the pre-damage was done and etch the substrate along that plane. This process is time consuming. Mechanical aids such as blades or wires can be used to break the substrate along the pre-damaged plane. The use of mechanical release agents speeds up the process, but can lead to unwanted fracture of the substrate outside the plane along which fracture should occur.

The publication US 7 052 978 B2 shows laser-induced damage to a substrate. The laser creates damage within the substrate through various effects. However, focusing of the laser is not known from this.

All of the prior art processes have the disadvantage that an additional component is required to produce the desired separation effect or that controlling the fracture process is either impossible or very difficult. In the SmartCut® process, for example, atoms must be implanted into the substrate. In a chemical etching process, chemicals must be made available. In a mechanical separation process, only mechanical separation agents such as a blade or wire would be required, but experience has shown that these only allow the separation of comparatively thick layers from a substrate. They are not suitable for separating thin layers from the substrate, and in particular not for separating layers with thicknesses in the micrometer or nanometer range. In addition, controlling the fracture along the desired plane using mechanical separation agents is difficult or even impossible. In processes in which a laser beam is shot into the substrate, optics are used to focus the laser beam.

Proceeding from this, the present invention has for its object to provide a method which simplifies the separation of substrates, in particular by eliminating the disadvantages described, and in particular is also suitable for providing substrate sub-layers which are as thin as possible after separating a substrate layer. The present invention solves the problem with the method for separating according to claim 1, with a device for separating according to claim 13 and a substrate according to claim 15. Further preferred embodiments of the invention emerge from the following description and from the drawings.

According to a first aspect of the present invention, a method for separating a first substrate layer along at least one separating line is provided, comprising:

- Providing the first substrate layer, and

- separating the first substrate layer along the separating line which is produced by irradiating the first substrate layer by means of laser light, wherein, to form the separating line, a power of the laser light is set such that the power of the laser light assumes a value above a critical power value for forming a self-focusing of the laser light in the first substrate layer.

In contrast to the prior art, the invention provides that the power of the laser light used is set in such a way that the phenomenon of self-focusing of the laser light occurs in the first substrate layer to be treated, in particular in the first substrate layer to be separated. The phenomenon of self-focusing is based on the Kerr effect and occurs in particular when a spatially profiled intensity profile of the laser light with a sufficiently high intensity amplitude interacts with the material of the first substrate layer. Non-linear effects occur here. In an alternative embodiment, the phenomenon of self-focusing is based on non-linear absorption. Since refractive index properties depend on the light intensity, a sufficiently high light intensity can lead to a refractive index profile in the material, which leads to the light being focused in a plane within the first substrate layer due to the changes in the refractive index profile initiated by the light itself. As a result, a partial area or section of the later dividing line or dividing plane is defined or determined in the focal point or focus defined by the self-focusing.

The occurrence of the phenomenon of “self-focusing” depends on the material and laser properties. Depending on the material selected, different critical powers or pulse energies arise, and if exceeded, self-focusing occurs. Based on experimental findings and/or simulations, it is possible to determine a critical power value for the laser type or the laser used. Laser light on the one hand and the material used in the first substrate layer on the other hand are to be determined, estimated or estimated. The effect is based on the fact that the refractive index of a material for high intensities depends on the intensity, particularly linearly. If a laser beam with an inhomogeneous, particularly Gaussian-shaped, intensity cross-section is shot into a transparent material, then polarization in the part of the laser beam close to the center leads to a different phase shift than in surrounding parts. It can be proven mathematically and physically that this circumstance leads to self-focusing of the laser beam, i.e. photons of the laser beam are focused on a point in a focal plane in the first substrate layer, even if the laser beam was not focused when it entered the material.

In particular, the self-focusing determines or creates a focus or focal point in the area through which the dividing line is to run. In particular, the dividing line is to be understood as a two-dimensional dividing plane. By means of a corresponding process or displacement of the first substrate layer relative to the laser light or the beam path of the laser light, a flat formation of the dividing line, which then forms the dividing plane, takes place.

Through self-focusing, it is advantageously possible to cause such high intensities within the first substrate layer that these high light intensities in turn lead to physical and/or chemical modification within a plane or within a point in which the self-focusing takes place. This modification in the material properties then leads to the formation of the dividing line or plane, for example a predetermined breaking point or even to independent separation along the dividing line or plane. Advantageously, the use of extensive etching agents or complex optical systems or the injection of ions into the first substrate layer can be dispensed with or the extent of this can be significantly reduced. This at least makes it possible to further reduce the effort involved in actually breaking or separating the first substrate layer, which can improve the separation of the first substrate layer.

The fact that only the power of the laser light needs to be adjusted to achieve self-focusing is particularly advantageous. Even if Since a large number of parameters must be taken into account to determine the critical power, it is sufficient to increase the power of the laser light in order to initiate self-focusing and, in particular, to determine the position of the self-focusing. This makes it possible to determine or influence the location of the interaction between the material of the first substrate layer and the laser light in a relatively simple manner. In particular, it is possible to react relatively quickly to a change in material, for example if the first substrate layers to be treated are exchanged or the position of the dividing line is changed. No extensive adaptation of the system is required to carry out the process.

In particular, it is provided that the first substrate layer is part of a substrate composite or is to become part of a substrate composite with the first substrate layer and a second substrate layer. For example, the first substrate layer is a wafer or part of a wafer and/or a bonding layer that connects two other wafer substrate layers to one another, for example. The first substrate layer and/or a second substrate layer preferably comprise materials such as silicon or germanium. The first substrate layer preferably has a thickness of between 1 nm and 5 mm, preferably between 1 nm and 1 pm and particularly preferably between 1 nm and 3,000 nm.

Using a Czochralski process, for example, an ingot can be drawn from a silicon melt. The process described is also suitable for separating sub-layers of substrate from this ingot, so that the separated sub-layer of substrate can then form its own substrate. The sub-layers of substrate, which are then processed as an independent substrate, would in this case have thicknesses in the millimeter range. In particular, separation takes place before the first substrate layer is bonded to the second substrate layer.

Substrate sublayers with thicknesses in the micrometer or nanometer range are mainly produced in processes in which a layer transfer is to take place. In this case, the first substrate layer to be separated is first bonded to a second substrate layer and only then is the process used to separate the first substrate layer. In particular, separation takes place after the first substrate layer has been bonded to the second substrate layer. The surface of the first substrate layer preferably has a roughness of between 0.01 and 300 nm, preferably between 0.01 and 30 nm and particularly preferably between 0.01 and 10 nm. The scattering of the laser beam coupled into the first substrate layer is lower the roughness of the substrate layer surface is lower. Low surface roughnesses on the substrate layer surfaces are therefore preferred.

In a preferred embodiment of the method, it is provided that the laser beams have a wavelength that is between 0.1 pm and 500 pm, preferably between 0.2 pm and 100 pm, particularly preferably between 0.3 pm and 50 pm, particularly preferably between 0.5 pm and 10 pm or even between 1 pm and 2.5 pm. In this way, the separating layer can be irradiated particularly efficiently and in a targeted manner.

In a preferred embodiment of the method, the laser has a power that is between 1 W and 1000 W, preferably between 5 W and 800 W, even more preferably between 7 W and 600 W, most preferably between 10 W and 500 W, most preferably between 20 W and 200 W.

In a preferred embodiment of the method, the laser area is smaller than 2000 pm ² , preferably smaller than 500 pm ² , preferably smaller than 80 pm ² and particularly preferably smaller than 20 pm ² , or even smaller than 1 pm ² . The area of the separating layer or separating line on which the laser beam acts is advantageously small and targeted in order to locally reduce the adhesive properties of the separating layer or to destroy it.

In a preferred embodiment of the method, there is provision for at least 0.1 pm, preferably at least 1 pm, preferably at least 5 pm, particularly preferably at least 10 pm or even at least 50 pm between the areas of action of the laser beams on the separating layer, so that the areas of action of the laser beams do not overlap. In this way, particularly simple and efficient separation is possible.

Preferably, ultrashort laser pulses are used as laser light, preferably ns (nanosecond) pulses, preferably ps (picosecond) pulses and particularly preferably fs (femtosecond) pulses or even as (attosecond) pulses. The duration of the light pulses is thus preferably less than 10' ⁹ s, preferably less than 10' ¹¹ s, preferably less than 10' ¹³ s, most preferably less than 10' ¹⁵ s or even less than 10' ¹⁶ s.

With correspondingly short pulse durations, it is advantageously possible to generate laser powers or light intensities that are required for self-focusing in a simple manner. Such pulses generally require the use of special optics, such as refractive, reflective and/or grating-like optical components. It is preferable to avoid the laser light passing through material with high intensity as far as possible. Such propagation through a medium can otherwise lead to modifications in the pulse shape and/or to a reduction in intensity. However, it is conceivable that beam expansion is used, whereby the overall intensity is initially reduced across the spatial profile of the laser pulse in order to refocus the laser beam preferably immediately before it enters the first substrate layer and to cancel the expansion. This makes it possible, for example, to modulate or influence the laser profile or intensity profile at low intensity in order to achieve the best possible self-focusing in the first substrate layer. In principle, the spatial intensity profile of the laser beam can also be influenced using simple apertures or other optical components. The intensity profile is preferably generated by refraction optics. It is also conceivable to use apertures, field mappers, lens arrays, integrators or beam homogenizers and/or axicons.

Preferably, the laser light is provided with a spatial profile, in particular a Gaussian, a Lorenzian or a Cauchy-shaped intensity profile. A corresponding intensity profile proves to be particularly advantageous because it is particularly suitable for the refractive index profiling that leads to self-focusing by means of the Kerr effect. Preferably, it is also conceivable that a corresponding spatial intensity profile is specifically realized by means of corresponding optical components, for example if the intensity profile provided by a laser source deviates from a preferred intensity profile. Finally, to generate the Kerr effect, it is preferably provided to generate an intensity profile of the laser radiation that is non-homogeneous in cross-section. Preferably, an intensity profile with a Gaussian, Lorentzian and/or Cauchy-shaped geometry is generated. Preferably, the dividing line extends along a plane that runs essentially parallel to the main plane of extension. This advantageously makes it possible to separate the first substrate layer along a plane that runs essentially parallel to its own main plane of extension. Separating along the dividing line that runs essentially parallel to the main plane of extension is particularly advantageous when comparatively thin substrate sublayers are to be produced. In particular, the separation leads to the separation of a primary substrate sublayer and a secondary substrate sublayer that emerge from the first substrate layer as a result of the separation. For example, coatings can be reduced to a desired dimension.

Alternatively, it is conceivable that the dividing line runs essentially perpendicular to the main extension plane, thus enabling separation, i.e. dicing, into individual parts of the first substrate layer, which are arranged adjacent to one another in the main extension plane, in particular before separation.

It is preferably provided that the first substrate layer is connected to a second substrate layer during irradiation. In particular, it is provided that the first substrate layer and the second substrate layer are connected to one another in a material-locking, form-fitting and/or force-locking manner and that the second substrate layer faces the incident laser light. In other words: the laser light initially propagates through the second substrate layer before it enters the first substrate layer and there creates a dividing line or a partial section of the dividing line due to the self-focusing that occurs. In particular, it is provided that the second substrate layer is passed by the laser light and the focus is created by self-focusing in the first substrate layer after passing through the second substrate layer. In particular, when setting the laser light power, care is taken to ensure that partial focusing preferably already takes place in the second substrate layer and the desired final focusing is realized in the first substrate layer.

The depth of the focal plane can be set by the laser parameters, in particular by the wavelength of the laser used. The closer the focal plane should be to the substrate layer surface, the shorter the wavelength of the laser radiation should be. In the laboratory, neodymium-YAG lasers with a wavelength of 1064 nm have proven to be effective for near-surface processing. With wavelengths around 1550 nm, depths of several 100 pm can be achieved.

Measurements have shown that with a wavelength of 1950 pm, a pulse duration of about 5 ps and a spot size diameter of about 20-30 pm, an energy of about 10 pJ per pulse is obtained. This is roughly the energy at which silicon begins to be damaged.

In one embodiment, the laser beam is positioned normally on the first substrate layer and the first substrate layer moves actively relative to the entire optical structure, i.e. to the laser beam. For this embodiment, the first substrate layer must be fixed to a substrate holder that can move as quickly as possible. The substrate holder should nevertheless have the highest possible resolution of the process axes in order to be able to position the laser beam precisely.

In another embodiment, the optical structure, in particular the laser beam, can move actively relative to the first substrate layer. A laser light alignment device is preferably used for this purpose.

In a further preferred embodiment, special optical elements, in particular in the form of a galvanometric scanner, are used as laser light alignment devices in order to scan the laser beam over the first substrate layer and thus form a flat separating layer running along a main extension plane. The laser light alignment device preferably comprises a telecentric lens, preferably a telecentric lens with a telecentric beam path on both sides. The laser beam is guided by the laser light alignment device through a lens or several lenses, which refract the laser beam in such a way that the laser beam always falls normally on the first substrate layer and/or second substrate layer. The lens used is usually smaller than the first substrate layer to be scanned. The first substrate layer must therefore be moved relative to the optical structure in a step-and-repeat process. The process is then repeated at several points until the entire first substrate layer to be scanned has been irradiated. It is conceivable to change the laser parameters, especially the laser power, so that the position of the focal plane can be changed.

It is preferably provided that the first substrate layer is connected to a third substrate layer, wherein the first substrate layer, the second substrate layer and the third substrate layer are arranged one above the other along a stacking direction running perpendicular to the main extension plane and the first substrate layer is the bonding layer between the second substrate layer and the third substrate layer. By corresponding self-focusing in the first substrate layer designed as a bonding layer, it is advantageously possible to enable an existing or manufactured substrate composite to be separated. Thus, with the method described here, it is also possible to carry out debonding in order to separate certain substrate layers, for example the second substrate layer and the third substrate layer, from one another.

To separate the second substrate layer from the third substrate layer, which are connected to one another via the first substrate layer, or to separate the first substrate layer, additional devices, in particular corresponding substrate holders with corresponding fixing means, can be used. For example, a first substrate holder and a second substrate holder are provided, which lie opposite one another, while the first substrate layer to be separated is arranged between the first substrate holder and the second substrate holder. A first device, in particular the first substrate holder, is located on a side facing the second substrate layer. A second device, in particular a second substrate holder, is located on a side facing the third substrate layer. The devices fix the substrates or the substrate layers with corresponding fixing means, in particular with a vacuum fixation and/or a clamping means. By moving the two devices away from one another relative to one another, the two substrate layers, i.e. the second and the third substrate layer, can be separated from one another and at the same time fixed to the devices, i.e. the first and second substrate holders. The same applies to a device that is fixed directly to an upper side of the first substrate layer to be separated.

To overcome the last remaining attraction between the substrate layers separated by means of the method, forces between 1 N and 100 kN, preferably between 1 N and 10 kN, more preferably between 1 N and 1 kN, particularly preferably between 1 N and 100 N or even between 1 N and 10 N are applied. The parting plane or parting line has a thickness between 1 nm and 10 pm, preferably between 1 nm and 1 pm, preferably between 1 nm and 100 nm and particularly preferably between 1 nm and 10 nm. If volume defects are generated in the parting plane or parting line, in particular holes, local melting and/or similar, then the size of these volume defects is between 1 nm and 10 pm, preferably between 1 nm and 1 pm, preferably between 1 nm and 100 nm and particularly preferably between 1 nm and 10 nm.

It is preferably provided that a lens is used to pre-focus the laser light before it enters the first substrate layer. This advantageously allows the light output of the laser beam to be set in a controlled manner when it enters the first substrate layer and, for example, to be adapted to the material properties of the first substrate layer. It is also conceivable to use a concave mirror for focusing, which advantageously makes it possible to both focus and deflect the laser beam, which can advantageously ensure that the device uses as little space as possible. Finally, a straight line is not required for the laser beam.

Alternatively, it is conceivable that no lens is arranged between the laser source and the first substrate layer, i.e. a beam path of the laser light is free of lenses and/or concave mirrors.

It is preferably provided that the separation is supported by a mechanical force, a further laser light irradiation, an introduction of ions, atoms and/or molecules, a chemical effect and/or an introduction of heat. In other words: the creation of the separation line is not yet completely sufficient to cause the separation of the first substrate layer, but a type of predetermined breaking plane or line is created as a separation line, which is finally broken by adding a further influence, such as a force, a further light effect and/or a chemical effect, in order to cause separation within the first substrate layer. For example, it is conceivable that corresponding suction devices adhere to the top and/or bottom of the first substrate layer and that a mechanical effect on the first substrate layer is caused by corresponding tensile and/or shear forces. If a force is necessary for separation, the force is less than 1000 N, preferably less than 500 N, more preferably less than 100 N, most preferably less than 10 N, most preferably less than 1 N.

In an exemplary embodiment, the laser beam is directed at at least one point of a focal plane in order to melt the material through a high energy density or even to sublimate it directly. If the material is melted, a separation process, for example an additional mechanical action, should preferably be carried out while the material is still in the molten state. Finally, the melting area is comparatively small due to the very focused introduction of energy by the laser beam and the heat stored in the melt is very quickly released into the environment via the substrate, so that very rapid cooling can be expected, especially before the actual separation process.

The structure of the solidified melt areas could still have a lower breaking strength than the original structure and thus serve as a predetermined breaking point. If the material is sublimated, it can only rarely resublimate in such a way that a new bond is created between the substrate halves of the first substrate layer that are to be separated. The material will still be in the border area, but in a technical sense it is considered to have been removed.

In another exemplary embodiment, the high intensity of the laser beam creates a chemical reaction in material within the focal plane, which subsequently leads to fracture.

It is preferably provided that, in particular in a preparatory process step, ions are introduced into the first substrate layer, in particular into the area of the later dividing line, in particular by accelerating them and shooting them into the first substrate layer. The focal plane is then created in the areas with ions by self-focusing. For example, the known SmartCut® process could be expanded so that the entire substrate stack no longer has to be heated, but that the effect of self-focusing allows the heat to be introduced by the laser in such a targeted manner that the hydrogen atoms recombine to form hydrogen molecules along the focal plane and thus lead to the structure breaking according to the SmartCut® process. It is preferably provided that ions are introduced into an implementation plane. A distance between the implementation plane and the focus plane is less than 1 mm, preferably less than 100 pm, preferably less than 1 pm, particularly preferably less than 100 nm, or even less than 10 nm. It is also conceivable that the first substrate layer is connected to the implementation plane via at least one coating with a second substrate layer before the first substrate layer is separated by supplying heat in the area of the implementation plane by means of self-focusing. Processing with the self-focusing light causes a physical or chemical effect that leads to a recombination of the atoms or molecules implanted in the implantation plane. In particular, when implanted hydrogen ions are used, the hydrogen atoms recombine to form hydrogen molecules. The hydrogen molecules have a higher molar volume. The hydrogen gas thus produced expands and causes damage along the implantation plane. By connecting to the second substrate layer via the at least one coating, it is advantageously possible to provide an overall substrate with the second substrate layer, which is provided with a comparatively thin substrate sublayer that remains after separation and is bonded to the second substrate layer via the coating. This makes it possible to form, for example, a silicon-on-insulator (SOI) substrate.

It is also conceivable that other atoms are present in the structure which recombine to form their corresponding molecular gases when subjected to thermal stress and produce an effect similar to the SmartCut® process. It would also be conceivable that halides, in particular fluorine, chlorine, bromine or iodine, could be ionized and shot into the substrate. It would also be conceivable to use nitrogen or oxygen atoms which recombine to form their corresponding molecular gases. The SmartCut® process is expanded by the invention in such a way that the entire substrate stack is no longer subjected to thermal treatment, which can have a gentle effect on existing functional units, in particular microchips, memory chips, MEMS, LEDs, etc. The introduction of ions, atoms and/or molecules preferably takes place before the dividing line is formed, forming a self-focusing effect.

In a further exemplary embodiment, atoms and/or molecules with a molecular structure suitable for the use are introduced into the substrate and/or the layer to be destroyed. The laser radiation is implanted with high absorption. As a result, the photons of the laser radiation are preferentially absorbed by the implanted atoms and/or molecules, which leads to a very strong thermal movement and thus to damage of the surrounding structure.

It is preferably provided that the method is used to produce a substrate layer composite, for example comprising a first substrate layer and a second substrate layer, wherein after separation a first thickness of the first substrate layer is smaller than a second thickness of the second substrate layer. The first thickness and second thickness are measured along a stacking direction running perpendicular to the main extension plane. This advantageously makes it possible to reduce coatings of substrate layers to a desired, in particular comparatively small, layer thickness. This makes it possible, for example, to realize layer thicknesses whose value is less than 1000 nm, preferably less than 500 nm, particularly preferably less than 100 nm.

Preferably, the first substrate layer comprises a ceramic material, a polymer material, silicon and/or germanium. Depending on the selected materials of the first substrate layer, the second substrate layer and/or the third substrate layer, the laser light output is adjusted accordingly in order to achieve the desired self-focusing effect.

It is preferably provided that during irradiation the first substrate layer is moved along a direction predetermined by a planned course of the dividing line within the first substrate layer. In other words: while the laser beam is stationary and in particular is not moved or shifted by pivoting mirrors, a movement takes place to create the dividing line by moving the first substrate layer relative to the laser source and in particular relative to the stationary laser beam. This can advantageously ensure that the laser light hits every point as perpendicularly as possible in order to avoid the dividing line becoming uneven due to changing alignment. Furthermore, it is advantageously possible to prevent laser light from being reflected to the side by the irradiated substrate. Instead, with appropriate constancy with regard to the surface alignment of the first substrate layer, the reflected light can be limited to a defined spatial volume. Preferably, the separation is brought about solely by irradiation with laser light. In other words: in a particularly preferred embodiment, the breaking is already caused by the self-focusing. This provides a particularly simple method, since no additional work steps are required to cause the breaking. In an exemplary embodiment, the laser beam is directed at at least one point of a focal plane in order to produce a direct break in the chemical bond there. In particular, covalent bonds in a ceramic or polymer material can be broken directly. The energy density of the self-focused laser beam is so high that the electrons that produce the covalent bond are removed from the molecular orbital.

A further subject of the present invention is a device for carrying out the method according to the invention with a laser source that is suitable for providing laser light outputs that achieve self-focusing in a first substrate layer. All of the advantages and features described for the method can be transferred analogously to the device. In particular, it proves to be advantageous to use an ultrashort pulse laser source, since the desired intensities can be achieved relatively easily with the laser pulses provided here. It is also conceivable that the source provides a control device for adjusting the laser power, whereby the laser power can advantageously be set to the desired level in order to initiate self-focusing in certain areas in a controlled manner.

It is preferably provided that the device comprises a holding element that is movable, in particular in a plane that runs parallel to the main extension plane of the first substrate layer. This advantageously makes it possible for a relative movement of the first substrate layer or the substrate layer composite to take place while the laser beam remains stationary. A further subject of the present invention is a substrate comprising at least one first substrate layer, produced using a method according to the invention. All features and properties for the method can be transferred analogously to the device. Further advantages, features and details of the invention will become apparent from the following description of preferred embodiments and from the drawings.

They show in:

Fig. 1 is a schematic representation of a method according to a first preferred embodiment of the present invention,

Fig. 2 is a schematic representation of a method according to a second preferred embodiment of the present invention,

Fig. 3 is a schematic representation of a method according to a third preferred embodiment of the present invention,

Fig. 4a is a schematic representation of a first method step of a method according to a fourth preferred embodiment of the present invention,

Fig. 4b is a schematic representation of a second method step of a method according to the fourth preferred embodiment of the present invention,

Fig. 4c is a schematic representation of a third method step of a method according to the fourth preferred embodiment of the present invention,

Fig. 4d is a schematic representation of a fourth method step of a method according to the fourth preferred embodiment of the present invention and

Fig. 4e is a schematic representation of a fifth method step of a method according to the fourth preferred embodiment of the present invention. In the figures, identical components or components with the same function are identified by the same reference numerals. The figures are not to scale. To increase clarity, a laser beam 2 intended for the method is shown wider. In particular, the laser beam 2 is drawn symbolically as a beam, although high-energy laser light pulses are preferably generated and directed at a first substrate layer 2 and/or a second substrate layer 4. Self-focusing 2k of the laser beam 2 is simplified by a convergent beam whose tip ends in a focal plane 3. Preferably, however, a large number of ultra-short laser pulses are directed along the path shown as laser beam 2 onto the first substrate layer 1 and/or the second substrate layer 4 and, due to non-linear effects in the materials of the first substrate layer 1 and/or the second substrate layer 4, leads to self-focusing 2k. Of particular importance is the high intensity in the focal point or focus 2f, which is created in particular by the self-focusing 2k.

Figure 1 shows a first embodiment of the present invention. The laser beam 2 is directed onto a focal plane 3 of a first substrate layer 1. The focal plane 3 can be determined and fixed by knowing the material properties of the substrate 1 in conjunction with the properties of the laser beam 2 used. A Kerr effect is preferably used to generate a chemical and/or physical (consequential) effect, in particular in the focal plane 3 or its immediate surroundings, which subsequently leads to the possibility of separating a primary substrate sub-layer 1o from a second substrate sub-layer 1u.

Figure 2 shows a second embodiment of the present invention. In addition to the first substrate layer 1, a substrate composite comprises at least one second substrate layer 4. It would also be conceivable to use several second substrate layers 4, which are in particular stacked, with the stack of several second substrate layers 4 being connected to the first substrate layer 1. The non-linear optical effects that lead to self-focusing of the laser beam 2 already begin in the at least one second substrate layer 4. In the present case, the focal plane 3 lies in the first substrate layer 1. If the focal plane 3 were to lie in the second substrate layer 4, an embodiment according to Figure 1 would be present. When using several layers 4, the focal plane 3 can also lie within one of the several second substrate layers and does not have to end in the substrate 1. It is therefore preferably provided to have at least one second substrate layer 4 and thus at least one further material in order to start the self-focusing of the laser beam 2 before the first substrate layer 1 and/or before the second substrate layer 4, which is actually to be separated.

Figure 3 shows a third embodiment of the present invention. In this embodiment, the self-focusing of the laser beam 2 is used to place the focal plane 3 in a first substrate layer 1 serving as a bonding layer, the task of which is to connect the second substrate layer 4 and the third substrate layer 5 to one another. As a result, the invention can also be used in particular for debonding two substrate layers of a substrate composite.

Figure 4a shows a first step of a method in which a preferably ionized atom or molecule beam 6 is shot into a first substrate layer 1. Ionized hydrogen atoms are preferably shot into the substrate layer 1. A mean penetration depth of the atoms or molecules in the first substrate layer 1 can be determined by a kinetic energy of the atoms or molecules in the atom or molecule beam 6. An implantation plane 8 is the plane in which the atoms or molecules accumulate on average.

Figure 4b shows a second method step of a method in which a coating 7 is applied to the substrate surface, via which the atoms or molecules of the atom or molecule beam 6 have been introduced. The coating 7 is in particular a thermal, preferably a native, oxide. The thickness of the coating 7 can be reduced by grinding back, thinning back and/or etching back. This method step is not shown additionally. After the coating 7 has been applied, it is optionally possible to produce a hybrid bond surface. To do this, holes are created in the coating 7 through several method steps (not shown or described), which preferably even extend into the substrate layer 1. The holes are then filled with an electric medium, preferably copper, through a coating process. The excess copper is then ground back down to the coating 7. The metal in the holes subsequently serves as an electrical contact. This process creates a hybrid surface consisting of electrically conductive copper contacts surrounded by the dielectric material of the coating 7. The person skilled in the art will be familiar with such Hybrid surfaces are known. To ensure the clarity of the drawing, the exact representation of such hybrid surfaces is omitted.

Figure 4c shows a third method step of a method in which the first substrate layer 1 is bonded via the coating 7 to a second, also coated, particularly preferably identically coated, substrate layer 4, for example without implementation level 8. The bond is preferably a fusion or direct bond. If the substrate layer 1 and the substrate layer 4 both have a hybrid bond surface, the contacts of both hybrid bond surfaces are aligned as precisely as possible with each other in this method step. Before the bond is made, functional units such as microchips, MEMS, LEDs, memory modules, etc. already exist in the substrate layer 4, which are thus electrically connected to the first substrate layer 1 via the contacts of the hybrid bond surface (not shown).

Figure 4d shows a fourth method step of the method in which the method for separating is used according to a further exemplary embodiment. The focal plane 3 of the laser beam 2 coincides approximately with the implantation plane 8. An average difference between the focal plane 3 and the implantation plane 8 is less than 1 mm, preferably less than 100 pm, preferably less than 1 pm, particularly preferably less than 100 nm, or even less than 10 nm. Due to the effect of the self-focusing of the laser beam 2 in the focal plane 3, and thus in the implantation plane 8, a very high temperature is generated along the implantation plane 8, which results in a physical and/or chemical effect on the implanted atoms or molecules. The physical or chemical effect is preferably a recombination of the atoms or molecules implanted in the implantation plane 8. In particular, when implanted hydrogen ions are used, the hydrogen atoms recombine to form hydrogen molecules. The hydrogen molecules have a higher molar volume. The hydrogen gas thus created expands and causes damage along the implantation plane 8. This effect is known from the SmartCut™ process and is now improved by separating with the self-focusing laser light in such a way that the entire substrate stack does not have to be heat-treated in an oven, but that the heat treatment is carried out very precisely, locally along the implantation plane 8. The process according to the invention thus significantly expands the SmartCut™ process known from the prior art. Any other atomic or Molecular type which, when exposed to the laser beam 2 according to the invention, results in a physical and/or chemical effect which causes damage along the implantation plane 8, can be used.

Figure 4e shows a fifth step of a process in which the primary substrate sublayer 1o (see Figure 4d) has been removed from the first substrate layer 1. All that remains is the secondary substrate sublayer 1u as a very thin layer on a coating 7, in particular an oxide layer. If the material of the secondary substrate sublayer 1u is silicon, the substrate thus created is called a silicon-on-insulator substrate (SOI). If the coating 7 was a hybrid bond surface, the copper contacts are now preferably exposed at the top. If functional units such as microchips, MEMS, LEDs, memory modules, etc. are now also produced in the secondary substrate sublayer 1u, these are automatically connected to the functional units of the substrate layer 4 (not shown).

In an alternative embodiment, the described method is carried out without a coating 7, but the first substrate layer 1 is bonded directly to the second substrate layer 4. This method is also a direct bond. The aim of this method is the direct connection of two materials, for example two different semiconductor materials.

For all embodiments described here, the focal plane 3 can be determined by knowing the material properties of the first substrate layer 1 in conjunction with the properties of the laser used. The Kerr effect is preferably used to generate a chemical and/or physical effect, in particular in the focal plane 3 or its immediate surroundings, which subsequently leads to the possibility of separating the primary substrate sub-layer 1o from the secondary substrate sub-layer 1u.

List of reference symbols

1 first substrate layer

1o primary substrate layer

1 u secondary substrate layer 2 Laser beam/laser light

2k self-focusing effect

2f focal point

3 Focal plane 4 Second substrate layer

5 third substrate layer

6 Atomic or molecular beam

7 Coating

8 implantation level

Claims

Claims Method for separating a first substrate layer (1) along at least one separating line, comprising: - providing the first substrate layer (1), and - Separating the first substrate layer (1) along the separating line which is produced by irradiating the first substrate layer (1) using laser light (2), characterized in that to form the separating line, a power of the laser light (2) is set such that the power of the laser light (2) assumes a value above a critical power value for forming self-focusing of the laser light (2) in the first substrate layer (1). Method according to claim 1, wherein ultrashort laser pulses are used as laser light (2), preferably ns pulses, preferably ps pulses and particularly fs pulses or even as pulses. Method according to one of the preceding claims, wherein laser light (2) with a spatial profile, in particular Gauss-shaped, Lorenz-shaped and/or Cauchy-shaped profile, is used. Method according to one of the preceding claims, wherein the separating line extends along a plane running essentially parallel to the main extension plane. Method according to one of the preceding claims, wherein the first substrate layer (1) is connected to a second substrate layer (4) during irradiation, wherein the second substrate layer (4) is preferably passed by the laser light (2) before the laser light (2) enters the first substrate layer (1) after passing the second substrate layer (4). Method according to claim 5, wherein the first substrate layer (1) is connected to a third substrate layer, wherein the first substrate layer (1), the second substrate layer (4) and the third substrate layer (5) are arranged along a plane perpendicular to the main extension plane. are arranged one above the other in the stacking direction and the first substrate layer (1) is the bonding layer between the second substrate layer (4) and the third substrate layer (5). 7. Method according to one of the preceding claims, wherein a lens and/or a concave mirror is used for prefocusing. 8. Method according to one of the preceding claims, wherein the separation is assisted by a mechanical force, a further laser light irradiation, an introduction of ions, atoms and/or molecules, a chemical action and/or an introduction of heat. 9. Method according to one of the preceding claims, wherein the method is used for producing a substrate layer composite which comprises at least the first substrate layer (1) and the second substrate layer (4), wherein after separation a first thickness (D1) of the first substrate layer (1) is smaller than a second thickness (D2) of the second substrate layer (4). 10. Method according to one of the preceding claims, wherein the first substrate layer (1) comprises a ceramic material, a polymeric material, silicon and/or germanium. 11. Method according to one of the preceding claims, wherein the first substrate layer (1) is moved during irradiation along a direction predetermined by a course of the separating line. 12. Method according to one of the preceding claims, wherein the separation is effected solely by irradiation by means of laser light (2). 13. Device for carrying out the method according to one of the preceding claims, with a laser source which is suitable for providing laser light powers which cause self-focusing in a first substrate layer (1). 14. Device according to one of the preceding claims, wherein the device comprises a holding element which is movable, in particular in a plane parallel to the main extension plane. Substrate comprising at least one primary substrate sublayer (1o) produced by a method according to one of claims 1 to 12, wherein the primary substrate sublayer (1o) results from the separation of the first substrate layer (1).