WO2022053274A1

WO2022053274A1 - Method for separating a transparent material

Info

Publication number: WO2022053274A1
Application number: PCT/EP2021/072958
Authority: WO
Inventors: Jonas Kleiner; Malte Kumkar; Daniel FLAMM; Myriam Kaiser; Tim Hesse
Original assignee: Trumpf Laser- Und Systemtechnik Gmbh
Priority date: 2020-09-11
Filing date: 2021-08-18
Publication date: 2022-03-17
Also published as: DE102020123787A1

Abstract

The present invention relates to a method for separating a transparent material (1), wherein localized material modifications (3) are introduced into the material (1) along a separation line (2) and at a spatial distance from one another by means of ultrashort laser pulses from an ultrashort pulse laser (6), wherein the laser beam (60) has a focal zone (600) that is elongate in the beam propagation direction, and the laser beam is preferably a non-diffractive laser beam, wherein the pulse energy is chosen so that the material modifications (3) generate material ejections (300) at a surface of the material, and the material (1) is subsequently separated along the separation line (2) within the scope of a separation step, wherein the spatial distance between the localized material modifications (3) is greater than 0.2-times to 10-times the diameter of the respective material ejections (300).

Description

Process for separating a transparent material

technical field

The present invention relates to a method for separating a transparent material, wherein localized material modifications are spaced apart from one another by means of ultrashort laser pulses of an ultrashort pulse laser and are introduced into the material along a separating line.

State of the art

The development of lasers with very short pulses, in particular with pulse durations of less than nanoseconds, and high average power in the kilowatt range enables new material processing. With these short pulse durations and high average powers, a non-linear absorption in the material to be processed becomes accessible, so that transparent or essentially transparent materials could also be processed for the laser light wavelength used. In particular, transparent materials can also be separated along a predetermined dividing line via the non-linear absorption.

A corresponding method for separating transparent materials is described in EP2754524.

According to the methods in the prior art, however, there are always breakouts on the parting surface and/or an undesirably high surface roughness and/or difficult separability/separability of the material parts.

Presentation of the invention

Proceeding from the known prior art, it is an object of the present invention to provide an improved method for separating a transparent material. The object is achieved by a method for separating a transparent material having the features of claim 1. Advantageous developments result from the dependent claims, the description and the figures.

Accordingly, a method for separating a transparent material is proposed, wherein by means of ultra-short laser pulses of an ultra-short-pulse laser, localized material modifications are spatially spaced from one another and introduced into the material along a separating line, the laser beam having a focal zone that is elongated in the direction of beam propagation and is preferably a non-diffracting laser beam, wherein the pulse energy is selected in such a way that the material modification generates material ejections on a surface of the material, and then the material is separated in a separating step along the separating line. According to the invention, the spatial spacing of the localized material modifications is greater than 0.2 to 10 times the diameter of the respective material ejections.

An ultra-short pulse laser provides the ultra-short laser pulses. Ultrashort laser pulses are understood to mean pulses in the picosecond range or in the femtosecond range.

The laser can also provide pulse trains made up of ultra-short sub-laser pulses, with each laser pulse train comprising the emission of a number of sub-laser pulses. In particular, so-called GHz laser pulse trains can also be provided. These pulse trains are also referred to as bursts.

The ultra-short laser pulses move over the material on a trajectory determined by the optics of the laser, with the trajectory following the desired dividing line, so that material modifications can be introduced into the material at a spatial distance from one another along the dividing line. The laser beam formed by the laser pulse and the material can be moved relative to each other with a feed in order to achieve the corresponding introduction of the material modifications along the parting line.

The laser beam and the material can be shiftable relative to each other with a feed.

Displaceable relative to one another means that both the laser beam can be displaced translationally relative to a stationary material and the material can be displaced relative to the laser beam, or both the material and the laser beam move. In particular, this allows the focus of the laser beam to be placed at different locations on the material in order to introduce material modifications. In addition to translational movements along the X, Y and Z axes, rotary movements are also possible, in particular rotations around the beam propagation direction. This makes it possible to orientate the laser beam along the dividing line.

The material can be a metal, or a semiconductor, or an insulator, or a combination thereof. In particular, it can also be a glass, a glass ceramic, a polymer or a semiconductor wafer, for example a silicon wafer. The material is transparent to the wavelength of the laser, with transparent meaning that 70% or more of the incident laser energy is transmitted through the material.

The laser pulse or laser pulses are at least partially absorbed by the material, for example by non-linear absorption processes, so that the material can be modified locally. A material modification is a permanent, material change in the material in thermal equilibrium, which is caused by the direct laser radiation.

The material modification can generally be a modification of the structure, in particular the crystalline structure and/or the amorphous structure and/or the mechanical structure, of the transparent material. For example, an introduced material modification of an amorphous material can consist in the material receiving a changed network structure through local heating only in this area. For example, the bond angles and lengths of the network structure can be changed by the modification. A material modification can in particular be a local change in density, which can also include areas without material, which can also be dependent on the selected material.

Depending on the specific material properties and the specific settings of the laser, such as pulse energy, pulse duration and repetition rate, different types of material modifications can occur. For example, with a first setting, the laser can provide a laser beam that results in an isotropic refractive index change in the material. With a further setting, however, the laser can also provide a laser beam which leads to a birefringent change in the refractive index in the material, so that the material has locally birefringent properties. In particular, so-called micro-explosions can occur at high pulse energies, in which highly excited, gaseous material is pushed out of the focal zone into the surrounding material and a less dense area or an empty core with surrounding compressed material is created. The size of the heated area is determined by the beam geometry, in particular by the focal zone of the laser beam and the beam cross section. The material modifications are introduced into the material by the local heat effect of the laser.

In contrast to material modification, the material modification area includes the entire area in which the effects of the laser pulse can be measured, for example using the tensile and compressive stresses. In particular, this is the area in which, spatially speaking, the material changes from the material modification back to the initial state of the untreated areas of the material.

As a result of the temperature gradients that arise as a result of the local heat effect, stresses can occur in the material modification area during heating and/or during cooling and formation of the material modification, which stresses promote crack formation. In particular, tensile and compressive stresses can arise in the material modification area, which run radially or orthoradially, for example. A material modification is therefore preferably accompanied by an indicated formation of cracks, i.e. targeted damage to the material.

Depending on the selected pulse energy, the material modification can generate material ejections on a surface of the material. The material ejections are a measure of the quality of the material modifications and thus also of the separability of the material.

Material ejections are accumulations of material on a surface of the material that occur around the location at which the laser pulses are introduced to produce a material modification. In particular, “a surface” means that it can be either the top or the bottom of the material relative to the beam propagation direction. Material ejections are consequences of the heating of the material, which protrudes from the volume of the material when the laser pulses are introduced, but part of the volume can also be lost through evaporation etc., so that there is no exact match between the displaced from the material and the around the Material modification around in the material ejections deposited material volumes must be present.

If the pulse energy is selected in such a way that no material is ejected, the material cannot be reliably separated because, for example, its volumetric strength is still too high. This can mean that the material modifications were not introduced deep enough into the material, or the material modification is too small due to the pulse energy being too low, or the type of modification, see above, is not suitable for cutting the material. However, if the pulse energy exceeds a certain limit, the spatial extent of the material modifications becomes too large, which leads to poor localization and poor surface quality of the interface. Material ejections in a certain size range are therefore desired as an indicator of an optimally introduced material modification and allow the determination of the optimal process parameters for the separation process.

The material modifications are introduced into the material along a desired parting line, so that material ejections occur at a distance along the parting line. A dividing line describes that line along which the material or parts of the material are to be separated or severed.

Due to the material modifications introduced along a dividing line in the material, the material is more or less perforated, so that a kind of predetermined breaking point is defined in the material by the dividing line. As a rule, however, this perforation does not lead to an independent separation of the material. Rather, the material modifications along the dividing line ensure, for example, a material weakening, so that when a subsequent separating step is applied, for example by applying a thermal stress and/or by applying a mechanical stress, preferably a tensile or bending stress, and/or by etching using at least one wet chemical solution, is separated along the dividing line.

A thermal stress can be achieved, for example, by heating the material along the parting line. For example, the parting line can be heated using a continuous wave CO2 laser, so that the material in the material modification area expands differently compared to the untreated material. As a result, the cracks favored by the material modification experience crack growth, so that a continuous and non-interlocked separating surface can form, separating the parts of the material from one another.

A tensile or bending stress can be generated, for example, by applying a mechanical load to the material parts separated by the dividing line. For example, a tensile stress can be applied if opposing forces act on the material parts separated by the dividing line in the material plane at a respective force application point, each pointing away from the dividing line. Are the forces not parallel respectively aligned antiparallel to each other, this can contribute to the development of bending stress. As soon as the tensile or bending stresses are greater than the binding forces of the material along the parting line, the material is separated along the parting line.

The material can also be separated by etching with a wet-chemical solution, with the etching process preferably starting the material at the material modification, i.e. the targeted material weakening. Since the material parts weakened by the material modification are preferably etched, this leads to a severing of the material along the severing line.

The material modifications have a spatial distance from one another along the dividing line which is greater than 0.2 to 10 times the diameter of the respective material ejections.

The spatial distance can be defined here as the distance between the centers of the material modifications. This can mean, for example, the geometric center of the material modification. The center of a material modification typically coincides with the center of the processing laser beam. Therefore, distance can also mean the spacing of the laser beams that have introduced the material modifications spatially spaced apart from one another along the parting line. However, it can also mean, for example, the distance between the centers of intensity of the respective beam cross sections.

The diameter of the material ejection can be the outside diameter of the material ejection along the parting line. In particular, the distance between the material modifications on the parting line can be related to the size of the material ejection. The diameter of the material spout can also be the minimum or the maximum diameter of the material spout. In particular, the material ejection by an elliptical jet can also be elliptical, see below. The diameter of the material chute is then the length along which the feed direction intersects the material chute. Thus, regardless of the orientation of the material modification in the material, further material modifications can be introduced into the material along a fixed distance.

For example, a round material modification with a diameter of 0.2 pm -1 pm can produce a round material ejection with an outside diameter of 1 pm - 5 pm. If the distance between the material modifications should be 5 times the diameter of the respective material ejections, then the distance between two adjacent material modifications is 5 pm - 25 pm. Due to the distance, which increases with the size of the material ejection, it can hereby be avoided that, if the spatial distance between the material modifications is too great, breakouts occur along the parting line when the material is separated. Furthermore, it can be avoided that the distance between the material modifications is too small in order to achieve an efficient machining process through the highest possible feed rate and avoidance of uncontrolled heat input into the material. In addition, a sufficiently large distance between the material modifications minimizes the risk of avoiding accumulation modifications, such as melting of the material on the surface.

In the embodiment mentioned, the material ejections also do not overlap, so that an undesired increase in strength in the area of the parting line due to material ejections connected to one another can be avoided.

The laser pulses can have a wavelength between 0.3 pm and 1.5 pm and/or the pulse length of the laser pulses can be 0.01 ps to 50 ps, preferably 0.3-15 ps and/or the mean power of the laser can be 10W to 1000W .

This has the advantage that the process can be optimized for the respective material over a large parameter range. In particular, this increases the probability of finding a laser wavelength available for a material at which the material is transparent.

The laser energy can be introduced into the material in the form of individual laser pulses, with the repetition rate of the individual laser pulses being 1 kHz to 2 MHz.

By introducing individual laser pulses, very localized processing of the material is possible. In particular, comparatively small material ejections occur here, so that the material modifications can be positioned close to one another. In this way, the material can also be reliably separated along separating lines with small curve radii.

The laser energy can be introduced into the material in the form of pulse trains comprising several sub-laser pulses, with the repetition frequency of the sub-laser pulses of the pulse train being between 100 MHz and 50 GHz, and with a pulse train preferably comprising 2 to 20 sub-laser pulses and /or the sum of the pulse energies of the sub-laser pulses of a pulse train can be between 10pJ and 300pJ per 100pm material thickness.

The sum of the pulse energy of the sub-laser pulses of a pulse train scales with the material thickness. The material thickness is the expansion of the material in the direction perpendicular to the surface. For example, in the case of a material with a material thickness of 100 pm, the sum of the pulse energies is 20 pJ in order to introduce a material modification with material ejection.

Accordingly, with a material thickness of 1000pm, the sum of the pulse energies is 200pJ. If the sum of the pulse energies is 200pJ and the pulse train consists of 10 sub-laser pulses, then each pulse has a sub-laser pulse energy of 20pJ.

By introducing a pulse train, it is possible to increase the material modification, so that the material ejection also increases. Since the distance between the material modifications is given by the diameter of the material ejection, it is thus possible to introduce the material modifications at a greater distance from one another. As a result, in particular a feed rate can be increased, so that the material processing can be carried out faster and more efficiently.

The repetition rate of the individual laser pulses or pulse trains can be freely selected or an individual pulse or pulse train can be emitted if required and the jitter or the start tolerance of the individual pulse or pulse train emission can be less than 100 ns.

Freely selectable can mean that the repetition rate is not fixed to a value. In particular, however, it can also mean that the repetition rate can be adjusted at any time. A freely selectable repetition rate makes it possible, in particular, to adapt the repetition rate to the shape of the dividing line. In particular, an overlapping of the material modifications can be avoided if the feed rate has to be reduced due to the process, for example with edges and corners in the parting line.

A jitter or a start tolerance of the single pulse or pulse train emission of less than 100 ns also allows a high positioning accuracy of the material modifications to be achieved, which corresponds to the product of the start tolerance and the feed rate.

For example, the start tolerance can be 50ns and the feed rate 100mm/s, so that a positioning accuracy of 5nm can be achieved.

The laser beam may have an elongated focal zone and is preferably a non-diffractive laser beam.

Non-diffracting rays obey the Helmholtz equation:

V ² U(x,y,z) + k ² U(x,y,z) = 0 and show a clear separability into a transversal and a longitudinal dependency of the

shape

U(x,y,z) = t/ _t (x,y) exp(i/c _z z) on. Here k=oj/c is the wave vector with its transversal and longitudinal components k ² =kz ² +kt ² and Ut(x,y) is an arbitrary complex-valued function that only depends on the transversal coordinates x,y. The z-dependence in the direction of beam propagation in U(x,y,z) leads to a pure phase modulation, so that the associated intensity I of the solution is propagation-invariant or non-diffractive:

I(x,y,z) = \U(x,y,z)| ² = I(x,y)

This approach provides different solution classes in different coordinate systems, such as Mathieu rays in elliptic-cylindrical coordinates or Bessel rays in circular-cylindrical coordinates.

Experimentally, a large number of non-diffracting beams can be realized to a good approximation, i.e. quasi non-diffracting beams. In contrast to the theoretical construct, these lead only to a finite performance. The length L of the propagation invariance of these quasi non-diffracting rays is also finite.

Based on the standard for laser beam characterization ISO11146 1-3, the beam diameter is determined using the so-called 2nd moments. The power of the laser beam or the 0th order moment is defined as:

P = f dx dy I(x,y).

The spatial moments of the 1 . order indicate the center of gravity of the intensity distribution and are defined as:

Based on the above equations, the spatial moments of the 2nd order of the transverse intensity distribution can be calculated:

With the spatial moments of the second order of the laser beam, which are completely defined in this way, the beam diameters in the main axes can be determined. The main axes here are the directions of the minimum and maximum extent of the transverse beam profile, which always run orthogonally to one another. The beam diameter d of the laser beam is then calculated as follows:

The focal zone d ^GF _x , _y of a Gaussian beam, the Gaussian focus, or the diameter of the Gaussian beam or the Gaussian profile, is defined by the second moments, i.e. the variance of the Gaussian curve, and the associated characteristic length, the Rayleigh length ZR =n(d ^GF _x , _y ) ² /4A, as the distance from the focal position at which the beam cross-section has increased by a factor of 2. In the case of a symmetric Gaussian beam, the focal zone has d ^GF o = d ^GF _x = d ^GF _y

Furthermore, we define the transverse dimensions of local intensity maxima as the transverse focus diameter for quasi-non-diffracting rays d ^ND o as twice the shortest distance between an intensity maximum and an intensity drop to 25% proceeding therefrom.

The longitudinal extent of the focal zone in the direction of beam propagation of these intensity maxima, which are almost propagation-invariant, indicates the characteristic length L of the quasi-non-diffracting beam. This is defined by the intensity drop to 50%, starting from the local intensity maximum in the positive and negative z-direction, i.e. in the direction of propagation. A quasi-non-diffracting ray is present if and only if for d ^ND o=d ^GF _x , _y , i.e. similar transverse dimensions, the characteristic length L clearly exceeds the Rayleigh length of the associated Gaussian focus, for example if L>10ZR.

As a subset of the quasi-non-diffracting rays, quasi-Bessel rays or Bessel-like rays are known. The transverse field distribution U*(x,y) in the vicinity of the optical axis obeys a Bessel function of the first kind of order n to a good approximation are widely used due to their ease of production. Thus, the illumination of an axicon in a refractive, diffractive or reflective design with a collimated Gaussian beam allows the formation of the Bessel-Gaussian beam. The associated transverse field distribution in the vicinity of the optical axis obeys a good approximation to a Bessel function of the first kind of order 0, which is enveloped by a Gaussian distribution.

Typical Bessel-Gauss beams, which can be used to process transparent materials, have a diameter of the central intensity maximum on the optical axis of d ^ND _x ,y=2.5 pm. The associated length L can easily exceed 1 mm. A Gaussian focus with d ^ND x,y=d ^GF _x ,y=2.5 pm, on the other hand, is characterized by a focus length in air of only ZR=5 pm at A=1 pm. In these cases, which are relevant for material processing, L»10ZR even applies to the Bessel-Gauss rays.

The extension do of the intensity maximum or the intensity maxima can be between 0.25 pm and 10 pm.

In particular, do can mean either d ^ND _x , _y or d ^GF _x , _y . This is determined by the dimension of the material ejections, since the size of the intensity maximum also approximately determines the size of the material modification.

If the intensity maximum d _x , _y exceeds a magnitude of 10 pm, then the pulse energy required to introduce a material modification begins to scale quadratically with the magnitude of the intensity maximum. Accordingly, if the size doubles, the required pulse energy will quadruple.

The non-diffractive laser beam can have a non-radially symmetrical beam cross section perpendicular to the beam propagation direction. A non-radially symmetrical beam cross section is present in particular when d _x is not equal to d _y , where d _x and d _y are defined via the second Moments of a transverse intensity distribution. In order to calculate the second moments, only intensity values of the intensity distribution that lie within a curve of the same intensity are taken into account, with the curve enclosing a local and/or global maximum of the intensity distribution. In particular, the intensity is 40% and in particular 60% of the maximum intensity of the local and/or global maximum.

The non-diffracting laser beam can have an enveloping non-radially symmetrical beam cross-section perpendicular to the beam propagation direction. An enveloping, non-radially symmetrical beam cross-section is present in particular when d _x is not equal to d _y , d _x and d _y being defined by the second moments of a transverse intensity distribution of the laser beam. In particular, only intensity values of the intensity distribution that are above an intensity threshold are taken into account for the calculation of the second moments.

For example, the intensity threshold is 17% of the maximum intensity of the global maximum of the intensity distribution.

The non-radially symmetrical beam cross section is preferably designed as an elliptical beam cross section.

The envelope of the beam cross section is preferably elliptical in shape.

However, a non-radially symmetrical beam cross-section can also mean that the beam cross-section is, for example, cross-shaped, or is triangular, or is N-sided, for example pentagonal. A non-radially symmetrical beam cross-section can also include further rotationally and mirror-symmetrical beam cross-sections.

In particular, the contours of the beam cross sections have locations with different radii of curvature. For example, in the case of an elliptical beam cross-section, the radius of curvature is particularly large at the point at which the semi-minor axis intersects the ellipse, while the radius of curvature is particularly small at the point at which the semi-major axis intersects the ellipse.

Elliptical, quasi non-diffracting beams have special properties that result from the analysis of the beam intensity. For example, elliptical quasi-non-diffracting rays have a main maximum, also referred to as a global maximum, which coincides with the center of the ray. The center of the beam is given by the place where the long axis intersects the short axis. In particular, elliptical quasi non-diffracting beams can result from the superimposition of multiple intensity maxima, in which If only the envelope of the intensity maxima involved is elliptical. In particular, the individual intensity maxima do not have to have an elliptical intensity profile.

The secondary maxima closest to the main maximum, which result from the solution of the Helmholtz equation, have a relative intensity of over 17%. Thus, depending on the transported laser energy in the main maximum, so much laser energy is also conducted in the secondary maxima that material processing is made possible. In addition, the nearest secondary maxima always lie on a straight line that is perpendicular to the long half-axis a or parallel to the short half-axis b and runs through the main maximum.

An elliptical quasi-non-diffractive beam can have a non-zero intensity along the long axis a, in particular an interference contrast lmax-lmin/(lmax+lmin)<0.9, such that the beam transports laser energy everywhere along the long axis a.

Here, Imax is the maximum beam intensity along the long axis a, while Imin is the minimum beam intensity. If lmin=0, then there is complete interference along the long axis a and an interference contrast of 1 . If lmin>0, then there is only partial or no interference along the long axis a, so the interference contrast is <1.

If, for example, the interference contrast along the long axis a is less than 0.9, there is no complete interference along the long axis a, but only partial interference, which does not lead to complete extinction of the laser intensity at the location of the intensity minimum Imin. This is the case, for example, when the quasi-non-diffracting beam is generated with a birefringent element based on quartz crystals. There are embodiments that produce an angular offset and embodiments that produce a spatial offset.

However, an elliptical quasi-non-diffractive beam can also have vanishing intensity along the long axis a and an interference contrast of 1, such that the beam does not transport laser energy everywhere along the long axis a. This is the case, for example, when the quasi-non-diffracting beam is generated with a modified axicon.

The focus zone, which is elongated in the beam propagation direction, can preferably project into the material at least in sections and penetrate at least one material surface. Protruding into the material in sections can mean that the focal zone lies partly in the volume of the material and partly outside of the material. If the focal zone penetrates the material surface, this can mean that the focal zone is intersected by the material surface, so that at least part of the laser energy penetrates the material.

In particular, the focal zone can penetrate a material surface so that, for example, laser energy is applied to the upper side of the material and into a part of the material volume, but no laser energy is applied to the underside of the material. A material modification is thus introduced on the upper side and the material is also ejected on the upper side. By treating or modifying the upper side and the lower side of the material differently, this can result in stress gradients in the material, which can be used to advantage when cutting the material.

However, it is also possible that the focal zone penetrates two opposite surfaces of the material. In particular, this means that laser energy is introduced into the material over the entire thickness of the material. Since the focal zone extends over the entire thickness of the material, each of the necessary material modifications can be generated in the material with a single laser pulse or a single laser pulse train. In other words, each material modification is introduced immediately throughout the entire thickness of the material, so that no further processing is necessary at the same position in order to introduce the material modification.

The focal zone elongated in the beam direction is preferably longer, preferably 1.5 times as long as the thickness of the material or 2 times 200 μm + the thickness of the material. In this way it can be achieved that the area of the highest intensity lies completely in the material and thus a particularly efficient processing of the material is achieved.

This results in material ejections on the top and bottom of the material. This leads to a lower required separating force in the subsequent separating process and thus to a lower surface roughness of the separating surface.

At the same time, because the focal zone extends over the entire thickness of the material, a high focal position tolerance can also be provided, so that the method can be constructed in a particularly robust manner. The long axis of the non-radially symmetrical beam cross section can preferably be oriented perpendicular to the beam propagation direction along the dividing line.

The special shape of the focal zone, in particular the non-radially symmetrical beam cross-sections of the quasi-non-diffracting beams and elliptical, quasi-non-diffracting beams, are transferred to the shape of the material modification in the material. Since the shape of the material modification determines the shape of the material ejection, material ejection occurs on the surface, which corresponds in shape to that of the jet cross section. For example, in the case of an elliptical jet cross section, elliptical material ejections form.

Typically, crack formation is induced in the material along a preferred direction of the non-radially symmetrical beam cross-section - for example, crack propagation mainly takes place in the direction of a longer extension of the beam cross-section, which is associated with smaller radii of the contour of the beam cross-section at the outer contour edges lying in this preferred direction.

This is due, among other things, to the targeted crack formation that emanates from the material modification or modifications. The material modification is introduced into the material in such a way that the resulting crack formation preferably extends to the adjacent material modification. In particular, it is also possible for the cracks, starting from two adjacent material modifications, to meet between the material modifications and thus form a common crack. This makes it possible to move the positions at which the material modification is introduced into the material further apart.

Oriented along the parting line means that the material modification lies on the parting line and the long axis of the material modification runs at least in sections tangentially to the desired parting line. For example, this can mean that in the case of an elliptical material modification cross section, the center of the ellipse lies on the desired parting line and the long axis a in the center of the ellipse lies tangential to the parting line.

This ensures that the crack propagation runs at least in sections along the desired parting line.

In particular, the long axis can also be oriented parallel to the feed rate vector or to the feed rate. Because the feed rate is always tangential to the parting line, this leads to crack formation along the parting line, with the long axis of the material ejection also always being oriented tangential to the parting line. Accordingly, the distance between the material modifications along the parting line can be selected to be large, so that the material can be processed efficiently.

The individual laser pulses and/or the pulse trains can be triggered by a position-controlled pulse release from the laser, the position preferably being given by the position of the laser beam on the material.

Position-controlled pulse triggering means that when the laser beam reaches a point or a distance on the dividing line or when it reaches a point in space relative to a predetermined reference value, a laser pulse or a laser pulse train is emitted by the laser.

For this purpose, for example, via an axis encoder of a feed device, with which the laser beam and the material are moved relative to each other, the current position of the feed device can be read and, for example, fed to a control device, which in turn processes the position and sends a trigger signal to the laser control of the ultrashort pulse laser .

By also processing the feed rate and the feed direction and thus the dividing line in the control device, the laser pulses or laser pulse trains can be emitted automatically.

In particular, the base frequency of the laser can thus provide the clock of the control device with a common time base, so that the control device can use the base frequency and the feed rate to cause the laser pulses or laser pulse trains to be emitted based on distance.

Since the laser pulses or pulse trains are emitted in a position-controlled manner, there is no need for time-consuming programming of the modification process. In addition, freely selectable process speeds can be easily implemented.

The support surface on which the material is placed is preferably neither reflective nor absorbent nor strongly scattering for the wavelength of the laser pulses. This prevents laser radiation from being reflected or scattered back into the material and heating, processing or modifying it in an uncontrolled manner. In particular, the bearing surface can thus also be transparent to the wavelength of the laser.

Brief description of the figures

Preferred further embodiments of the invention are explained in more detail by the following description of the figures. show:

FIG. 1 A, B, C shows a schematic representation of how the method is carried out;

FIG. 2A, B shows a microscope image and cross section of a material modification

Material ejection:

FIG. 3A, B, C, D shows a schematic representation of beam cross sections of quasi-non-diffracting beams;

FIG. 4A, B, C, D an analysis of the beam cross-sections of quasi-non-diffracting beams;

Figure 5 is a schematic representation of a composite elliptical quasi-non-diffracting beam;

FIG. 6A, B shows a further schematic representation for carrying out the method;

FIG. 7A, B, C, D shows a schematic representation of elliptical beam cross-sections and material modifications and their orientation at a parting line; and

Figure 8A, B is a schematic representation of the device for performing the

procedure.

Detailed description of preferred exemplary embodiments

Preferred exemplary embodiments are described below with reference to the figures. Elements that are the same, similar or have the same effect are provided with identical reference symbols in the different figures, and a repeated description of these elements is sometimes dispensed with in order to avoid redundancies.

FIG. 1 shows schematically the separation process described here for separating a material

1 shown. In order to separate the material 1, laser pulses 60 of an ultrashort pulse laser 6 in the material 1 focuses. The laser pulses, which are at least partially absorbed by the material 1 in the focal zone 600 of the laser beam, run in the laser beam 60 in order to introduce a material modification 3 into the material 1 in this way. The hatched plane here shows the plane below the dividing line 2 along which the material 1 is separated. Ideally, this level corresponds to the future parting line.

The linear and/or non-linear absorption of the laser pulses in the material 1 can result in so-called micro-explosions, in which the material 1 is abruptly vaporized in the focal zone 600 of the laser beam. The highly excited, gaseous material 1 is pushed into the surrounding material 1 by the high pressure, so that the material 1 is compressed at the shock front. In the area of the focal zone 600, a less dense or empty core (engl. “void”) arises, which is surrounded by the compacted material. In particular, as a result of the micro-explosions, part of the material from the focal zone 600 can penetrate to the outside, where it is deposited on the surface of the material 1 and the material ejections 300 form.

These modifications result in the material modification 3. A material modification region 30 is formed around the material modification 3. In the material modification region 30, the material gradually returns to its original state from the state that is present in the material modification 3, the further away the material is viewed from the material modification 3. The original state can be, for example, the unprocessed state of the material, which is present, for example, in neighboring points in material 1. However, the original state is also understood to mean the state of the material 1 that existed before the material modification 3 was introduced.

Depending on the selected pulse energy of the ultra-short laser pulses, material ejections 300 can form on the material 1 . In this case, the material ejections 300 are a measure of the quality of the material modification 3 and thus also of the separability of the material 1 .

In particular, the material ejections 300 are caused by the heating of the material 1 in the focus zone 600, the heated material emerging from the material 1 due to thermal expansion and being deposited next to the material modification 3

For this purpose, the laser pulses can have a wavelength between 0.3 pm and 1.5 pm and/or the pulse length of the laser pulses can be 0.01 ps to 50 ps, preferably 0.3-15 ps and/or the average power of the laser can be 10 W to 1000 amount to W. The laser energy can be introduced into the material in the form of individual laser pulses, with the repetition rate of the individual laser pulses being 1 kHz to 2 MHz. However, the laser energy can also be in the form of pulse trains, comprising several sub-laser pulses, are introduced into the material, wherein the repetition frequency of the sub-laser pulses of the pulse train can be between 100MHz and 50GHz, further wherein a pulse train can preferably comprise 2 to 20 sub-laser pulses and/or the sum of the pulse energies of the sub - Laser pulses of a pulse train can be between 10pJ and 300pJ per 10Opm material thickness.

For example, a material modification 3 with material ejections 300 can be produced using a laser with a wavelength of 1 pm, a pulse duration of 1 ps and an average power of 100W. The laser pulse can be introduced into the material 1 in the form of a single pulse, the repetition rate of the laser being 100 kHz, for example.

The spatial distance between the material modifications 3 is in particular greater than 0.2 to 10 times the diameter of the respective material ejection 300.

In the material modification 3 and the material modification area 30, local stresses can occur that promote crack formation. For example, the material 1 can have a different density—for example, a lower density—due to local heating, and a compressive stress can thereby build up in the material modification region 30 . However, a higher density can also be present in the heated area and thus a tensile stress can be built up in the material modification area 30 . If the tensile and/or compressive stress becomes too great, for example greater than the tensile or compressive strength of the untreated material, a crack may form spontaneously.

As shown in FIG. 1, several material modifications 3 are introduced into the material 1. Material modification regions 30 form around each material modification 3 . The material modifications 3 are placed along the desired parting line 2. The parting line 2 is an imaginary line along which the material 1 is to be separated.

The material modifications 3 introduced into the material along a dividing line 2 virtually perforate the material, so that the dividing line 2 defines a type of predetermined breaking point in the material 1 . As a rule, however, this perforation does not lead to an independent separation of the material 1. Rather, the material modifications 3 along the dividing line 2 ensure, for example, a targeted material weakening and/or a targeted introduction of cracks, which cause a material weakening along the dividing line 2. After the material modifications 3 have been introduced into the material 1 by means of the laser beam 6, the material 1 can be physically separated in a subsequent separating step, for example by applying a tensile force FZ to the material halves 10 and 12 separated from one another by the separating line 2. In particular, it is also possible to separate the material 1 by applying a bending stress to the material halves 10, 12 (not shown).

FIG. 1B shows an analogous method in which the halves of the material are not separated in a separating step with mechanical stress, but by applying thermal stress.

After the material modifications 3 have been introduced, a thermal gradient 620 can be generated over the material modifications 3 . For example, a continuous wave CO2 laser 62 may be used to introduce the thermal gradient 620 .

The focal zone of the continuous wave CO2 laser 62 can be placed a few micrometers below the surface to generate the thermal gradient 620, so that the material is separated with little damage and a smooth fracture edge or separation surface is created.

By positioning the focal zone of the continuous wave CO2 laser 62, the temperature at the top surface of the material 1 is higher than at the bottom surface. This creates a thermal gradient T(z). Due to the thermal expansion of the material 1, which is linear in terms of temperature in a first approximation, the material 1 expands more on the upper surface than on the lower surface. This results in different material stresses along the Z-axis.

The various material stresses run through the introduced material modifications 3. The material stresses can preferably relax there, which leads to the formation of cracks. The cracking takes place between the different adjacent material modifications 3 . This leads to crack formation, which finally separates the material 1 into the two material halves 10 and 12 .

Another analogous method is shown in FIG. 1C, in which the material halves are separated in a separating step by means of a wet-chemical reaction. For this purpose, the material 1 perforated with the material modifications 3 is placed in a chemical bath 11 . In this case, the chemical bath 11 contains a solvent which is able to remove and etch the material 1 . In particular, the etching process takes place on the previously introduced Material modifications 3 instead, because there the material weakening is particularly large and the change in the physical and / or chemical properties at the site of the material modification allows the reaction to take place particularly advantageously. To a certain extent, a material modification 3 can act as a catalyst of the etching reaction. The reaction is shown in FIG. 1C by the formation of reaction bubbles 110 in the chemical bath 11.

As soon as the material 1 has been etched through, the material 1 is separated into the two material halves 10,12. If the material 1 has not yet been separated after the chemical bath 11, for example because the chemical bath 11 has only etched away the material modifications 3, the material 1 has been specifically damaged further along the separating line 2, so that, for example, by applying a tensile or Bending stress the material 1 in the material halves 10, 12 can be separated.

In FIG. 2A, a microscopic image of the surface of a processed material 1 is shown. Round material modifications 3 were introduced into the material 1 along the parting line 2 at a distance dM=5 pm. The material modifications 3 are in the form of a perforated channel, with the material of the outer lateral surface of the perforated channel being compacted by microexplosions when the material modification 3 was introduced. Around the round opening of the material modification 3, or the hole channel, round material ejections 300 arise on the surface of the material 1. These material ejections 300 have an outer diameter dA. The outer diameter of the material ejections 300 is 3 pm here. According to the method, the material modifications can thus be spaced apart from one another by between 0.6 pm and 30 pm. In particular, the material modifications 3 shown could be introduced into the material 1 when using quasi-non-diffracting rays by means of a pulse or a pulse train.

In Figure 2B, a thickness cross-section through Figure 2A is shown. It can be clearly seen that the material ejections have a height above the surface of the material 1 of 50 nm to 200 nm. The diameter and the height of the material ejections 300 are specified here by the pulse energy and the beam cross section of the laser beam.

FIG. 3A shows the intensity profile and beam cross section of a quasi-non-diffracting laser beam. In particular, the quasi-non-diffracting beam is a Bessel-Gaussian beam. In the beam cross-section in the x-y plane, the Bessel-Gaussian beam has radial symmetry, so that the intensity of the laser beam only depends on the distance from the optical axis. In particular, the transverse beam diameter d ^ND o is between 0.25 pm and 10 pm. FIG. 3B shows the longitudinal beam cross section, ie the beam cross section in the direction of beam propagation. The beam cross-section has an elongated focal zone that is about 3mm in size. The focus zone in the direction of propagation is therefore significantly larger than the beam cross section, so that there is an elongated focus zone.

Analogously to FIG. 3A, FIG. 3C shows a non-diffracting beam which has a non-radially symmetrical beam cross-section. In particular, the beam cross-section appears stretched in the y-direction, almost elliptical.

FIG. 3D shows the longitudinal focal zone of the Bessel beam, which again has an extension of about 3 pm. Accordingly, the Bessel beam also has a focal zone that is elongated in the direction of beam propagation.

FIG. 4 shows a detailed analysis of the beam cross section from FIG. 3C,D. FIG. 4A shows the transversal intensity distribution of the laser beam 60, the main maximum and the secondary maxima resulting from the solution of the Helmholtz equation.

FIG. 4B shows the so-called iso-intensity lines of the intensity distribution from FIG. 4A, the lines being drawn in where the relative intensity of the laser beam is 25%, or 50%, or 75%. It is clearly visible that the main maximum 41 of the intensity distribution has an approximately elliptical shape, with the extension along the x-axis being significantly larger than the extension along the y-axis. In particular, the main maximum is followed by two kidney-shaped secondary maxima 43, which have a significantly lower relative intensity.

FIG. 4C shows a cross section through the intensity distribution from FIG. 4A through the center of the main maximum along the x-axis. The intensity distribution shows its maximum in the center of the main maximum, with the relative intensity here by definition being 100%. The intensity distribution decreases along the positive and negative x-direction until a minimum in the relative intensity distribution is reached at about 0.003 mm, which, however, differs from 0%. Accordingly, laser energy is also transported between the main maximum 41 and the secondary maximums 43 of the laser beam 60 .

FIG. 4D shows a cross section through the intensity distribution from FIG. 4A through the center of the main maximum 41 along the y-axis. Again, the intensity maximum can be found in the center, but the intensity drop is much faster along the y-direction, so that the intensity minimum is reached at about 0.002 mm. In this case, the intensity minimum is exactly zero, since there is complete interference for the laser beam 60 here. In particular, with larger values on the y-axis, secondary maxima can again be found which, for example, lie above a relative intensity value of 25%. This is not the case in the x-axis cross section from FIG. 4C. The properties of the elliptical beam cross section thus differ along the different propagation directions.

In particular, it is shown in FIGS. 4C and 4B that the long semi-axis a is measured from the center of the main maximum to the drop in the relative intensity to 50%. Analogously, the length of the short semi-axis b is measured from the center of the main maximum until the relative intensity drops to 50%. Here, the long and short semi-axes are perpendicular to each other.

FIG. 5 shows that elliptical, quasi non-diffracting beams can result from the superimposition of a plurality of intensity maxima, in which case only the envelope of the intensity maxima involved is elliptical. In particular, the individual intensity maxima do not have to have an elliptical intensity profile.

In the present case, the beam cross section also has two kidney-shaped secondary maxima 43 in addition to the pronounced main maximum 41 . At least 25% of the laser energy of the main maximum 41 is transported in the secondary maxima. If the laser pulse energy is high enough, the laser pulse energy transported in the secondary maxima 43 is also sufficient to bring about a material modification. The focal zone within which material processing takes place is thus given by the 25% iso intensity line instead of a 50% iso intensity line.

The main maximum and the two secondary maxima each form, for example, overlapping material modification regions 30 so that an elliptical material modification 3 results overall, the long axis of which extends in the y-direction. Crack formation along the y-direction is therefore to be expected.

In particular, this will also result in an elliptical material ejection, the long axis of which is aligned analogously along the y-axis.

FIGS. 6A, B show that the elongated focal zone can be introduced into the material 1 in different ways. In Figure 6A, the elongated focal zone has a larger one Length up than the material is thick. This makes it possible to position the focal zone 600 such that the focal zone 600 penetrates the top surface and the bottom surface. This makes it possible, in particular, for the material modification 3 to be introduced over the entire material thickness. Material ejections 300 can also occur on the upper side and the underside of the material 1 . This leads to a lower required separating force in the subsequent separating process and thus to a lower surface roughness of the separating surface.

FIG. 6B shows that the focal zone 600 can also be introduced into the material 1 in such a way that only one material surface is penetrated by the elongated focal zone. In the present case the upper surface is penetrated. In particular, material ejection 300 can only occur on the upper surface, since no material modification 3 is produced on the underside of the material.

An elliptical material modification 3 in a material 1 is shown in FIG. 7A. The material modification 3 is introduced into the material 1 by the laser beam 60 of the laser 6 . In this case, the shape of the material modification 3 is predetermined by the beam cross section 4 of the laser beam 60 . A material modification area 30 is formed around the area of the material modification 3, in which the laser beam 60 has a direct effect on the material 1 for the time of the laser pulse. With a suitably selected pulse energy, a material ejection forms on the surface of the material 1 300, which corresponds in the form of the introduced material modification 3, or the beam cross section 4 of the laser beam.

Accordingly, material stresses can occur both in the material modification 3 itself and in the material modification area 30, which promote crack formation. For example, in the case of an elliptical material modification, cracking can be promoted at the points on the ellipse where the curve radius of the boundary line is particularly small. A small curve radius ensures that the stress introduced into the glass 1 by the material modification 3 can drop particularly quickly in many different directions. There is therefore a higher probability of relaxation of the material stress at this point than at locations where the material stress can only relax in a few directions. As a result, the points of the material modification 3 in the material 1 that have a small curve radius are particularly unstable.

The formation of the crack 32 then takes place preferentially in the direction of the long axis of the elliptical

Material modification 3 instead. Thus, it is possible the crack propagation through the orientation of the Material modification 3 to control. So it is possible in particular the crack propagation of a

Material modification 3 to another material modification 3 to control.

Several material modifications 3 have been introduced into the material 1 in FIG. 7B. The material modification 3 are again elliptical. As a result, the cracks form particularly preferably along the long axis of the ellipse at the points of the smallest radii of curvature of the ellipse. The material modifications 3 are placed so close together in the figure that the respective cracks of adjacent material modifications overlap. This allows the cracks to coalesce and form a common crack between two adjacent material modifications. In particular, this state can be achieved by crack growth, such as by applying a tensile force. For example, cracks can be introduced into the material 1 along any parting lines 2 by this method.

In FIG. 7C it is shown that the long axes of the material modifications 3 and the material ejections 300 are aligned along the parting line 2 . The diameter of the material ejection 300 above the dividing line 2 can be used here, for example, to determine the distance between the material modifications 3 . Since the long axes of the material modifications 3 are aligned along the dividing line 2, this means at the same time that the long axis of the beam cross section of the laser beam 60 was aligned along the dividing line 2 when the material modifications 3 were introduced.

Accordingly, FIG. 7D shows that the long axis of the beam cross section 4 is aligned parallel to the feed rate V, so that the long axis is always aligned parallel to the dividing line 2 .

A structure for carrying out the method is shown in FIG. 8A. The laser beam 60 of the ultra-short-pulse laser 6 is directed onto the material 1 by beam-shaping optics 9 and an optional mirror 70 . The material 1 is in this case arranged on a support surface of the feed device, the support surface neither reflecting nor absorbing the laser energy which the material does not absorb nor strongly scattering it back into the material 1 . The material of the bearing surface is therefore preferably also transparent to the wavelength of the laser.

In particular, the laser beam 60 can be coupled into the beam shaping optics 9 through a free space section with a lens and mirror system. However, the laser can also be coupled into the beam shaping optics through a hollow-core fiber 65 with coupling and decoupling optics, as shown in FIG. 8B. The beam-shaping optics 9 can be, for example, a diffractive optical element or an axicon, which generates a non-diffracting laser beam 60 from a Gaussian laser beam 60 . In the present example, the laser beam 60 is directed by the mirror 70 in the direction of the material 1 and is focused onto or into the material 1 by a focusing optics 72 . The laser beam 60 causes material modifications 3 in the material 1. In particular, the beam cross section of the laser beam is retained by the mirror 70 and the focusing optics 72.

In this case, the feed device 8 can move the material 1 under the laser beam 60 with a feed rate V, so that the laser beam 60 introduces material modifications 3 along the desired parting line. In particular, in the figure shown, the feed device 8 is composed of a first part 80 which can move the material along an axis. In particular, the feed device can have a second part 82, which is set up to rotate the laser beam 60 about the z-axis, or about the beam propagation direction, so that the long axis of the beam cross section perpendicular to the beam propagation direction is always tangential to the desired dividing line 2, in order to To cause crack propagation along the parting line 2.

For this purpose, the feed device 8 can be connected to a control device 5 , the control device 5 converting the user commands of a user of the device into control commands for the feed device 8 . In particular, predefined cutting patterns can be stored in a memory of the control device 5 and the processes can be automatically controlled by the control device 5 .

The control device 5 can in particular also be connected to the laser 6 . The control device 5 can set the laser pulse energy of the laser pulses of the laser 6, or request or trigger the output of a laser pulse or laser pulse train. The control device 5 can also be connected to all the components mentioned and thus coordinate the material processing.

In particular, a position-controlled pulse triggering can be implemented in this way, with an axis encoder of the feed device 8 being read out, for example, and the axis encoder signal being able to be interpreted by the control device as location information. It is thus possible for the control device 5 to automatically trigger the delivery of a laser pulse or laser pulse train if, for example, an internal adder unit that adds the distance covered reaches a value and resets it to 0 after it has been reached. For example, a laser pulse or laser pulse train can be emitted automatically into the material at regular intervals. Because the feed speed and the feed direction and thus the dividing line 2 are also processed in the control device, the laser pulses or laser pulse trains can be emitted automatically.

The control device can also calculate a distance dM or location based on the measured speed and the fundamental frequency provided by the laser 6, at which a

Delivery of a laser pulse train or laser pulse is to take place.

Since the laser pulses or pulse trains are emitted in a position-controlled manner, there is no need for time-consuming programming of the cutting process. In addition, freely selectable process speeds can be easily implemented. As far as applicable, all individual features that are presented in the exemplary embodiments can be combined with one another and/or exchanged without departing from the scope of the invention.

material

10 first half of material

12 second half of material

2 dividing line

20 parting surface

3 material modification

30 material modification area

300 material ejection

32 crack

4 beam cross-section

41 main order

43 bylaws

5 control device

6 lasers

60 laser beam

600 focus zone

62 continuous wave CO2 laser

620 temperature gradient

65 hollow core fiber with coupling and decoupling optics

7 focusing unit

70 mirrors

72 focusing optics

8 feed device

80 first part of the feed device

800 bearing surface

82 second part of the feed device

9 beam shaping optics

11 chemical bath

110 reaction bubbles

D Thickness of the thinnest glass dA Outside diameter of the material ejection dM Distance of the material modifications FZ tensile stress

Claims

Expectations

1 . Method for separating a transparent material (1), in which localized material modifications (3) are introduced into the material (1) along a separating line (2) and spatially spaced apart from one another by means of ultra-short laser pulses from an ultra-short-pulse laser (6), the laser beam (60) producing a has a focal zone (600) that is elongated in the direction of beam propagation and is preferably a non-diffracting laser beam, the pulse energy being selected such that the localized material modifications (3) each produce material ejections (300) on a surface of the material (1), and then the material (1) is separated in a separating step along the separating line (2), characterized in that the spatial distance (dM) of the localized material modifications (3) is greater than 0.2 to 10 times the diameter (dA) of the respective material ejections (300) is.

2. The method according to claim 1, characterized in that the separating step comprises the application of a thermal stress and/or the application of a mechanical stress, preferably a tensile or bending stress, and/or etching by means of at least one wet-chemical solution.

3. The method according to any one of claims 1 or 2, characterized in that

- the laser pulses have a wavelength between 0.3 μm and 1.5 μm, and/or

- the pulse length of the laser pulses is 0.01 ps to 50 ps, preferably 0.3-15 ps, and/or

- the average power of the laser (6) is 10W to 1000W.

4. The method according to claim 3, characterized in that the laser energy is introduced into the material (1) in the form of individual laser pulses, the repetition rate of the individual laser pulses being 1 kHz to 2 MHz.

5. The method according to claim 3, characterized in that the laser energy in the form of pulse trains comprising a plurality of sub-laser pulses are introduced into the material (1), the repetition frequency of the sub-laser pulses of the pulse train being between 100 MHz and 50 GHz, further wherein a pulse train preferably comprises 2 to 20 sub-laser pulses, and/or - the sum of the pulse energies of the sub-laser pulses of a pulse train is between 10pJ and 300pJ per 100pm material thickness.

6. The method according to any one of claims 4 to 5, characterized in that the repetition rate of the individual laser pulses or pulse trains can be freely selected or an individual pulse or pulse train is emitted if required and the jitter or the start tolerance of the individual pulse or pulse train emission is less than 100 ns.

7. The method according to any one of the preceding claims, characterized in that the extent dO of the intensity maximum or the intensity maxima is between 0.25 pm and 10 pm.

8. The method according to any one of the preceding claims, characterized in that the laser beam (60) has a non-radially symmetrical beam cross section perpendicular to the beam propagation direction, wherein the beam cross section or the envelope of the beam cross section is preferably elliptical in shape.

9. The method as claimed in one of the preceding claims, characterized in that the focal zone (600) which is elongated in the direction of beam propagation protrudes at least in sections into the material (1) and penetrates at least one material surface.

10. The method according to any one of claims 8 to 9, characterized in that the long axis of the non-radially symmetrical beam cross-section is oriented perpendicular to the beam propagation direction along the dividing line (2).

11 . Method according to one of the preceding claims, characterized in that the individual laser pulses and/or pulse trains are triggered by position-controlled pulse triggering from the laser (6), the position preferably being given by the position of the laser beam (60) on the material (1).

12. The method according to any one of the preceding claims, characterized in that the bearing surface on which the material (1) rests is designed to be neither reflective nor absorbent nor strongly scattering for the wavelength of the laser pulses. Method according to one of the preceding claims, characterized in that the laser beam (60) and the material (1) are moved relative to one another with a feed.