Method and device for inserting a hollow-channel-shaped filament into a brittle fracturing workpiece, and use of same

Abstract

The invention relates to a method and a device for inserting a hollow-channel-shaped filament (1) into a brittle fracturing workpiece (2), in which a focused laser beam (4) of a pulsed laser (3) with a wavelength λ is directed onto the workpiece for processing. The workpiece is transparent for the wavelength λ of the laser, and the filament (1) is generated by the irradiation of at least one burst packet of laser pulses, wherein the burst packet has a burst duration Tb and comprises a sequence of ultra-short laser pulses which have an individual pulse energy Ep and an intra-burst delay Ti. The individual pulse energy Ep0 of the laser pulse at the start of the burst duration Tb is greater than the individual pulse energy Epx of the laser pulse at the end of the burst duration. The invention also relates to advantageous uses.

Description

[0001] Method and device for inserting a hollow channel-shaped filament into a brittle workpiece and its use

[0002] The invention relates to a method for inserting a hollow-channel-shaped filament into a brittle workpiece. A device for producing a hollow-channel-shaped filament in such a workpiece is also subject to the invention, as is the application of the method and / or device.

[0003] The invention uses laser-assisted processes for this purpose. Brittle-breaking workpieces generally include workpieces consisting of, or at least the filament region of, a brittle-breaking material. For the purposes of the invention, these include, in particular, glass, glass ceramics, silicon, etc.

[0004] In laser processing, a filament is generally referred to as an elongated, thread-like damage to the material of the workpiece, which occurs as a result of laser irradiation.

[0005] A well-known application of filamentation is laser cutting. This involves inducing a damage into the glass. If several damages are then placed side by side, a dividing line is created along which the glass can be separated by applying tension.

[0006] The basic mechanism is that microcracks originating in the filament weaken the material. If damage occurs adjacent to each other, the cracks connect, allowing separation along this connecting line.

[0007] WO 2012 / 006736 A2 describes a laser cutting process based on the nonlinear optical Kerr effect. It is based on two nonlinear effects: the self-focusing of the pulse due to the optical Kerr effect and its defocusing due to the plasma generated by the laser in the material. It is stated that the self-focusing is maintained over a focal length of approximately 500 to 1000 m and then spatially disperses when the pulse energy is insufficient to refocus itself and create a new plasma channel.

[0008] EP 3169635 A1 uses a Bessel beam with an axicon optic to extend the laser focus line during laser cutting. The inventors of this document argue that the difference from a method based on the Kerr effect, such as WO 2012 / 006736 A2, is that refocusing the laser beam according to WO 2012 / 006736 A2 requires a modification of the refractive index of the workpiece material, in this case glass. This inevitably results in a different damage pattern, which can be undesirable.

[0009] WO 2018 / 189080 A1 uses a double axicon to focus a Bessel beam with a long laser focus line into a workpiece. This optical element allows the use of high-energy laser beams.

[0010] WO 2018 / 210746 A1 describes a laser cutting process in which the laser focus line of an ultrashort pulse laser system is produced using an optical arrangement with chromatic aberration.

[0011] A common feature of the aforementioned prior art is that the microcracks extend radially around the produced filament and thus also penetrate the volume of the workpiece, even at right angles to the intended cutting line. These microcracks can then reduce the strength of the cut edge.

[0012] The object of the invention is therefore to control microcracking during filamentation. This object is achieved by the inventive method and the inventive device according to the independent claims. Preferred embodiments are set out in the dependent claims.The inventors have discovered that a hollow channel-shaped filament is inserted into a brittle-breaking workpiece by means of a method in which a focused laser beam of a pulsed laser with a wavelength A is directed onto the workpiece for processing, the workpiece being transparent to the wavelength A of the laser, the filament being produced by irradiating it with at least one burst packet of laser pulses, the at least one burst packet having a burst duration Tb and comprising a sequence of ultrashort laser pulses having a single pulse energy Ep and an intra-burst delay Ti, the single pulse energy of the laser pulse at the beginning of the burst duration Tb being greater than the single pulse energy of the laser pulse at the end of the burst duration. A hollow channel-shaped filament can be produced by using a plurality of burst packets. It is advantageous if each hollow channel-shaped filament is produced by a single burst packet.If a plurality of burst packets are used to produce a hollow-channel-shaped filament, it is sufficient for the purposes of this description if a single burst packet, in particular the first one, has the aforementioned property that the first laser pulse of the burst packet has a higher energy than the last. This applies particularly advantageously to all respective burst packets. Within a burst packet, it is especially advantageous that the individual pulse energy of each individual pulse is lower than that of the previous individual pulse.

[0013] The hollow-channel filament is characterized by the hollow space along the filament axis. Such a hollow channel goes beyond mere material damage and / or refractive index changes and enables efficient further processing of the workpiece, for example, cutting or further expanding the filaments.

[0014] A filament is generally understood to be an elongated, thread-like structure whose diameter is significantly smaller than its length.

[0015] A brittle-breaking workpiece within the meaning of the invention comprises or consists of a brittle material, at least in the region of the filament. A brittle material can only be plastically deformed to a limited extent and is therefore characterized by low ductility. Brittle fracture occurs at low elongation and usually close to the yield point. A brittle material and / or a brittle-breaking workpiece within the meaning of the invention includes, in particular, glass, glass ceramic, and / or silicon.

[0016] To insert the hollow-channel filament into the workpiece, the focused laser beam of a pulsed laser is directed onto the workpiece. The laser emits the laser beam at a wavelength A, whereby the workpiece is transparent to the laser's wavelength A, at least in the area of ​​incidence of the laser beam. This means that the transmission of the workpiece at this point is greater than 85%, advantageously greater than 90%, and particularly advantageously greater than 95% at wavelength A.

[0017] To still achieve interaction of the laser beam with the workpiece material, an ultrashort pulse laser is used according to the invention. With such a laser, the pulse width is very short, i.e., in the range of less than 1 ns, in particular in the range of a few ps to fs, in particular in the range of 100 ps or 10 ps to 100 fs or 10 fs. The range from 10 ps to 200 fs is particularly advantageous.

[0018] To efficiently provide sufficient energy to produce the filament, even with industrially available laser sources, the filament is created by irradiating at least one burst of laser pulses. A burst consists of a sequence of ultrashort laser pulses, also referred to as individual pulses. A burst therefore contains at least two individual pulses. Typically, however, approximately 2 to 20 pulses are used, with 3 to 10 being advantageous, although upper limits of around 100 individual pulses are also possible.

[0019] The pulse duration of a single pulse corresponds to the pulse duration of the laser. The duration of the burst packet, called the burst duration Tb, depends on the pulse duration of the single pulse, the number of single pulses in the corresponding burst packet, and the temporal spacing of the single pulses, called the intra-burst delay Ti, measured at the time of the maximum intensity of a single pulse. Each single pulse has a single pulse energy Ep, which corresponds to the integral of the intensity curve of a single pulse over the single pulse duration and, at a constant single pulse duration, correlates with the maximum intensity of a single pulse in the burst packet.

[0020] According to the invention, the single pulse energy of the laser pulse at the beginning of the burst duration Tb is greater than the single pulse energy of the laser pulse at the end of the burst duration.

[0021] Until now, it was assumed that the individual pulses of the burst packet were intended to accumulate as much energy as possible in the workpiece. The inventors realized that it is not necessary to apply as much energy as possible, but that a temporally controlled energy input can, on the one hand, efficiently produce a hollow-channel-shaped filament, while, on the other, controlling its mechanical interaction with the surrounding workpiece material.

[0022] An advantageous embodiment of the method provides that the burst packet has a burst degression of 40% to 90%, advantageously 50% to 70%, particularly advantageously 60% to 70%.

[0023] The burst degression BG considers the ratio of the single pulse energy Ep xof the last single pulse of the burst packet at the end of the burst duration and the single pulse energy of the first single pulse Epo of the burst packet at the beginning of the burst duration in percent. The burst degression BG is therefore the quotient of Epx and Epo, preferably expressed as a percentage. The formula applies accordingly. For example, if Epo is 100 units and Ep x 60 units, BG = 0.6 = 60%

[0024] The individual pulse energy is significantly reduced at the end of the burst duration compared to the first individual pulse. This allows for fewer and / or shorter microcracks around the filament extending from the filament into the workpiece. This helps prevent unintentional weakening of the filamented workpiece.

[0025] In particular, it has proven advantageous if the method is designed in such a way that from the second individual pulse of a burst packet onwards, the individual pulse energy of a laser pulse is lower than the individual pulse energy of the previous individual pulse.

[0026] This, of course, applies from the second individual pulse onwards. One can also formulate it equivalently by saying that the individual pulse energy of a laser pulse in the burst packet is greater than the individual pulse energy of the subsequent individual pulse.

[0027] In other words, this means that the individual pulse energy in the burst packet decreases over time. Although significantly less energy is deposited in the workpiece and / or filament by the burst packet of individual pulses, it is still possible to efficiently produce a hollow-channel filament. This is particularly surprising given that the goal is to apply as much energy as possible.

[0028] It has been shown that it is advantageous if the single pulse energy decreases linearly within the burst packet or particularly advantageous if the single pulse energy decreases exponentially during the burst duration.

[0029] It is particularly advantageous if the single pulse energy during the exponential decrease is determined according to the formula

[0030] E(t) = E decreases. E(t) symbolizes the intensity of an individual pulse at time t of a burst packet, Eo the individual pulse energy of the first individual pulse of the burst packet, and AF the decrease factor of the entire burst packet, i.e., the decrease in the individual pulse energy of the last pulse of the burst packet compared to the first pulse. In the aforementioned advantageous embodiment, in which the last individual pulse has, for example, 60% of the individual pulse energy, the decrease is 40%.

[0031] An embodiment of the described method according to the invention provides that a filament-shaped initial channel is generated during the first individual pulse of a burst packet, in particular an initial hollow channel, which is widened by the following individual pulses of the burst packet.

[0032] It is assumed that this describes the mechanism of action of the filamentation described herein. The first single pulse alone creates a filament-shaped initial channel. This can, in particular, already be a hollow channel. The subsequent single laser pulses in the burst packet then further expand this initial channel, which essentially means increasing its diameter. The result is the hollow-channel-shaped filament in the workpiece.

[0033] Even though we are referring to the diameter of a hollow channel here, it should be noted that in reality, a laser-generated filament in a workpiece rarely corresponds to a uniform channel. This approach is an idealized, simplified description. Local variations in diameter are certainly possible along the filament axis, i.e., the axis in the longitudinal direction of the filament. It is also possible, and encompassed by the invention, for the filament to have a different diameter at the beginning and / or end than in between.

[0034] A further embodiment of the method according to the invention includes producing a region of compressed workpiece material around the hollow channel-shaped filament, also called a compression zone or compression region, which has a higher density, in particular compared to the starting material.

[0035] The compaction of the surrounding material appears to be a result of the widening of the initial channel caused by the laser pulse. Simply put, a cylindrical region of compacted workpiece material is located around the hollow-channel-shaped filament, with the filament axis and the cylinder axis essentially overlapping. This region of compressed material has the advantage that the hollow-channel-shaped filament, especially the filament wall, can be mechanically stabilized by this region. The filament wall is thus, in a sense, hardened.

[0036] It is astonishing that, firstly, the first single pulse of the burst packet is sufficient to generate the initial channel without any energy pooling taking place, and that, furthermore, the decrease of the single pulse energies in the burst packet still leads to the compression of the surrounding workpiece material.

[0037] A particularly advantageous method provides that the individual pulses of a burst packet press material of the workpiece into the workpiece during the expansion of the initial channel, in particular perpendicular to the channel axis.

[0038] The described compression also means that workpiece material is not removed from the workpiece by the interaction with the laser pulses, but is pressed into the filament wall and thus into the workpiece. One advantage of the described process is that, as a rule, significantly less material is removed by inserting the hollow-channel-shaped filaments than is pressed into it. This is because removed material, which is ablated, is released as dust in the production facilities. This dust can not only contaminate the sensitive optics used for steel forming, but, if it is suspended in the beam path and / or even in the initial channel, absorbs the incoming laser radiation, especially since the ablated material is at least essentially the same as the material being processed by the interaction with the laser radiation. Such parasitic absorptions negatively impact the efficiency of the process.

[0039] In particular, it is assumed that the described reduction of the single pulse energy in the burst packet contributes to reducing the proportion of ablated material in favor of compressed material.

[0040] The finding that the aforementioned compression occurs with the process described here was based on measurement tests according to which a workpiece to be filamented was first weighed. The workpiece was then filamented using the described process, i.e. the hollow channel-shaped filaments were introduced into the workpiece as described. The diameter and length of the hollow channels were measured and their volume determined from this. If this measured volume of the hollow channels is correlated with the specific gravity of the workpiece material, it was found that significantly less weight and therefore material was removed from the workpiece by filamentation than would be expected from the inserted hollow channel volume. The material must therefore still be present in the workpiece, plausibly as compressed material as previously described.

[0041] In the described process, the relative weight loss per filament is less than 10%.

[0042] The relative weight loss per hollow channel-shaped filament was determined to be less than 10% as follows (here as an example for a 2 mm thick soda-lime glass, which was cut to a size of 150 x 250 mm by manual scoring and breaking with a commercially available glass cutter). 2Particles, fingerprints, and other contaminants were removed by cleaning both sides with glass cleaner, ethanol, and compressed air, either manually or in an automated cleaning process (washing machine). The prepared sample was then weighed on a precision balance (Mettler Toledo AB204-S). The entire setup was enclosed in a transparent plastic box to increase measurement accuracy. The sample was then mounted on the holder of an XY-axis system (Aerotech Inc., 3D Micromac microSTRUCT). It was aligned using two stops and secured against slipping. Laser processing was performed as previously described using an ultrashort pulse laser with a pulse width of 10 ps, ​​a wavelength of 1064 nm, and an energy of 400 pJ in a burst packet, specifically with 3 to 7 individual pulses. A two-dimensional grid of hollow-channel-shaped filaments measuring 90 x 160 mm was created. 2The sample was then weighed again. To rule out any falsification of the measurement results that could be caused by particle deposition during the laser process, the previously mentioned cleaning steps (glass cleaner, ethanol, compressed air) or automated cleaning procedures were repeated, and the sample was weighed again. The following calculation procedure was used to calculate the relative weight loss per filament.

[0043] The theoretical weight loss mtheo per hollow channel-shaped filament, if the complete material of the hollow channel were removed from the sample, is calculated as:

[0044] Where d: Diameter of the hollow cylindrical filament, determined from SEM images of fracture edges along the filamentation h: Glass thickness p: Density of the sample, from material data sheet The actual weight loss m actualper pre-damage is obtained if the measured weight loss m mess divided by the number of previous damages N:

[0045] The relative mass loss is determined by forming the quotient:

[0046] Here, as an example, for a 2 mm thick glass: With a sample thickness of h = 2 mm and a density of p = 2.5 g / cm 3 and a determined filament diameter of d = 600 ± 60 nm, this results in a theoretical mass loss of rritheo = 1413 ± 60 pg per filament.

[0047] The actual (average) weight loss is obtained at mmess = 6 ± 1 mg and N = 5.8 ■ 10 7 to mactuai = 104 ± 17 pg per filament. The relative weight loss is calculated as 7.4 ± 2.7%. All calculations take error propagation into account.

[0048] An advantageous method therefore provides that the relative weight loss m reiper hollow channel shaped filament is less than 10%, measured by

[0049] IDrel = actual / mtheo and / or particularly advantageously, the workpiece has a material densification of at least 1% compared to the actual material density, measured at a radius of 3 pm around the longitudinal axis of the respective hollow channel-shaped filament.

[0050] The following measurement method was used to determine that the glass or glass-ceramic exhibits a material densification of at least 1% compared to the actual material density within a radius of 3 pm around the longitudinal axis of the respective cylindrically symmetric pre-damage: Lena Bressel, Dominique de Ligny, Camille Sonneville, et al. “Femtosecond laser induced density changes in GeO2 and SiO2 glasses: Fictive temperature effect [Invited]”, Optical Materials Express, Vol. 1 , No.

[0051] 4, 605 613 (1 August 2011 ) DOI: 10.1364 / OME.1 .000605. The evidence of material densification was indirectly demonstrated via a spectral shift of the Nd peak in the Raman spectrum. For this purpose, the Raman microscope had to have a spectral range of 10000 cm -1so that the Nd luminescence can be measured at approximately 890 nm upon excitation at 488 nm, and be equipped with suitable motorized XYZ stepper motor axes and a suitable microscope objective (NA > 0.7) to achieve a spatial resolution of better than one micrometer. The components according to the invention were scanned in the area to be measured using the stepper motors. At each point, a Raman spectrum of the Nd peak was recorded. Its spectral position is recorded, and pressures and densities were calculated using standard spectra. These standard spectra were recorded from glass or glass-ceramic bodies that were compacted using belt presses. Further verification was possible using spatially resolved Brillouin spectra. The radius of 3 pm around the longitudinal axis or filament axis represents the measurement location of the compression or compaction and is also referred to below as the compression measurement line.The densification zone can extend further around the filament into the material of the workpiece.

[0052] The diameter of the hollow channel-shaped filament in the sense of this description refers to the distance from one channel wall to the other perpendicular to the filament axis. In a cylindrical channel, the diameter is essentially the same everywhere. However, it is possible for the hollow channel-shaped filament to taper towards one of the surfaces of the workpiece or, depending on the point of view, to widen. In this case, one speaks of V-shaped filaments. It is also possible for the hollow channel-shaped filament to have a local constriction, thus creating an hourglass or X-shape, so to speak. The shape of the filament can be influenced by the formation of the focus line and its position. An advantageous method in the sense of the invention provides for the hollow channel-shaped filament to have a diameter of 100 nm to 3 pm. Ranges from 250 nm to 1.5 pm or from 500 nm to 1 pm are particularly advantageous.

[0053] If the filament is not formed as a hollow cylinder, or as a hollow cylinder with significant local deviations, but rather in a V-shape or X-shape, then the aforementioned diameter refers to the diameter at the location with the maximum value, or, in other words, at the widest point. In this description, the diameter of the filament is measured at the surface of the workpiece, specifically at the surface that is and / or was facing the laser. This side can also be clearly determined on the workpiece being processed.

[0054] As already described, the invention balances the need for efficient production by producing the largest possible filaments, especially filaments with the largest diameters, which can, however, adversely weaken the workpiece due to unwanted cracking around the filament, with the control of energy input and thus crack formation. As described, the workpiece material is particularly advantageously compacted around the filament axis.

[0055] This is particularly advantageously done according to a process in which the focused laser beam generates a plasma in the workpiece, which forms the hollow channel-shaped filament.

[0056] It is assumed that the burst of ultrashort laser pulses ignites a plasma in the workpiece, particularly within the elongated laser focus line. As described, the plasma explosion can push material into the workpiece perpendicular to the filament axis, thus creating the hollow channel and / or the hollow channel-shaped filament with the previously described geometries. It has proven particularly advantageous if the process is designed such that the first individual pulse of the burst packet generates an initial plasma, which is energized and thus maintained by the subsequent individual pulses of the burst packet.

[0057] The individual pulses following the first individual pulse of the burst packet essentially feed the plasma and keep it burning. However, because the pulses following the first individual pulse of the burst packet have lower individual pulse energies, the plasma explosion appears to be controllable.

[0058] According to an advantageous embodiment of the method according to the invention, microcracks are generated in the workpiece by the plasma, which run from the hollow channel-shaped filament into the volume of the workpiece.

[0059] Such crack formation is particularly important when applying the method for separating the workpiece. In principle, crack formation occurs perpendicular to the filament axis and runs from the filament into the volume of the workpiece. The cracks are usually evenly distributed around the filament axis. This means that, when viewed from above, the filament and / or the filament axis exhibit essentially the same crack density within a 360° area around the filament. It has been shown that the method according to the invention reduces the number of cracks and thus the crack density as well as the average crack length compared to a method in which the individual pulse energy within the burst packet is not reduced.

[0060] A particularly advantageous embodiment of the method provides that at least 90% of the microcracks end within the aforementioned compression zone, particularly advantageously within a radius of 3 pm around the longitudinal axis of the hollow channel-shaped filament.

[0061] This is a statistical analysis. The number of microcracks and their length are observed, and it is evaluated how many of the microcracks end within the compression range and how many extend beyond it. It is advantageous if at least 10% of the microcracks end within the compression range. As described, the degree of compression is measured within a radius of 3 pm around the longitudinal axis of the hollow-channel filament. This means that, particularly advantageously, 90% of the microcracks end within this 3 pm radius. The compression range thus advantageously limits the vast majority of microcracks.

[0062] In order to provide a particularly efficient separation process, a further advantageous embodiment of the process provides that the elongated focus line is widened in a direction perpendicular to the axis of the filament and the length and / or number of microcracks is greater in the direction of the widening of the focus line than in the direction perpendicular to the widening.

[0063] At this point, it should be mentioned again that the limitation of the crack length to the compression range can surprisingly be advantageously achieved by making the single pulse energy Epo of the laser pulse at the beginning of the burst duration Tb greater than the single pulse energy Epx of the laser pulse at the end of the burst duration.

[0064] The direction of the focal line expansion perpendicular to the filament axis is also referred to as the preferred direction. The expansion of the focal line in one direction also causes the plasma to expand in that direction. The plasma explosion is presumably stronger in the direction of expansion. In any case, the number of microcracks in the preferred direction is greater than perpendicular to the preferred direction and / or the average length of the microcracks is greater in the preferred direction than perpendicular to the preferred direction.

[0065] In particular, the preferred direction can be selected to lie in the direction of and / or at least substantially congruent with a provided dividing line of the workpiece. This means that the widening of the laser focus relative to the workpiece is controlled accordingly. This can be achieved by movable optical devices and / or movable workpiece holders, which are controlled in particular by microcontrollers and / or data processing devices. This and the meaning of the invention also encompasses the possibility of the preferred direction being individually adjustable at any desired point on the workpiece, i.e., the preferred direction can be variable and, in particular, adjustable depending on the position of the focus line on the workpiece.

[0066] The shape of the focal line widening can be chosen arbitrarily within the meaning of the invention. A laser focus in the shape of an oval is particularly common. The main axis of the oval then corresponds to the preferred direction. Linear, asymmetrical, and any other suitable shape of widening are also conceivable and encompassed by the invention.

[0067] The described method can be designed such that the plasma emits X-ray radiation which is attenuated and / or shielded by at least one shielding element.

[0068] The inventors have recognized that the previously described plasma explosion presumably causes the emission of X-rays. In particular, this can be in the range of 4 keV to 20 keV and is thus surprisingly intense. This can be used, for example, to specifically examine materials and / or workpieces and / or the workpiece to be processed, especially simultaneously with its laser processing. It has also been shown that the reduction of the individual pulse energies in the burst packet according to the invention also contributes to a reduction in the emitted X-rays.

[0069] Nevertheless, it is advantageous to contain the resulting X-ray radiation within a defined spatial area. In particular, a production environment is to be protected from the emitted X-ray radiation. Therefore, at least one shielding element is provided, which at least attenuates and / or shields the X-ray radiation emitted by the plasma.

[0070] The shielding element can, in particular, be designed to at least partially enclose the space around the workpiece. In particular, an inlet for the laser beam, which must impinge on the workpiece, and / or an inlet and / or outlet through which the workpiece can be fed for processing and removed again after processing are provided. In the simplest case, the shielding element forms a chamber around the workpiece, in particular with the aforementioned passages. The shielding element can, in particular, be formed as a single piece or in multiple pieces.

[0071] An advantageous method includes that the shielding element is attached such that it at least partially surrounds the workpiece, and wherein the shielding element consists of or comprises a material which absorbs X-radiation, preferably the material is selected from the group of metals, in particular comprising iron, steel, stainless steel, tungsten, or selected from the group of plastic composites, in particular comprising metal-filled plastic composite, in particular tungsten-filled plastic composite.

[0072] In general, the shielding element aims for the highest possible shielding efficiency. However, it should also be efficient to manufacture and operate. The shielding behavior essentially correlates with the atomic number of the atoms in the material comprising the shielding element. The interaction with the thickness of the shielding element then determines the overall ability of the shielding element to at least reduce and / or shield X-rays.

[0073] Metals fulfill this compromise particularly well. For example, lead is known as a good X-ray absorber. However, it is not the most advantageous material in operation. For the purposes of the invention, ferrous materials are advantageously used, in particular steel and / or stainless steel, particularly according to EN 10020:2000, issue date 07 / 2000, including unalloyed stainless steels and alloyed stainless steels. Tungsten has also proven advantageous. Metal-filled plastic composites, in particular tungsten-filled plastic composites, are also possible.

[0074] As previously described, it is advantageous to direct the laser beam onto the workpiece in the form of an elongated focal line. Accordingly, an advantageous method comprises directing the focused laser beam onto the workpiece in an elongated focal line. Particularly advantageously, the elongated focal line is generated by beam-shaping optics comprising optical elements selected from the group consisting of SLM (Spacial Light Modulator), axicon, truncated cone axicon, double axicon, spherically aberrated lens, and / or chromatically aberrated lens.

[0075] To achieve spatial confinement of the laser radiation across the focal line, so-called Bessel beams are ideally used. These consist of a central, bright maximum surrounded by fainter rings. Unlike the Gaussian beams emitted by the laser, their radius remains constant in the direction of propagation. They enable both processing over a larger depth range and greater tolerance in workpiece alignment. Furthermore, they are virtually diffraction-free.

[0076] The generation of Bessel beams and a long, extended line focus is achieved, among other things, with so-called axicons as beam-shaping lenses. The axicon as an optical element was described by McLeod in 1954 (John H. McLeod: The Axicon: A New Type of Optical Element. J. Opt. Soc. Am. / Vol. 44, No. 8, August 1954). According to this, an axicon is an optical element that images the light from small point sources onto a straight, continuous focal line. McLeod describes various forms of axicons, but emphasizes the glass cone as the most important axicon.

[0077] The truncated cone axicon has a flattened surface at the apex of the axicon cone, in the laser beam axis. The center beam is not refracted by the truncated cone axicon. This facilitates the adjustability of the optical system, as it is usually very difficult to align the center beam precisely with the apex of the axicon. Likewise, tilting of the axicon and / or its manufacturing precision, i.e., the sharpness and symmetry, especially of the cone apex, are less critical. The double axicon has an entrance surface designed such that the laser beam is formed into a ring beam in the double axicon.The advantage of the double axicon is that by selecting the energy of the laser, the refractive index of the double axicon and the axicon angle of the entrance surface, the intensity of the laser beam within the double axicon can be set lower than the threshold intensity of the material of the double axicon, so that the material of the double axicon is not destroyed during the beam passage.

[0078] It may also be advantageous to create the elongated laser focus and / or the line focus using Gaussian optics.

[0079] Beam extension can then be achieved through the spherical aberration of a Gaussian lens. Spherical lenses and / or cylindrical lenses, each with aspherical components, are particularly suitable for this purpose.

[0080] When exploiting spherical aberration, it can be provided, in particular, that the laser beam is not aligned with the geometric center of the lens and / or that the lens is tilted. This can also be implemented dynamically, i.e., that the lens alignment relative to the laser beam can be controlled by actuators and microcontrollers and / or data processing devices.

[0081] Instead of a monochromatic laser, a pulsed, polychromatic laser beam with a specific pulse duration and specific wavelengths of the laser beam can also be used. Using an optical arrangement with chromatic aberration for wavelength-dependent focusing of the laser beam and advantageously with at least one filter for wavelength-dependent filtering of the laser beam, this can generate a focal line along the beam direction of the laser beam, allowing the machining depth of the workpiece to be selectively and precisely adjusted. In particular, the length of the focal line can be adjusted by creating different focuses. In some cases, the presence of an elongated laser focus is not essential. In these cases, conventional Gaussian optics are usually sufficient. These cases are, of course, also encompassed by the invention.

[0082] The method is advantageously designed such that the filament is a channel that connects one surface of the workpiece to another surface, thus completely penetrating the workpiece as a through-channel, or is a channel that ends in the volume of the workpiece as a blind hole.

[0083] As described, the shape of the hollow channel can be cylindrical, V-shaped, and / or X-shaped. This can be adjusted, in particular, by adjusting the focus position of the line focus relative to the workpiece in the beam direction and the intensity profile of the focus line perpendicular to the beam direction.

[0084] In an advantageous method, the laser wavelength is from 500 nm to 1100 nm, advantageously 1020 nm to 1080 nm, and / or the pulse duration of a single pulse is from 200 fs to 11 ps and / or the burst duration is from 1 ps to 500 ns and / or the single pulse energy of the first single pulse at the beginning of the burst duration is from 100 pJ to 1000 pJ, advantageously from 200 pJ to 500 pJ, and / or the workpiece consists of or comprises glass and / or glass ceramics and / or silicon and / or ceramic.

[0085] The workpiece comprises the material mentioned, in particular glass and / or glass ceramics and / or silicon and / or ceramics, in particular in the area in which the interaction with the laser beam occurs, i.e. at the location of the hollow channel-shaped filament.

[0086] The laser wavelength is determined by the laser used. Its wavelength bandwidth is typically very limited. The above wavelength ranges indicate that lasers can be used in the specified wavelength ranges. As described, the laser wavelength is selected so that the workpiece is essentially transparent within this wavelength range. In the previous paragraphs, it was explained that the microcrack density and / or microcrack length around the hollow channel-shaped element can be controlled by reducing the individual pulse energy in the burst packet.

[0087] To enable efficient separation of the workpiece through filamentation, the spacing between two adjacent filaments is adjusted accordingly. The spacing between two adjacent filaments is advantageously between 3 pm and 20 pm, particularly between 3 pm and 10 pm or between 4 pm and 8 pm. For glasses with a thickness of over 0.5 mm, the empirical formula for the spacing between the filaments is particularly advantageous:

[0088] Distance in mm = thickness of the workpiece ■ Z where the factor Z can assume values ​​from 3 to 7. This empirical formula for selecting the distance has proven particularly applicable for workpieces made of glass and / or glass ceramic with a thickness or material strength in the range of 0.5 mm to 30 mm. For example, applying this formula for a workpiece with a thickness of 1 mm results in an advantageous distance between the filaments of 3 pm to 7 pm.

[0089] This distance is significantly smaller compared to processes with conventional burst packets, in which the individual pulse energy remains the same and is chosen to be as large as possible.

[0090] A burst degression of 90% to 70% is particularly advantageous, especially with burst packets with 2 to 7 individual pulses, in particular 3 to 5 individual pulses.

[0091] The above parameters apply in particular to glasses containing silicon and boron.

[0092] The combination of the aforementioned parameters enables a particularly efficient method for machining workpieces. In addition to the method described above, the invention also encompasses a device described below. In particular, the device is suitable for carrying out and / or applying the described method. All aspects described with regard to the method also apply to the device. Likewise, all aspects described below with regard to the device also apply to the method.

[0093] The invention accordingly also relates to a device for the laser-assisted introduction of a hollow channel-shaped filament into a brittle-breaking workpiece, in particular a device for applying the method described above, with an ultrashort pulse laser with a wavelength A, in which the workpiece is transparent to the irradiated wavelength, and a beam-shaping optics which is designed such that the laser beam is directed onto the workpiece with an elongated focus line and the focus line lies at least partially within the workpiece, wherein the laser is set up such that it emits at least one burst packet of laser pulses, wherein the at least one burst packet has a burst duration Tb and comprises a sequence of ultrashort laser pulses,which have a single pulse energy Ep and an intra-burst delay Ti and wherein the single pulse energy of the laser pulse at the beginning of the burst duration Tb is greater than the single pulse energy of the laser pulse at the end of the burst duration.,

[0094] The device can be integrated into a production environment and enables efficient machining of workpieces.

[0095] The aforementioned device is advantageously provided with a device by means of which the reduced energy of the last individual pulse at the end of the burst duration can be adjusted compared to the energy of the first individual pulse at the beginning of the burst duration. The energy drop of the individual pulses within the burst duration can be adjusted particularly advantageously, in particular as a linear or exponential decay behavior. These decay behaviors naturally apply analogously to the method described above. An advantageous embodiment of the device comprises a device by means of which the laser and / or the optical device can be moved relative to the workpiece. In particular, the device comprises a movable workpiece holder and / or a device for moving the laser and / or the optical device, and a control unit for controlling the relative movement.

[0096] In particular, the control device can be configured such that filaments can be introduced into the workpiece in a predefined pattern, in particular adjacent along a predefined dividing line.

[0097] By moving the workpiece relative to the laser, particularly the focal line, and repeatedly impinging a burst packet at different locations on the workpiece, this can be structured and / or a parting line can be created in the workpiece, at least as a preliminary parting line along which the workpiece can be separated. In the case of the preliminary parting line, the workpiece is still continuous and is later separated, usually by applying a mechanical breaking force or thermal stress. This can have advantages in the production of serial workpieces, particularly during their transport and / or subsequent processing.

[0098] In order to generate a previously described preferred direction, the beam-shaping optics of the device contains optical elements which widen the focus line in a direction perpendicular to the filament axis, the so-called preferred direction, preferably in the direction of an adjacent filament on a predefined dividing line.

[0099] The device can comprise a control unit with which the position of the preferred direction can be adjusted, in particular by positioning the optical elements. This positioning can be achieved, as previously described in connection with the method, in particular with the aid of microcontrollers and / or data processing systems. As also already described, the device advantageously comprises a shielding element for X-ray radiation, which at least partially encloses the workpiece.

[0100] Particularly advantageously, the shielding element is formed as a single piece. However, it is of course also possible to have more than one shielding element, even at different locations on the device.

[0101] The materials advantageously used for the shielding element have already been mentioned. An advantageous device is one in which the shielding element consists of or comprises a material selected from the group of metals, in particular iron, steel, stainless steel, tungsten, or selected from the group of plastic composites, in particular tungsten-filled plastic composites.

[0102] The applications of the method and / or device are also subject of the invention. The method and / or device is advantageously used for separating or structuring or preparing for etching glass and / or glass-ceramic and / or silicon elements.

[0103] When structuring workpieces, hollow-channel filaments are introduced into the workpiece in a predefined pattern. The diameter of the hollow-channel filaments can be expanded through an etching process. It has been shown that etching allows material to be removed more quickly in the area of ​​the channel wall than on the surface of the workpiece.

[0104] It is assumed that the microcracks around the filament represent a stronger point of attack for the etching medium. Possible etching media include HF, HCl, NaOH, and / or KOH.

[0105] The invention is explained in more detail below with reference to the figures. The figures are schematic and not necessarily to scale. The figures also represent exemplary embodiments. The same reference numerals in different figures have the same meaning. They show: Fig. 1: A schematically illustrated device

[0106] Fig. 2: A schematically illustrated device for processing a workpiece strip

[0107] Fig. 3: Burst packets of laser pulses

[0108] Fig. 4a to 4d: Examples of energy profiles of the laser pulses within a burst packet

[0109] Fig. 5a to 5c: Examples of filament shapes in the workpiece

[0110] Fig. 6: Top view of a filamented workpiece

[0111] Fig. 7: Filaments with microcracks in the workpiece according to the state of the art in plan view

[0112] Fig. 8: Filaments with microcracks according to the description in top view Fig. 9: Filaments with shortened microcracks in top view

[0113] Fig. 10: Filaments with microcracks in preferred direction in plan view

[0114] Fig. 11 : Dependence of the microcrack length on the preferred direction

[0115] Fig. 12: Examples of focus areas with preferred direction in plan view

[0116] Fig. 1 schematically shows a device according to the present description. In particular, the device (100) is suitable for applying the described method. With the device (100), at least one hollow channel-shaped filament (1) is inserted into the workpiece (2). The device comprises an ultrashort pulse laser (3) with beam-shaping optics (30). This focuses the laser beam (4) with the wavelength A as an elongated focal line into the workpiece (2). At least part of this focal line lies within the workpiece, where the filament (1) is produced by interaction with the ultrashort pulses of the burst packet.

[0117] The workpiece (2) is arranged on a workpiece holder. The positioning device (200) can be used to laterally position the point of incidence of the laser beam (4) of the ultrashort pulse laser (3) on the surface of the workpiece (2). As described, the workpiece (2) is a brittle workpiece, in particular comprising glass, glass ceramic, and / or silicon. The wavelength A of the laser (3) is selected, as already described, such that the workpiece (2) is transparent to this wavelength.

[0118] As described, the beam-shaping optics (30) can also be designed such that the focal line is widened perpendicular to the filament axis to provide a preferred direction (DR). It is also possible to incorporate apertures into the beam path to create a correspondingly shaped focus.

[0119] In the illustrated example, the positioning device (200) comprises an xy table on which the workpiece (2) rests on its underside. Alternatively or additionally, it is also possible to design the laser (3) and / or the beam-shaping optics (30) to be movable in order to move the laser beam (4) so ​​that the point of impact of the laser beam 4 can be moved even when the glass element (1) is held in place. A combination of the two methods is also possible.

[0120] The ultrashort pulse laser (3), the positioning device (200) and / or the beam-shaping optics (30) are connected to a control device (210), for example a microcontroller and / or a data processing device.

[0121] As described above, X-rays can be released by the plasma generated by the burst of ultrashort laser pulses, which subsequently produces the filament. The shielding element (300), also shown, serves to protect the surroundings from these radiations. For this purpose, it comprises a material of a suitable thickness that at least partially absorbs the X-rays. In the case shown, the shielding element (300) is made of stainless steel sheet, which encloses at least the area of ​​the laser focus. The sensor device (310) is provided as an optional element, which can, for example, detect light reflections of the incoming laser beam or, in particular, the emitted X-rays. Other measured variables are also conceivable, particularly if they correlate with the energy of the radiated laser power.In the present case, the energy of the emitted laser power can be used as a control variable for the laser (3) and / or the positioning device (200). For this purpose, it is connected to the control device (210). In this way, high process stability can be achieved, for example, by using the X-ray emission as a measure of the incoming laser power and controlling the laser (3) in such a way that the desired processing parameters are maintained using these measured variables.

[0122] Fig. 2 shows a device (100) in which the laser filamentation described herein is carried out on a workpiece (2) as strip material. In the present example, this is in particular ultra-thin glass that is obtained from a drawing tank (250) by means of a downdraw process. A corresponding application of the method according to the invention is also possible with alternative ultra-thin glass production processes such as the overflow fusion process. The ultra-thin glass here has a material thickness of approximately 100 pm at most, in particular the material thickness in this example is from 20 pm to 50 pm. By means of a deflection device (251), the strip material is deflected in the direction of the laser processing device with the laser (3). Due to the distance to the drawing tank, filamentation takes place here in the cold area of ​​the system. However, it is equally possible for the laser processing to take place in the hot area (220) closer towards the drawing tank (250).The strip material can be picked up and / or stored, for example, on a drum in the form of a roll, here a glass roll (221). For this purpose, the drum is rotated according to the withdrawal speed of the glass strip. Also shown is a shielding element (300), which at least partially surrounds the filamentation area and, as described, shields the surrounding area from the emitted X-rays. In the present case, a shielding element (300) is also provided below the workpiece.

[0123] Fig. 3 shows three burst packets of laser pulses as a plot of the intensity I of the laser pulse over time t. Since, as described, the energy of an individual laser pulse corresponds to the integral of the intensity over time, the energy E of a laser pulse correlates with its intensity. Accordingly, the representation also corresponds to a schematic plot of the laser pulse energy over time. Three burst packets of four laser pulses each are shown, with the burst duration being Tb. Tb is determined from the time of the highest intensity of the first laser pulse (41 ) of the burst packet to the time of the highest intensity of the last pulse of a burst packet.

[0124] In the example given, the number of pulses in a burst packet is four. Advantageous, for example, are from two to 20 pulses and / or from 2 to 10 pulses in the burst packet. The pulse width is advantageously from 300 fs to 11 ps. Tb is advantageously in the range from 1 ps to 500 ns and Ti from 400 fs to 400 ns. To produce a hollow channel-shaped filament (1), the irradiation of a single burst packet onto the workpiece may be sufficient. However, the use of two to 10 burst packets is also possible and advantageous, in particular from 2 to 5 burst packets. The time interval between the individual burst packets is referred to as the inter-burst delay Td. 1 / Td can advantageously be from 1 to 1000 kHz.

[0125] These aforementioned parameters are selected depending on the material of the workpiece and the desired geometries of the filament and distances between filaments.

[0126] The energy Epo of the first laser pulse (41) of a burst pact is, as described in the sense of the invention, greater than the energy Ep x of the last laser pulse (42) of the same burst packet.

[0127] 4a to 4d show exemplary advantageous curves of the energies of the individual laser pulses over time t in a single burst packet. For the sake of simplicity, the curves are given as curves and not as discrete value points. The number of pulses in the burst packet is arbitrary, advantageously within the ranges mentioned above. The simplest energy curve is a linear curve as shown in Fig. 4a. In this case, the individual pulse energies within the burst packet decrease by at least essentially the same amount. It is also possible for the first individual pulse(s) of the burst packet to have at least essentially the same energy, and for the following pulses to have a lower, in particular equally constant, energy. This corresponds to a two-stage energy curve, as shown in Fig. 4b. Non-linear curves are also possible, for example an exponential decrease in the individual pulse energies in the burst packet as shown in Fig. 4c.It can also be provided that the individual pulse energy in the burst packet initially increases and then decreases again, with the first individual pulse, as described, having a higher energy than the last. The laser pulse(s) following the first individual pulse then have a higher energy, which then decreases towards the last individual pulse. Such a curve is shown in Fig. 4d.

[0128] 5a to 5c show cross-sections through filamented workpieces in order to schematically illustrate possible shapes of hollow channel-shaped filaments. Fig. 5a shows the simplest case when cylindrical hollow channels are introduced into the workpiece as filaments (1). The filament axis A passes through the center of the filament. The filament walls are axially symmetrical. In the example shown, the filaments (1) connect the top side (0) of the workpiece (2) with its underside (U). The hollow channel-shaped filament (1) thus represents a through-channel or a through-opening. As described above, it is also possible for the filament to end in the workpiece. In this case, not shown, it represents a blind hole. The workpiece has the workpiece thickness (s). The invention is applicable to almost any workpiece thickness.Accordingly, (s) covers the range from ultra-thin glass with a thickness of a few micrometers to standard glass with thicknesses of several centimeters. Particularly applicable are the range from 10 pm to 200 pm, i.e., the ultra-thin glass range, and the range from 0.5 mm to 3 mm, i.e., the substrate or flat glass range.

[0129] Fig. 5a also shows the compression region (22) that surrounds the filament. The compression region, or compression zone, and its formation have been described in detail above. It extends essentially perpendicular to the filament axis into the material of the workpiece (21). According to Fig. 5b, a truncated cone-shaped filament (1) can also be provided, which is synonymously referred to as a V-shaped filament. This can also be described as having a hollow channel. The truncated cone-shaped filament is, in particular, also axisymmetric to the filament axis (A). In the present illustration, the filament opens out towards the underside of the workpiece. This means that the diameter (dU) of the filament (1) on the underside of the workpiece (2) is greater than the diameter (dO) on the top. Not shown, but again possible, is a design as a blind hole.The expansion of the filament toward one side of the workpiece can be achieved by adjusting the focus position of the beam-shaping optics, particularly by adjusting the convergence of the laser beam and the distance from the workpiece. Typical channel angles a range from 0.1° to approximately 30°, in particular from 0.1° to 10° or from 0.2° to 3°. This channel angle is measured between the perpendicular line through the workpiece and the inner wall of the filament and thus corresponds to half the aperture angle.

[0130] Fig. 5c shows hourglass-shaped or synonymously x-shaped filaments (1), again designed as continuous hollow channels. This embodiment can be produced in particular by placing the region of the narrowest focusing of the laser in the volume of the workpiece. The same ranges as mentioned for the v-shaped filaments are advantageous for the channel and / or opening angle α. The x-shaped filaments are characterized by a constriction of the filament diameter in a region along the filament axis (A). In the example shown, the constriction is in the middle of the workpiece (2), i.e. at the same distance from its surfaces. There are also embodiments in which the constriction is arranged closer to one surface of the workpiece than to the opposite one.

[0131] The compression region (22) is not shown in Figs. 5b and 5c for the sake of clarity. However, it can of course also be present in the case of V- and / or X-shaped filaments (1). In this case, it follows the respective channel wall in particular. Fig. 6 shows a top view of a filamented workpiece (2). The hollow channel-shaped filaments (1) are arranged in a predetermined pattern in the workpiece (2). This is also referred to as a structured workpiece (2). The filaments (1) are surrounded by the compression region (21), in which, as described above, the material of the workpiece is compacted by the effect of irradiation with the ultrashort laser pulses. As also described above, a compression of at least 1% exists within a radius of 3 pm, the so-called compression measurement line (23), around the filament axis. This means that the degree of compaction, synonymously referred to as compression, is determined at this compression measurement line (23).

[0132] Fig. 7 schematically shows the mechanism assumed according to the prior art for separating workpieces using laser filamentation. It shows a top view of a workpiece. The filaments (1) are surrounded by microcracks (50) of length RL. According to the common explanation, the microcracks are caused by the shock wave generated by the plasma explosion in the focus area of ​​the ultrashort laser pulse in the material of the workpiece (21). If microcracks (50) combine to form a connected microcracks (51) in a neighboring filament, these connected microcracks (51) can form a continuous crack line along which the workpiece can be separated.

[0133] Fig. 8, on the other hand, shows a schematic top view of a workpiece that has been filamented according to the present invention. Because the energy of the last individual pulse of a burst packet is lower than the energy of the first individual pulse of a burst packet used to produce a filament (1), the crack length RL is shorter than according to the prior art. It is assumed that reducing the laser pulse energies in the burst packet at least reduces uncontrolled energy input into the material of the workpiece (21). This is surprising because it is usually assumed that it is more advantageous to deposit as much laser energy as possible in the material of the workpiece during filamentation. Viewed in top view, the microcracks (50) generally continue radially from the filament (1) as the center into the material of the workpiece.Here too, the microcracks of adjacent filaments can combine to form one or a plurality of connected microcracks (51) and thus form a dividing line.

[0134] The shortened crack length RL does have the disadvantage that the filaments (1) have to be arranged closer to one another in order to form connected filaments (51) and thus a parting line. However, this is balanced by the advantage that shorter and / or fewer microcracks continue away from the parting line into the material of the workpiece (21). These would otherwise weaken the edge of the workpiece resulting from the parting, which manifests itself in reduced edge strength. However, high edge strength is advantageous because it allows for more efficient further processing and / or application of the filamented workpiece and, in particular, results in less waste.

[0135] In the present example of Fig. 8, the compression region (22) around the filament (21) is also shown. In this exemplary embodiment, the crack length RL is longer than the thickness of the compression region. This means that the majority of the microcracks continue beyond the compression region, in particular beyond the compression measurement line. At this point, it should be noted that the explanations are statistical. The focus is on the properties of the majority of the microcracks. The crack length RL is therefore to be understood as the average crack length, which is determined by the arithmetic mean of all measured crack lengths. This of course also means that in an actual exemplary embodiment, microcracks can occur that are significantly longer or shorter than described using the present diagram. However, these are in the minority, in particular the clear minority. Fig.Figure 9 explains an advantageous embodiment in a representation analogous to Figures 7 to 8. According to this embodiment, the microcracks (50) run within the compression region (22). The crack length RL is thus less than the width of the compression ring around the filaments (1). In order to form a continuous parting line with connected microcracks (51), the compression regions (22) must at least be adjacent to one another and / or overlap. This applies in particular to the compression measuring line. If the filaments (1) are too far apart from one another, no microcracks (52) form, and thus a parting line is interrupted at this point. This can be particularly advantageous for setting the breaking force required for the actual separation of the parts of the workpiece along the parting line.In many cases, it is advantageous if the filamented workpiece does not separate or detach on its own, but requires an additional mechanical force, the separation force. This can facilitate further processing, especially the transport, of filamented workpieces.

[0136] In the present example of Fig. 9, it is assumed that the microcracks (50) in the compression region (22) are essentially deadlocked. Crack propagation is apparently inhibited in the compacted material, so that the laser pulse energies, which decrease as described, in the burst packet can be adjusted particularly well so that the microcracks (50) terminate within the compression region (22), particularly within the compression measurement line. This line is not shown here for clarity.

[0137] A particularly advantageous embodiment is shown in Fig. 10. The representation is analogous to Figs. 7 to 9. In the plan view, the laser focus is widened in the direction of the parting line, in particular the intended parting line. This creates a preferred direction as described above. This has the consequence that the crack length RL of microcracks (50) in the preferred direction is greater than the crack length of microcracks (50) that do not run in the preferred direction. This makes it possible to arrange the filaments (1) further apart from one another in the preferred direction, microcracks in the preferred direction form connected microcracks (51) and thus create a parting line, but only fewer and / or shorter microcracks continue into the material of the workpiece (21) and thus less weakening of the edge occurs.The described reduction in individual pulse energies in the burst packet makes it possible, in particular, to confine microcracks outside the preferred direction in the compression region (22), particularly within the compression measurement line, while microcracks (50) propagate in the preferred direction outside the compression region. This achieves a particularly good compromise between machining efficiency and edge stability.

[0138] The relationship between the crack length RL of an individual microcrack and the angle Q of its extension to the preferred direction is schematically shown in Fig. 11. The information derived from Fig. 11 is qualitative. It shows the average crack length RL plotted against the angle Q. As mentioned, the angle Q indicates the angle of the propagation direction of a microcrack to the axis of the preferred direction, illustrated here using the example of an elliptical focus expansion with the main axis of the ellipse in the preferred direction.

[0139] There is a highly nonlinear relationship between RL and Q. Perpendicular to the preferred direction, at a value of 90° for Q, the crack length is shortest and RL has the lowest value. In contrast, in the preferred direction, at a value of 0° for Q, the crack length is greatest. This demonstrates that the efficient alignment of the preferred direction toward the parting lines has a disproportionate advantage for the effectiveness of the process and edge strength.

[0140] The shape of the focus expanded in the preferred direction is not limited to the previously described elliptical shape in plan view. In principle, all applicable shapes are encompassed by the invention. Figure 12 shows some examples of possible plan views of foci expanded in the preferred direction. VR denotes the preferred direction as a vector. This lies on the intended dividing line (53). The first example is the elliptical expansion with the main axis on the dividing line (53). Also possible are a drop-shaped expansion, a heart-shaped expansion, and / or a triangular expansion with the axis of symmetry on the dividing line (53).

[0141] As has been shown, the invention has the overall advantage that microcrack propagation is controlled and occurs to a reduced extent in the material of the workpiece (21). This improves the material strength of structured workpieces and thus supports further processability and avoids scrap. Overall, the filaments (1) can also be arranged closer to one another without unduly jeopardizing the material strength. In addition to the potential for progressive miniaturization of components, this also enables the avoidance of scrap. In particular, when separating brittle workpieces (2), the conflicting objectives are resolved: on the one hand, to provide efficient separation of brittle workpieces through the shortest possible processing time and thus the greatest possible distance between the individual filaments (1) on a parting line, and on the other hand, to ensure high edge strength of the machined workpiece.

[0142] List of reference symbols

[0143] 1 filament

[0144] 2 Workpiece

[0145] 3 lasers

[0146] 4 laser beam

[0147] 21 Material of the workpiece

[0148] 22 compression range

[0149] 23 Compression measuring line

[0150] 30 beam-shaping optics

[0151] 41 First pulse of a burst packet

[0152] 42 Last pulse of a burst packet

[0153] 50 microcrack

[0154] 51 connected microcrack

[0155] 52 no microcrack

[0156] 53 dividing line

[0157] 100 device

[0158] 200 positioning device

[0159] 210 Control device

[0160] 220 hot area

[0161] 221 roll

[0162] 250 pulling tank

[0163] 300 shielding element

[0164] 310 Sensor device

[0165] Tb burst duration

[0166] Ti Intra Burst Delay

[0167] Td Inter Burst Delay

[0168] BG Burst Degression

[0169] Epo pulse energy of the first pulse of a burst packet

[0170] Ep x Pulse energy of the last pulse of a burst packet

[0171] A filament axis

[0172] O Top

[0173] U Bottom side dO Filament diameter at the top side dU Filament diameter at the bottom side s Workpiece thickness a Channel angle

[0174] RL crack length

[0175] VR preferred direction

[0176] Q Angle to the preferred direction

Claims

Patent claims 1. A method for inserting a hollow channel-shaped filament (1) into a brittle-breaking workpiece (2), in which a focused laser beam (4) of a pulsed laser (3) with a wavelength A is directed onto the workpiece (2) for processing, wherein the workpiece (2) is transparent to the wavelength A of the laser, wherein the filament (1) is produced by the irradiation of at least one burst packet of laser pulses, wherein the at least one burst packet has a burst duration Tb and comprises a sequence of ultrashort laser pulses having a single pulse energy Ep and an intra-burst delay Ti, wherein the single pulse energy Epo of the laser pulse at the beginning of the burst duration Tb is greater than the single pulse energy Ep x of the laser pulse at the end of the burst duration.

2. Method according to the preceding claim, wherein the burst packet has a burst degression BG of 40% to 90%, preferably 50% to 80%, particularly preferably 60% to 70%, wherein BG is determined by the ratio of the individual pulse energy Ep x of the last single pulse of the burst packet at the end of the burst duration Tb and the single pulse energy of the first single pulse Epo of the burst packet at the beginning of the burst duration is calculated in percent.

3. The method according to at least one of the preceding claims, wherein, starting with the second individual pulse of the burst packet, the individual pulse energy of a laser pulse is lower than the individual pulse energy of the previous individual pulse; preferably, the individual pulse energies decrease exponentially during the burst duration Tb.

4. Method according to at least one of the preceding claims, wherein in the first single pulse of a burst packet a filament-shaped initial channel is generated, in particular an initial hollow channel, which is widened by the following individual pulses of the burst packet.

5. Method according to at least one of the preceding claims, wherein a compression region (22) of compressed workpiece material is produced around the hollow channel-shaped filament (1), which has a higher density, in particular compared to the starting material.

6. The method according to at least one of claims 4 to 5, wherein the relative weight loss m rei per hollow channel-shaped filament (1 ) is less than 10%, measured by m rei = rriactuai / mtheo , particularly preferably the workpiece has a material compaction of at least 1% compared to the original material density, measured in the radius of 3 pm (23) around the longitudinal axis of the respective hollow channel-shaped filament (1 ).

7. Method according to at least one of the preceding claims, wherein the filament-shaped hollow channel (1) has a diameter of 100 nm to 3 pm.

8. The method according to at least one of the preceding claims, wherein the focused laser beam (4) generates a plasma in the workpiece (2), which forms the hollow-channel-shaped filament (1); preferably, the first individual pulse of the burst packet generates an initial plasma, which is supplied with energy and thus maintained by the following individual pulses of the burst packet.

9. The method according to claim 8, wherein the plasma creates microcracks (50) in the workpiece (2) which extend from the hollow channel-shaped filament (1) into the volume of the workpiece.

10. The method according to claim 9, wherein the focused laser beam (4) is expanded in a direction perpendicular to the axis of the filament (VR) and the Length and / or number of microcracks (50) in the direction of widening of the focus line is greater than in the direction perpendicular to the widening.

11. Method according to at least one of the preceding claims 5 to 10, wherein at least 90% of the microcracks (50) end within the compression region (22), preferably within a radius of 3 pm (23) around the longitudinal axis of the hollow channel-shaped filament (1).

12. The method according to at least one of claims 8 to 11, wherein the plasma emits X-rays which are attenuated and / or shielded by at least one shielding element (300); preferably the shielding element (3) is attached such that it at least partially surrounds the workpiece (2), and wherein the shielding element consists of or comprises a material which absorbs X-rays; particularly preferably the material is selected from the group of metals, in particular comprising iron, steel, stainless steel, tungsten, or selected from the group of plastic composites, in particular comprising metal-filled plastic composite, tungsten-filled plastic composite.

13. Method according to at least one of the preceding claims, wherein the hollow channel-shaped filament (1) is a channel which connects one surface (O, U) of the workpiece to another surface (O, U) and penetrates the workpiece (2), or is a channel which ends as a blind hole in the volume of the workpiece (2).

14. Method according to at least one of the preceding claims, wherein the laser wavelength A is 500 nm to 1100 nm and / or the pulse duration of a single pulse is from 200 fs to 11 ps and / or the burst duration is from 1 ps to 500 ns and / or the single pulse energy of the first single pulse at the beginning of the burst duration is from 100 pJ to 1000 pJ and / or wherein the workpiece consists of or comprises glass and / or glass ceramics and / or Silicon and / or ceramic; preferably the distance between two adjacent filaments is from 3 pm to 10 pm.

15. Device for the laser-assisted introduction of a hollow channel-shaped filament (1) into a brittle-breaking workpiece (2), in particular a device for applying a method according to claims 1 to 14, with an ultrashort pulse laser (3) with a wavelength A, in which the workpiece (2) is transparent to the irradiated wavelength, and a beam-shaping optics (30) which is designed such that the laser beam (4) is directed onto the workpiece with a focus line and the focus line lies at least partially within the workpiece (2), wherein the laser (3) is set up such that it emits at least one burst packet of laser pulses, wherein the at least one burst packet has a burst duration Tb and comprises a sequence of ultrashort laser pulses which have an individual pulse energy Ep and an intra-burst delay Ti, and wherein the individual pulse energy Epo of the laser pulse at the beginning of the burst duration Tb is greater than the individual pulse energy Ep xof the laser pulse at the end of the burst duration.

16. Device according to claim 15, comprising a device with which the reduced energy of the last individual pulse Ep x at the end of the burst duration Tb in comparison to the energy Epo of the first individual pulse at the beginning of the burst duration; preferably, the energy drop of the individual pulses within the burst duration is adjustable, in particular as a linear or exponential decay behavior.

17. Device according to at least one of claims 15 to 16, with a shielding element for X-ray radiation, which at least partially encloses the workpiece; preferably the shielding element consists of a material or comprises a material selected from the group of metals, in particular comprising iron, steel, stainless steel, tungsten, or selected from the group of Plastic composites, in particular plastic composites containing tungsten.

18. Use of a method or a device according to at least one of the preceding claims 15 to 17 for separating and / or structuring and / or preparing for etching glass and / or glass-ceramic and / or silicon elements.

Images