WO2023237613A1

WO2023237613A1 - Method and device for processing at least one sub-region of a layer system

Info

Publication number: WO2023237613A1
Application number: PCT/EP2023/065254
Authority: WO
Inventors: Daniel Grossmann; Marc Sailer
Original assignee: Trumpf Laser Gmbh
Priority date: 2022-06-10
Filing date: 2023-06-07
Publication date: 2023-12-14
Also published as: DE102022114637A1

Abstract

The present invention relates to a method and a device for processing at least one sub-region (1) of a layer system, in particular a micro LED, by means of ultrashort laser pulses of a laser beam (20) of an ultrashort pulse laser (2), wherein the at least one sub-region (1) is positioned at a boundary layer (130) on a substrate (30), wherein the substrate (30) is substantially transparent to the wavelength of the ultrashort laser pulses of the laser beam (20), wherein the ultrashort laser pulses (200) of the laser beam (20) are applied to the boundary layer (130) through the substrate (30), as a result of which the at least one sub-region (1) is detached, wherein the laser beam (20) is shaped into a multifocus distribution (22) by means of a beam-shaping device (4), and the multifocus distribution (22) is applied to the boundary layer (130).

Description

Method and device for processing at least a portion of a layer system

Technical area

The present invention relates to a device and a method for processing at least a portion of a layer system, in particular a micro-LED, using ultrashort laser pulses of a laser beam of an ultrashort pulse laser.

State of the art

In modern display systems, micro-LEDs are used as light sources to make conventional, global backlighting of the display panel unnecessary. For this purpose, the micro-LEDs are applied to a so-called active matrix so that each micro-LED can be controlled individually. The displays with micro-LEDs therefore have the advantage that the micro-LEDs on the matrix can be controlled locally. This means that individual micro-LEDs can be switched on and off, which means, for example, a high contrast ratio and a very good black level can be achieved. In addition, so-called bleeding of the background lighting through the display panels can be avoided due to their design.

However, the production of such displays with micro-LEDs is challenging because the micro-LEDs are first produced on a substrate and then have to be transferred to a corresponding active matrix. In addition, only functioning micro-LEDs may be transferred to the active matrix in order to avoid pixel errors, for example. The use of lasers has proven successful for transferring the micro-LEDs to a corresponding active matrix. In particular, the Laser Lift-Off (LLO) process, with which a micro-LED can be detached from the substrate, and the so-called Laser Induced Forward Transfer (LIFT) process, with which the micro-LED is transferred from the substrate to the active matrix can become relevant. In both cases, a laser pulse is applied to the boundary layer between the substrate and the micro-LED in order to achieve appropriate processing of the micro-LED. “Laser Processing of Micro-LED - a Coherent Whitepaper” from January 23, 2018 describes that excimer lasers are typically used for the aforementioned processes.

However, excimer lasers have a low temporal pulse quality. In particular, the difficult to control transient response and decay behavior of the laser pulses influences the processing quality. In addition, the pulse length of excimer lasers is in the range of a few 10ns, so that the laser pulses interact with the boundary layer between the substrate and the micro-LED over a long period of time and accordingly have a thermal influence on the exposed materials. This can damage the semiconductor junctions, in particular the pn junctions, of the microLEDs.

In addition, the laser beam of an excimer laser does not have a defined polarization, so that they are particularly not suitable for high-performance applications, since the optical components necessary for high laser powers often require a defined polarization of the laser beam. In addition, the boundary layer between the micro-LED and the substrate is typically exposed to a flat-top beam. However, pure flattop beam shaping is inflexible and does not allow adaptation to the actual geometry of the micro-LED.

US8742289 discloses a laser lift-off process in which a masked laser beam is intended to process the microLEDs, so that individual processing of individual microLEDs is difficult. In particular, the geometry of the component to be removed, or the micro-LED, is not replicated.

JP2017221969 discloses a laser lift-off device in which the laser beam has a line shape, the line shape being formed from multiple beams.

Presentation of the invention

Based on the known prior art, it is an object of the present invention to provide an improved method for processing at least a portion of a layer system and a corresponding device.

The object is achieved by a method for processing at least a portion of a layer system with the features of claim 1. Advantageous further developments result from the subclaims, the description and the figures.

Accordingly, a method for processing at least a partial area of a layer system, in particular a micro-LED, by means of ultra-short laser pulses of a laser beam of an ultra-short pulse laser is proposed, wherein the at least one partial area on a Boundary layer is arranged on a substrate, wherein the substrate is essentially transparent to the wavelength of the ultra-short laser pulses of the laser beam, the boundary layer being acted upon by the ultra-short laser pulses of the laser beam through the substrate, whereby the at least one partial area is processed. According to the invention, the laser beam is shaped into a multi-focus distribution by means of a beam shaping device and the boundary layer is exposed to the multi-focus distribution.

The ultra-short pulse laser provides the laser pulses of the laser beam, with the individual laser pulses forming the laser beam in the direction of beam propagation. The pulse duration of the ultra-short laser pulses can be between 50fs and 1000ps, preferably between 100fs and 100ps, particularly preferably between 50fs and 20ps.

The laser pulses can also be part of a so-called laser burst, with each laser burst comprising several laser pulses. During the length of the laser burst, the laser pulses can follow one another very closely, for example at intervals of a few picoseconds to nanoseconds. The laser bursts can in particular be GHz bursts, in which the sequence of successive laser pulses of the respective burst takes place in the GHz range.

The sub-area of the layer system can in particular be a sub-area of a multi-layer system. For example, such a subarea is a micro-LED, a mini-LED or another LED. However, a layer system can also only comprise one layer of a material. The method is described below primarily using a microLED.

Micro-LEDs are generally localized pn junctions of semiconductors, in particular III/V semiconductors, such as gallium nitride GaN, which emit light of a specific wavelength, preferably a visible wavelength, when a voltage or current is applied. However, it is also possible for a micro-LED to also emit light in the UV range or in the IR range.

A micro LED has external dimensions of a few micrometers. For example, micro-LEDs can be round or rectangular and have a diameter or edge length of a few 10pm. For example, a micro LED can be square and 20x20pm2 or 30x30pm2.

The micro-LEDs are typically manufactured on a substrate. For this purpose, the p- and n-doped semiconductors are deposited one after the other onto the substrate. The first layer of the micro-LED is accordingly arranged at a boundary layer on the substrate. In addition, the first layer on the substrate has, for example, a p-doping and the second layer on the first layer has an n-doping. This occurs at the interface between the p- and n-doped layers a pn junction that produces light when a voltage or current is applied. The layers are then separated along the layer normal, so that a large number of micro-LEDs are created on the substrate.

What is particularly problematic here is that the substrate can influence the functionality of the micro-LEDs, since, for example, the substrate only allows insufficient removal of the heat generated when the micro-LEDs are operated. In addition, the substrate can deteriorate the optical properties if the light from the microLEDs first has to penetrate through the substrate before it reaches the observer. For this reason, it is advantageous to remove the substrate from the micro LED.

For this purpose, the substrate can be a substantially transparent material such as sapphire or a polymer or a plastic, or consist of silicon or comprise silicon, such as silicon carbide SiC. Substantially transparent means that the material can partially or completely transmit the laser light of the given wavelength, for example more than 80% or more than 85% or more than 90% or more than 95% or more than 99% of the light , so that the laser beam can penetrate through the substrate and the micro-LED underlying the beam propagation direction can also be processed. The laser beam can therefore be guided through the at least partially transparent substrate.

Finally, the laser energy can be at least partially absorbed in the boundary layer, since the micro-LED is typically not transparent to the laser wavelength. The boundary layer here comprises in particular a substrate-side section of the micro-LED, which is determined by the penetration depth of the laser beam into the micro-LED. However, the boundary layer can also or additionally comprise a buffer layer that has been applied to the substrate in order to reduce the lattice mismatch between the substrate and the micro-LED.

For example, in the case of a gallium nitride GaN micro-LED, the introduced laser energy in the boundary layer can separate the GaN into liquid gallium and gaseous nitrogen. This creates a vapor pressure in the boundary layer that separates the microLED from the substrate. However, it can also be the case that the micro-LED is detached from the substrate and, for example, transferred to a carrier substrate, for example to an underlying active matrix or a transfer layer. The processing of the sub-area, in particular the micro-LED, can therefore in particular be the laser-induced forward transmission of the sub-area and/or the laser lift-off of a sub-area of a substrate. The laser beam can be converted into a multi-focus distribution by a beam shaping device, this focus distribution being introduced into the boundary layer between the substrate and the partial region.

The term “focus” in general can be understood as a targeted increase in intensity, whereby the laser energy converges into a “focus area” or a “focus zone”. In particular, the term “focus” will be used below regardless of the beam shape actually used and the methods used to bring about an increase in intensity. By “focusing” the laser pulses, the location of the intensity increase along the beam propagation direction can be influenced. For example, the intensity increase can be quasi-point-shaped and the focus area can have a Gaussian-shaped intensity cross section, as provided by a Gaussian laser beam. The intensity increase can also be designed in a line shape, resulting in a Bessel-shaped focus area around the focus position, as can be provided by a non-diffracting beam. Furthermore, other more complex beam shapes are also possible, the focus position of which extends in three dimensions, such as a multi-focus distribution of Gaussian laser beams and/or non-Gaussian intensity distributions, as described below.

The multifocus distribution is a spatial distribution of individual focus zones, so-called individual foci. The focus zone is understood to be the part of the intensity distribution of the laser beam that is greater than the modification threshold of the material to be processed. The term focus zone makes it clear that this part of the intensity distribution is specifically provided and that an increase in intensity in the form of the intensity distribution is achieved through focusing.

A focus distribution means that the multifocus distribution includes at least two individual foci, with the individual foci being spatially separated from one another. However, the individual foci can all lie in one focal plane, so that the individual foci can all be found in one plane along the beam propagation direction of the laser beam, but have different coordinates in the focal plane perpendicular to the beam propagation direction.

A multi-focus distribution is provided by a so-called beam shaping device, whereby the incident laser beam can be divided into a plurality of partial laser beams that are guided to different individual foci. Beam shaping includes the design of the multi-focus distribution. The beam shaping can also include the design of the individual foci, such as the formation of Gaussian or non-diffractive laser beams. By distributing the laser energy to different individual foci, the laser energy is also distributed in the boundary layer. In this way, in particular, a flat processing of the partial area can be achieved, whereby lateral material stresses are reduced. In a sense, the multifocus distribution can be used to imitate a homogeneous flattop beam shape with the individual foci, although each individual focus is, for example, Gaussian. Accordingly, the requirements for the phase front of the laser beam are significantly lower than with a flattop beam. For example, the ultrashort pulse laser can be operated in its basic mode and does not require multimode operation. The short pulse duration of ultra-short pulse lasers compared to excimer lasers also enables the partial areas to be processed with significantly reduced heat input.

The repetition rate of the laser pulses can be greater than 1 kHz, preferably greater than 50 kHz, particularly preferably greater than 1 MHz.

The processing speed in particular can be adjusted by the repetition rate of the laser pulses. For example, a partial area can be processed with a laser pulse. Accordingly, more partial areas can be processed if a high repetition rate of the laser pulses is selected.

The repetition rate of the laser pulses can be, for example, 10MHz, so that over ten million partial areas can be processed per second. This enables a particularly high throughput of partial areas.

The wavelength of the laser beam can be between 50nm and 300nm, preferably between 250nm or 270nm.

The wavelength can in particular be designed so that the subregions at the wavelength have a particularly high absorption and the substrate has a low absorption. This allows the partial areas to be processed through the substrate. In addition, smaller focus zones can also be formed with small wavelengths, so that particularly small structures on the substrate can be targeted.

The ultrashort pulse laser can be operated in its basic mode, with the diffraction coefficient M ² being less than 1.5.

Accordingly, the basic transversal electrical mode TEM00 can be used as the laser mode, which has a pure Gaussian beam shape in the beam cross section and therefore in principle has the highest beam quality. In particular, TEOO fashion is... Beam cross section smaller than that of the higher modes, enabling more targeted processing of the partial areas. The diffraction coefficient indicates the deviation of the laser beam from the ideal TEOO mode.

The fluence can be between 0.05J/cm2 and 10J/cm2, preferably between 0.1J/cm2 and 1J/cm2.

The processing quality can be adjusted using the fluence, as the energy deposited in the boundary layer directly influences the LLO or LIFT process. For example, the deposited energy has an influence on the magnitude of the resulting vapor pressure, which detaches the partial areas from the substrate.

Preferably, all individual foci of the multifocus distribution can lie in the boundary layer. This allows particularly effective processing of the partial areas to be achieved.

The focus zones of the multi-focus distribution can be arranged within the geometric extent of the sub-regions and/or an envelope of the multi-focus distribution can have a shape that corresponds to the geometric extent of the sub-region.

For example, a partial area can have a dimension of 10x20pm2. The multifocus distribution can have 6 individual foci, with four in the corners of the sub-area and 2 in the middle of the long sides. For example, a partial area can have a dimension of 5x10pm2. The multifocus distribution can include 2 individual foci, with both individual foci being arranged at the opposite longitudinal ends. For example, a partial area can be elliptical and the multi-focus distribution can include two individual foci that are arranged in the foci of the ellipse. For example, a sub-area can be round, have a diameter of 30pm and the multi-focus distribution can include 5 individual foci that are evenly spaced on a circle with a diameter of 20pm. For example, a further individual focus can also be arranged in the middle of the partial area. For example, a micro-LED can be square and the individual foci can be arranged on a square grid.

The multi-focus distribution can be adapted to different partial area geometries. For example, the distribution of the individual foci in the boundary layer can be adjusted by the beam shaping device. This means that different sub-areas can be processed with a single device, so that particularly efficient processing of different sub-areas is achieved. This means that production costs in particular can be reduced. In particular, the individual focus can be smaller than the geometric dimensions of the partial area, preferably smaller than half, particularly preferably smaller than a tenth, of the geometric dimensions of the partial area.

For example, a partial area can be 20x30pm2, while an individual focus can have a diameter of less than 5pm, for example 1, 2, 3 or 4 pm. For example, an elliptical partial area can have a minor and a major semi-axis with a length of 5pm and 25pm respectively. Then the diameter of the individual foci can be, for example, smaller than 2.5pm or smaller than 0.5pm.

This makes it possible for the partial area to be processed in a targeted manner without damaging surrounding structures on the substrate. In particular, this allows several individual foci to be distributed in the geometry of the partial area without the individual foci overlapping and the laser beams running therein interfering with one another.

The spatial distance between the individual foci of the multifocus distribution can be smaller than five times the focus diameter, preferably smaller than twice the focus diameter. The focus diameter here is determined in particular by the beam waist of the individual laser beams that form the individual foci. The beam waist is in particular related to the wavelength of the laser beam, with a smaller wavelength enabling a smaller beam waist. The spatial distance between the individual foci of the multifocus distribution can be smaller than 1 opm. The distance between the individual foci is calculated from the center of the respective individual foci.

Adjacent individual foci of the multifocus distribution can have polarization that is orthogonal to one another.

Polarization describes the alignment, in particular the spatially or temporally variable alignment of the electric field vector of the laser beam relative to the direction of propagation of the laser beam.

Polarization is given here by a linear polarization or a circular or an elliptical polarization of the laser beams. The polarizations are orthogonal if the electric field vectors of the laser beams are orthogonal to one another at all times. For example, laser beams that are polarized perpendicular and parallel to the plane of incidence are orthogonal to each other. For example, left and right circularly polarized laser beams are perpendicular to each other. Orthogonal polarization means that the laser beams of the respective individual foci do not interfere with one another at small distances, so that the phase front of the laser beams in the individual foci is preserved. This allows the distribution of the laser energy to be particularly homogeneous be designed because, so to speak, the gaps between the individual foci of a first polarization are filled with individual foci of a second polarization, the first and the second polarization being orthogonal to one another. Material tensions in particular can be avoided by introducing laser energy in this way and the process quality can therefore be increased.

The multifocus distribution can have an intensity gradient; in particular, the individual foci can have at least partially different intensities.

An intensity gradient can be understood as an increase or decrease in intensity from one individual focus to another individual focus.

For example, a multifocus distribution can consist of 3x3 individual foci, with the middle individual focus having a higher intensity than the neighboring individual foci. However, it can also be the case that the central individual focus has a lower intensity than the neighboring individual foci. The latter can be particularly advantageous since the laser energy deposited in the middle of the sub-area spreads radially, so to speak, and processes the sub-area, while at the edge of the sub-area the laser energy deposited can act beyond the edge of the sub-area. Accordingly, a higher intensity at the edge of the multifocus distribution can enable better processing of the partial area overall.

The exposure of the boundary layer to the laser energy can thus be adjusted by means of an intensity gradient in order in particular to take into account the geometric properties of the partial area. For example, the distance between the individual foci can also be taken into account.

However, it is also possible for the individual foci to have a homogeneous intensity distribution.

The boundary layer can be acted upon simultaneously in the individual foci of the multifocus distribution, the loading occurring simultaneously if the maximum time interval of the loading in the individual spots is shorter than 10 ns, preferably shorter than 1 ns, particularly preferably shorter than 100 ps.

The impact on the boundary layer in the individual foci of the multifocus distribution is therefore simultaneous when the laser pulses reach the boundary layer within the typical pulse duration of an excimer laser.

This means that the laser beams that generate the multi-focus distribution can also reach the boundary layer in different ways. For example, one can Angular divergence of the individual partial laser beams means that the boundary layer is acted upon at different times for geometric reasons, but within the limits set above.

However, it is also possible for the individual foci to be specifically exposed to the laser energy at different times. For example, the laser pulse can be a laser burst, as shown above. The individual pulses of the laser burst can have a time interval that is less than the limit mentioned above. However, by definition, the individual pulses of the laser burst are still introduced into the boundary layer at the same time. For example, the laser burst can comprise 5 pulses at a distance of 0.1 ns, with the pulse length being 10ps. The individual pulses of the laser burst actually reach the boundary layer at different times. However, the individual pulses reach the boundary layer within 0.5 ns, so that by definition they are introduced into the boundary layer at the same time. The definition of simultaneity emphasizes that the individual pulses do not act as separate pulses, but rather function as a common unit in the boundary layer. For example, due to the small distance between the individual pulses, the accumulation of laser energy in the boundary layer can be faster than the removal of the laser energy through material-specific relaxation processes.

It is also possible for each individual pulse of the laser burst to be directed into a specific individual focus, so that each individual focus is subjected to a different individual pulse.

Such a temporal offset of the laser pulses within the above-mentioned limits further allows the individual foci to be tightly packed, i.e. also to overlap spatially, for example. However, because the individual pulses do not overlap in time, they cannot interfere with each other and distort the phase front of the laser beam. However, since the individual pulses simultaneously act on the boundary layer within the set limits, a further homogenization of the radiation field and thus an improvement in the process quality can be achieved.

The above object is further achieved by a device for processing a partial area of a layer system with the features of claim 12. Advantageous developments of the method result from the subclaims as well as the present description and the figures.

Accordingly, a device for processing a portion of a layer system, in particular a micro-LED, is proposed, comprising an ultra-short pulse laser, a beam shaping device and processing optics, the ultra-short pulse laser also being used is set up to provide a laser beam made up of ultra-short laser pulses, the partial region being arranged at a boundary layer on a substrate, the substrate being essentially transparent to the wavelength of the ultra-short laser pulses of the laser beam, the beam shaping device being set up to impress a multi-focus distribution on the laser beam , whereby the processing optics is set up to transfer the multi-focus distribution of the laser beam into the boundary layer and to apply the ultra-short laser pulses to it, whereby the partial area is processed.

For example, such processing optics can be a lens or a mirror or a telescope, with which in particular the position of the focal plane in the beam propagation direction can be determined. This allows the focal plane to overlap with the boundary layer between the substrate and the partial area.

The ultrashort pulse laser can provide a basic laser beam with a wavelength of 1030 nm and include at least one frequency doubling crystal, which is designed to double the frequency of the incident laser beam or to halve the wavelength.

The basic laser beam can, for example, be guided twice through the frequency doubling crystal, preferably sequentially through a first frequency doubling crystal and then through a second frequency doubling crystal, whereby the frequency of the basic laser beam is quadrupled overall, or the wavelength of the basic laser beam is quartered overall. As a result, the ultrashort pulse laser can provide a total laser beam with a wavelength between 250nm and 270nm.

The ultra-short laser pulses are particularly suitable for such frequency doubling due to their high pulse peak intensity, as these non-linear effects typically scale with the intensity used. This also makes it particularly easy to provide wavelengths in the UV range or DUV (deep-UV) range.

The beam shaping device may be an acousto-optical deflector and/or a microlens array and/or a diffractive optical one.

A diffractive optical element is designed to influence the incident laser beam in one or more properties in two spatial dimensions. Typically, a diffractive optical element is a specially shaped diffraction grating, whereby the incident laser beam is brought into the desired beam shape through diffraction. In an acousto-optical deflector unit, an acoustic wave is generated using an alternating voltage on a piezo crystal in an optically adjacent material, which periodically modulates the refractive index of the material. The wave can propagate through the optical material, for example as a propagating wave or as a wave packet, or as a standing wave. The periodic modulation of the refractive index creates a diffraction grating for an incident laser beam. An incident laser beam is diffracted at the diffraction grating and thereby at least partially deflected at an angle to its original beam propagation direction. The grating constant of the diffraction grating and thus the deflection angle depends, among other things, on the wavelength of the acoustic wave and therefore on the frequency of the applied alternating voltage. For example, deflections in the x and y directions can be created using a combination of two acousto-optical deflectors in a deflector unit.

The acousto-optical deflector unit can in particular be a polarization-dependent acousto-optical deflector unit and can therefore be particularly suitable for performance. For example, the actus-optical deflector unit can be a quartz-based deflector unit.

Microlens arrays include arrangements of multiple microlenses. Microlenses are small lenses, in particular lenses with a typical distance from lens center to lens center (“pitch”) of 0.1 to 10 mm, preferably 1 mm, whereby each individual lens in the arrangement can have the effect of a normal, macroscopic lens. With the multiple microlens arrays, an angular spectrum is generated from the (at least essentially) collimated input laser beam, with a large number of partial laser beams being created depending on the distance between the microlens arrays due to interference and diffraction effects. The variable change in the interference pattern results in a variation in the number of partial laser beams.

At least one optical component of the device can comprise CaF or crystalline quartz, or can be made of CaF or crystalline quartz and/or can be designed as a reflective element.

For example, the processing optics can be such an optical component. However, it is also possible for a mirror and/or a lens system, which guides the laser beam from the ultrashort pulse laser to the beam shaping device, to be such an optical component.

Optical components made of CaF or crystalline quartz are particularly suitable for UV and/or DUV applications because they have very low absorption for the short wavelengths. Reflective elements have the advantage that the UV rays do not penetrate the material optical components penetrate and interact with it. In particular, undesirable non-linear effects can be avoided.

The device can have a controller that is communicatively connected to the ultrashort pulse laser and the beam shaping device and is set up to control the ultrashort pulse laser and the beam shaping device.

For example, the controller can be an FPGA and/or a computer and/or a microchip. By communicatively connecting the controller to the beam shaping device and the ultrashort pulse laser, the controller can send and receive corresponding control signals.

In particular, it is possible for the pulse output to be controlled by the ultrashort pulse laser. In particular, this makes it possible for the adjustment of the focus position to be controlled by the beam shaping device. In particular, it is possible for the adjustment of the focus position and the delivery of the laser pulses to be coordinated with one another.

The control commands or their execution can be synchronized with the seed frequency of the laser in all connected devices, so that a common time base exists for all components.

The device can have a feed device, preferably a scanner, particularly preferably a galvano scanner, which is set up to direct the laser beam in the focal plane, the feed device being communicatively connected to the controller.

Both the feed device and scanner optics can be synchronized via the seed frequency, so that a common time base exists for the feed, beam deflection, beam shaping and control of the pulsed laser.

Short description of the characters

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

Figures 1A, B show a schematic representation of the method;

Figure 2A, B, C, D schematic representations of different multifocus distributions;

Figure 3 shows a further schematic representation of the method;

Figure 4 shows a schematic representation of the multi-focus distribution with individual foci with mutually orthogonal polarizations; Figure 5 shows a schematic representation of a laser burst;

Figure 6A, B, C shows a schematic representation of different multi-focus distribution;

Figure 7 shows a further schematic representation of the method;

Figures 8A, B show an intensity distribution of a multi-focus distribution and one with it

Processed micro LED; and

Figure 9 is a schematic representation of the device according to the invention.

Detailed description of preferred embodiments

Preferred exemplary embodiments are described below with reference to the figures. The same, similar or identical elements in the different figures are given identical reference numbers, and a repeated description of these elements is partly omitted in order to avoid redundancies.

1 A schematically shows a simplified processing method, in particular an LLO or LIFT method, in which a partial area 1 in the form of a micro-LED 1 is transferred from a substrate 30 directly to an active matrix 32. The simplified processing procedure applies analogously to general subareas 1 of the shift system.

For this purpose, a laser beam 20 is provided by an ultra-short pulse laser 2 (not shown), in which the ultra-short laser pulses 202 run. The ultra-short laser pulses 202 have a pulse duration of between 50fs and 1000ps, in particular between 100fs and 10ps. In addition, the wavelength of the laser beam 20 is between 50nm and 300nm, preferably between 250 and 270nm. The repetition rate of the laser pulses 202 can be greater than 10 kHz, preferably greater than 50 kHz, particularly preferably greater than 1 MHz and/or the fluence can be between 0.05J/cm2 and 10J/cm2, preferably between 0.1J/cm2 and 1J /cm2.

In addition, the laser 2 can be operated in its basic mode. Accordingly, the basic transversal electrical mode TEM00 can be used as the laser mode, which has a pure Gaussian beam shape in the beam cross section and therefore in principle has the highest beam quality. In particular, in the TEM00 mode the beam cross section is smaller than in the higher modes, so that a more targeted processing of the partial areas 1 is possible.

The partial area 1 in the form of a micro-LED 1 is arranged in the initial state at a boundary layer 130 on the substrate 30. For example, the substrate 30 can be made of sapphire and can have a material thickness between 100 pm and 5 mm. However, it is also possible for the substrate 30 to be a carrier layer or a carrier substrate, such as a film, an adhesive or a wafer, from which the micro-LEDs 1 are transferred to a display panel in a further step. In particular, the layer thicknesses in Figures 1 A, B are not drawn to scale.

The micro-LEDs 1 can have several layers 11, 12, 13, for example a layer 11 made of n-doped gallium nitride GaN and a layer 13 made of p-doped GaN. A pn junction 12 is formed between the two layers 11, 13, in which light of a characteristic wavelength is produced when a current or voltage is applied. The layers 11, 13 can be between 1 nm and 10 pm thick, for example. For example, the vertical cuts may already be present in the layer system, or may have been introduced into the layer system in a previous process step, so that the micro-LEDs 1 to be removed are only connected to the substrate via the boundary layer 130.

A multi-focus distribution 22 is impressed on the laser beam 20 by a beam shaping device 4, which in the example shown has three different individual foci, all of which lie in the boundary layer 130. As a result, the boundary layer 130 between the partial region 1 and the substrate 30 is exposed to the laser beam 20 or the laser pulses 202 running therein. Since the substrate 30 is transparent to the wavelength of the laser beam 20, the laser beam 20 can be guided with the multi-focus distribution 22 through the substrate 30 to the boundary layer 130.

For example, the GaN of the first layer 11 of the micro-LED 1 may not be transparent to the wavelength of the laser beam 20, so that the laser energy introduced by the ultra-short laser pulses 202 is absorbed there. The energy introduced can separate the GaN into liquid gallium and gaseous nitrogen at the interface, whereby a vapor pressure is formed at the interface 130, which detaches the micro-LED 1 from the substrate 30. At the same time, the micro-LED 1 can be separated from the substrate, for example by laser ablation at the interface 130. To a certain extent, targeted damage along the boundary layer 130 can be achieved through the interaction of the laser beam 20 with the first layer, so that the micro-LED 1 can be separated along the boundary layer 130. The detachment of the micro-LED 1 from the substrate 30 is called laser lift-off (LLO).

At the same time, the micro-LEDs 1 can be transferred to a carrier substrate 32 by the resulting forces. In particular, the micro-LEDs 1 can be mounted directly on a carrier substrate attached, for example glued or transferred to a carrier substrate, for example another wafer. In particular, the micro-LED 1 can be applied to an active matrix as a carrier substrate 32, which is designed to provide a voltage and/or power supply to the micro-LED 1.

In Figure 1B, the carrier substrate 32 is approximately an adhesive layer. Since the micro-LEDs 1 were separated from the substrate 30 at the boundary layer 130 and transferred to the adhesive layer 32, the desired micro-LEDs 1 remain on the carrier substrate 32 when the substrate 30 is removed.

Figure 2 shows a top view of the boundary layer 130 with the multi-focus distribution 22. The multi-focus distribution 22 is formed by a beam shaping device 4 from the laser beam 20 of the ultra-short pulse laser 2 and then introduced into the boundary layer 130. In particular, the multi-focus distribution 22 includes several individual foci 220. These individual foci can be arranged within the geometric extent of the partial area 1. This makes it possible for the boundary layer 130 of the partial region 1 to be exposed to the laser energy of the laser beam 20 over a certain area, so that the partial region 1 can be processed.

Figure 2A shows schematically the boundary layer 130 of a rectangular sub-region 1, within which a multi-focus distribution 22 made up of six individual foci 220 is arranged. The individual foci 220 are preferably evenly spaced and evenly distributed over the geometry of the partial area 1, since this enables particularly precise and gentle processing of the partial area 1.

2B shows schematically a boundary layer 130 of a rectangular subregion 1, within which a multifocus distribution 22 made up of four individual foci 220 is arranged, the individual foci 220 being arranged along a line.

2C shows schematically a boundary layer 130 of a round partial area 1, within which a multi-focus distribution 22 made up of six individual foci 220 is arranged. Here, an individual focus is located in the middle of the sub-area, while the five remaining individual foci are evenly spaced radially from the center.

2D shows schematically a boundary layer 130 of an elliptical partial area, within which a multi-focus distribution 22 made up of two individual foci 220 is arranged. The two individual foci 220 are in the two focal points of the ellipse. The focus diameters of the individual foci are significantly smaller than the diameter of the partial area 1. For example, the diameter of the individual foci can be approximately 6pm if the partial area 1 has geometric dimensions of the order of 30pm. In all of the above examples the Center distance of the individual foci 220 smaller than five times the focus diameter of the individual foci. This allows a particularly homogeneous impact on the boundary layer 130 to be achieved. In particular, in the present cases, the center distance between neighboring individual foci can also be smaller than 10pm. In particular, the envelope of the multifocus distribution has a shape that corresponds to the geometric extent of the partial area. 3 shows a method with which several partial areas 1 can be processed successively. For this purpose, the multifocus distribution 22 is reoriented on the substrate 30 between each laser pulse 202 of the ultrashort pulse laser 2. Each laser pulse 202 of the ultrashort pulse laser 2 is divided into six individual foci 220 and introduced into the boundary layer 130. Each laser pulse 202 has a pulse energy of more than 1 pJ, so that about a sixth of it arrives per individual focus. With a diameter of the individual foci of 20pm, the fluence is approximately 0.05J/cm2. In particular, the partial areas 1 can be processed with the repetition frequency of the ultra-short pulse laser 2. If the repetition frequency is around 10MHz, ten million partial areas 1 can be processed per second. An active matrix 32 of a display with 3840 by 2160 micro-LEDs 1 could theoretically be populated in less than a second.

A further embodiment of the method is shown in FIG. The center distance corresponds approximately to a focus diameter of the individual foci 220 or even less. The individual foci 220 can therefore partially overlap, so that the associated partial laser beams can interfere with one another and thus distort the phase front. In particular, this can lead to an attenuation of the laser intensity in the individual foci 220. However, if neighboring individual foci 220 have different polarizations, no such unfavorable influence on neighboring individual foci 220 takes place. In particular, a first individual focus 220 can have an s-polarization (as perpendicular to the plane of incidence) and an adjacent individual focus 220 can have a p-polarization (i.e. parallel to the plane of incidence). However, it is also possible for a first individual focus 220 to have an r-polarization (as right-hand circular) and an adjacent individual focus 220 to have an I-polarization (i.e. as left-hand circular). In both cases, the polarization of adjacent individual foci 220 is orthogonal to one another. This allows the individual foci 220 to be packed more densely, so that a more homogeneous impact on the boundary layer 130 can be achieved and thus the process quality can be improved.

The different multifocus distributions 22 can also have intensity gradients. For example, in FIG. 2C, the boundary layer 130 in the central individual focus 220 can be subjected to a lower intensity. For example, in FIG. 2A there may be a homogeneous intensity distribution, so that the boundary layer 130 is exposed to the same intensity in each individual focus 220. For example, in Figure 3 an additional Intensity gradient runs diagonally across the interface 130 so that the intensity in the individual foci 220 is greatest in the top left corner and drops towards the bottom right corner. With such a tailored intensity gradient, the intensity distribution and thus also the process quality can be optimized.

Figure 5 shows the time course of a laser burst. The laser burst 200 comprises several individual pulses 202, the distance between the individual pulses 202 being less than 10 ns. By introducing the individual laser pulses 202 simultaneously within this time window, by definition, further measures can be taken to pack the individual foci 220 even more densely, as shown in Figure 6.

Figure 6A shows a multi-focus distribution 22 of 2 times 3 individual foci at a first time t1, where the center distance is approximately one focus diameter. Figure 6B shows a further multi-focus distribution 22 of 2 times 3 individual foci at a second time t2. The second multi-focus distribution 22 is spatially shifted to the first multi-focus distribution 22. In the superposition of the two figures in FIG. 6C it becomes clear that the second multi-focus distribution 22 fills the gaps in the first multi-focus distribution 22. If the two multifocus distributions are introduced into the boundary layer at a time interval of less than 10ns, for example in less than 1 ps or less than 1fs, then the boundary layer 130 is, by definition, simultaneously acted upon in both multifocus distributions 22 in FIG. 6C, so that through the actual Temporal offset enables a particularly dense packing of the individual foci 220.

7 shows how such a method can be used to process several partial areas 1. Here, analogous to FIG. 3, the multifocus distribution 22 is repositioned between each laser burst. For each partial area 1, the boundary layer 130 is acted upon simultaneously because the time interval between the individual pulses 202 is significantly smaller than 10 ns. This means that the partial areas 1 can be processed over a large area.

In Figure 8A, the intensity distribution of a multi-focus distribution 22 is shown schematically, which consists of 4 by 4 individual foci. White stands for high intensity, while black stands for low intensity. In Figure 8B, a micro-LED 1 was removed with the intensity distribution shown. It can be clearly seen that the boundary layer 130 carries the signature of the intensity distribution from FIG. 8A and that a flat detachment and thus processing of the micro-LED 1 was made possible. A device according to the invention is shown schematically in FIG. Here, the ultra-short pulse laser 2 is set up to provide a laser beam 20 made up of ultra-short laser pulses 202. The beam shaping device 4 is set up to impose a multi-focus distribution on the laser beam 20, while the processing optics 5 is set up to transfer the multi-focus distribution 22 of the laser beam 20 into the boundary layer 130 between partial area 1 and substrate 30 and to apply the ultra-short laser pulses 202 to this, whereby partial area 1 is processed.

The ultra-short pulse laser 2 can in particular provide a basic laser beam with a wavelength of 1030 nm and include a frequency doubling crystal (not shown) which is designed to double the frequency of the incident laser beam or halve the wavelength, the basic laser beam being guided, for example, successively through two frequency doubling crystals . As a result, the frequency of the basic laser beam is quadrupled overall, or the wavelength of the basic laser beam is quartered overall. A laser beam 20 with a wavelength between 250nm and 270nm is therefore provided.

The beam shaping device can in particular be an acousto-optical deflector and/or a microlens array and/or a diffractive optical element which generates a multi-focus distribution 22 from the laser beam 20. The partial laser beams, which are formed from the laser beam 20 by the beam shaping device, can be transferred via processing optics 5 into the respective individual foci 220 of the multi-focus distribution 22.

The optical components of the device can comprise CaF or crystalline quartz, or are made of CaF or crystalline quartz and/or can be designed as reflective elements, making the optical components particularly suitable for UV and DUV applications.

In addition, the device can have a feed device 6 in order to move the laser beam 20 and/or the substrate 30 with the partial region 1 relative to one another. For example, such a feed device can be an xyz table.

The device can have a controller 7 which coordinates the various components of the device with one another. For example, the ultra-short pulse laser 2 can be connected to the controller 7, so that the controller 7 can trigger the delivery of a laser burst. The controller 7 can further be connected to the beam shaping device 4, so that the multifocus distribution 22 can be quickly repositioned. However, this can also be used, for example, to set a different geometric shape of the multi-focus distribution 22, so that the multi-focus distribution 22 is adapted to the geometric shape of the partial area 1. The Control 7 can also be connected to the feed device 6, so that the feed device 6 can be moved, for example, between the laser bursts or the laser pulses 202 in order to achieve a rough positioning of the substrate under the processing optics. To the extent applicable, all individual features shown in the exemplary embodiments can be combined and/or exchanged with one another without departing from the scope of the invention.

Reference character list

1 section

11 first layer

12 pn junction 13 second layer

2 ultra-short pulse lasers

20 laser beam

200 laser burst

202 laser pulse 22 multi-focus distribution

220 single focus

30 substrate

32 additional substrate

4 Beam shaping optics 5 Processing optics

6 feed device

7 Control

Claims

Expectations

1 . Method for processing at least one partial region (1) of a layer system, in particular a micro-LED, by means of ultra-short laser pulses of a laser beam (20) of an ultra-short pulse laser (2), the at least one partial region (1) being at a boundary layer (130) on a substrate ( 30), wherein the substrate (30) is essentially transparent to the wavelength of the ultra-short laser pulses of the laser beam (20), the boundary layer (130) passing through the substrate (30) with the ultra-short laser pulses (200) of the laser beam ( 20) is applied, whereby the at least one partial area (1) is detached, the laser beam (20) being shaped into a multi-focus distribution (22) by means of a beam shaping device (4) and the boundary layer (130) being acted upon by the multi-focus distribution (22). .

2. The method according to claim 1, characterized in that the processing of the sub-area (1) is a laser-induced forward transmission of the sub-area (1) and / or a laser lift-off of the sub-area (1).

3. Method according to one of claims 1 or 2, characterized in that

- the pulse duration of the ultra-short laser pulses is between 50fs and 1000ps, preferably between 100fs and 100ps, particularly preferably between 50fs and 20ps and/or

-the repetition rate of the laser pulses is greater than 10kHz, preferably greater than 50kHz, particularly preferably greater than 1 MHz and/or,

-the fluence is between 0.05J/cm2 and 10J/cm2, preferably between 0.1 J/cm2 and 1J/cm2 and/or

- The wavelength of the laser beam (20) is between 50nm and 300nm, preferably between 250nm and 270nm and/or

- The laser (2) is operated in its basic mode, with the diffraction coefficient M ² being less than 1.5.

4. Method according to one of the preceding claims, characterized in that the individual foci (220) of the multi-focus distribution (22) lie in the boundary layer (130). Method according to one of the preceding claims, characterized in that the individual foci (220) of the multi-focus distribution (22) are arranged within the geometric extent of the sub-area (1) and/or that an envelope of the multi-focus distribution (22) corresponds to the geometric extent of the sub-area (1) has a corresponding shape. Method according to one of the preceding claims, characterized in that multi-focus distribution (22) is adapted to different partial area geometries. Method according to one of the preceding claims, characterized in that the individual foci (220) of the multi-focus distribution (22) are smaller than the geometric dimensions of the partial area (l), preferably smaller than half, particularly preferably smaller than a tenth, of the geometric dimensions of the Partial area (1). Method according to one of the preceding claims, characterized in that the spatial distance of the individual foci (220) of the multi-focus distribution (22) is less than five times the focus diameter, is preferably smaller than twice the focus diameter and/or the spatial distance of the individual foci (220) the multifocus distribution (22) is less than 10pm. Method according to one of the preceding claims, characterized in that adjacent individual foci (220) of the multi-focus distribution (22) have polarization that is orthogonal to one another. Method according to one of the preceding claims, characterized in that

- the multifocus distribution (22) has an intensity gradient, in particular the individual foci (220) have at least partially different intensities or

- The individual foci (220) have a homogeneous intensity distribution. Method according to one of the preceding claims, characterized in that the boundary layer (130) in the individual foci (220) of the multi-focus distribution (22) is acted upon simultaneously, the loading taking place simultaneously when the maximum time interval of the loading in the individual foci (220) shorter than 10 ns, preferably shorter than 1 ns, particularly preferably shorter than 100 ps. Device for processing a partial area (1) of a layer system, in particular a micro-LED, comprising an ultra-short pulse laser (2) of a beam shaping device (4) and processing optics (5), the ultra-short pulse laser (2) being set up to produce a laser beam (20). from ultra-short laser pulses, the partial region (1) being arranged at a boundary layer (130) on a substrate (30), the substrate (30) being at least partially transparent to the wavelength of the ultra-short laser pulses of the laser beam (20), the Beam shaping device (4) is set up to impose a multi-focus distribution (22) on the laser beam (20), the processing optics (5) being set up to transfer the multi-focus distribution (22) of the laser beam (20) into the boundary layer (130) and this with the ultra-short laser pulses, whereby the partial area (1) is processed. Device according to claim 12, characterized in that the ultra-short pulse laser (2) provides a basic laser beam with a wavelength of 1030nm and comprises at least one frequency doubling crystal which is set up to double the frequency of the incident laser beam or to halve the wavelength, the basic laser beam being used twice is guided through the frequency doubling crystal, preferably sequentially through a first frequency doubling crystal and then through a second frequency doubling crystal, whereby the frequency of the basic laser beam is quadrupled overall, or the wavelength of the basic laser beam is quartered overall and thus a laser beam (20) with a wavelength between 250 nm and 270nm is provided. Device according to one of claims 12 or 13, characterized in that the beam shaping device (4) is an acousto-optical deflector and/or a microlens array and/or a diffractive optical element. Device according to one of claims 12 to 14, characterized in that at least one optical component of the device comprises CaF or crystalline quartz, or is made of CaF or crystalline quartz and / or is designed as a reflective element. Device according to one of claims 12 to 15, characterized by a control (7) which communicates with the ultra-short pulse laser (2) and the beam shaping device (4). is connected and is set up to control the ultra-short pulse laser and the beam shaping device (4). Device according to claim 16, characterized by a feed device (6), preferably a scanner, particularly preferably a galvano scanner (62), for moving the laser beam (20) in the focal plane, the feed device (6) being communicative with the control (7 ) connected is.