DE102008033358A1 - Production of laser beam for surface processing, comprises emitting the laser beam with first beam parameter product from laser beam source and then deforming into laser beam with predeterminable beam parameter product via optical unit - Google Patents

Abstract

The method of producing a laser beam for surface processing, comprises emitting the laser beam with a first beam parameter product from a laser beam source and then deforming into a laser beam with a predeterminable beam parameter product via an optical unit, where the beam parameter product is less than or equal to the predeterminable beam parameter product; the beam parameter product in x-segment is not equal to the predeterminable beam parameter product in x-segment; and the beam parameter product in y-segment is not equal to the predeterminable beam parameter product in y-segment. The method for the production of a laser beam for surface processing, comprises emitting the laser beam with a first beam parameter product from a laser beam source and then deforming into a laser beam with the predeterminable beam parameter product via an optical unit, where the beam parameter product is less than or equal to the predeterminable beam parameter product; the beam parameter product in x-segment is not equal to the predeterminable beam parameter product in x-segment; the beam parameter product in y-segment is not equal to the predeterminable beam parameter product in y-segment; and all the products are orthogonal components. The laser beam has a beam cross section segmented into two segments, which are rhombic or rectangular, or represent disjoint surface of an ellipse with straight cut edges. The two segments correspond to the beam cross sections of two partial beams produced from the laser beam, where the partial beams are spatially separated from each other. The two segments are subjected to a transformation relevant to the spatial relative position, shape and/or size of the segments. The transformation is carried out in the frame of a reflection, a rotation and/or a lateral shifting by optical functional elements such as optical lenses, diffractive optical elements, curved or plane mirrors or prisms. The two transformed segments are joined to single deformed laser beam. The laser beam emitted by the laser beam source and/or the laser beam having the predeterminable beam parameter product are subjected to an optical stretching or compression in two orthogonal directions with respect to beam cross section. The two partial beams have a longitudinal beam cross section with a longitudinal extension and two lateral flank areas limiting the longitudinal beam cross section on both sides. The two partial beams are joined along their lateral flank areas under formation of an uniform, longitudinal laser beam or are joined under formation of a laser beam in the beam cross section in a stepped manner, where the beam cross sections of the partial beams are parallely directed with respect to respective longitudinal extension, or are arranged to each other in such a manner that the cross-sectional centre of gravity associated with the beam cross sections lies on a common connecting axis. The longitudinal beam cross sections are bent along their longitudinal extension around a uniform angle alpha of 0-90[deg] . An independent claim is included for a device for the production of a laser beam with a predeterminable beam parameter product.

Description

Technical area

The invention relates to an apparatus and a method for redistributing the beam parameter product of a laser beam,
hereafter referred to as SPP. The beam parameter product describes the focusability of a light beam, especially a laser beam, and can be mathematically described by the following relationship: SPP = θw 0 ,

Here, θ describes the opening angle of the bundle in the far field and w ₀ the beam radius or diameter at the location with the narrowest bundle diameter.

It is common to specify either the product of half the opening angle and radius or the product of full angle and diameter. For the frequent case of a so-called Gaussian beam, the SPP results specifically to SPP = θw 0 = M 2 · Λ / π,

Therein, θ is half the far-field angle, w _{0 is} the beam radius at the location of the narrowest bundle diameter, the so-called beam waist, λ is the wavelength and M ^{2 is} the so-called diffraction factor. θ and w ₀ denote the radial position at which the beam intensity has dropped to a fraction of 1 / e ² = 0.135 of its maximum axial value. M ² is a dimensionless number greater than or equal to 1 and describes the departure of a real laser beam from an ideal coherent Gaussian beam consisting of a TEM ₀₀ mode.

Basically, the larger the SPP or M ² , the worse the beam is to focus, ie, the larger the smallest possible focus diameter.

To describe the beam quality of a general laser beam, the SPP can expediently be represented in two components so z. For example, as a product of its orthogonal components SPP = SPPx * SPPy, for Gaussian beams specifically over the product of the SPP-proportional diffraction coefficients M ² = Mx ² * My ² . Excimer lasers and diode lasers, for example, have an anisotropic distribution with SPPx> SPPy where y describes the so-called short and x the so-called long beam cross-sectional axis. In contrast, many solid-state lasers have a largely isotropic distribution SPPx = SPPy or Mx ² = My ² . Problems exist in which the spatial distribution of the SPP of a given radiation source does not correspond to the distribution suitable for a particular application.

State of the art

The Necessity of laser beam shaping arises z. B. in connection with surface treatment applications where the beam a laser formed by means of an optical system and ultimately defined on a surface.

An application of current interest is, for example, in the field of the production of TFT and OLED displays, currently especially in the production of small high-resolution displays. There is typically one of the process steps in the crystallization of the provided for receiving the switching transistors, a few 10 nm thick layers of amorphous silicon on the glass substrates of the display. Such a silicon layer is locally melted by the laser radiation, and solidifies in the subsequent cooling in a polycrystalline structure. Both mask-imaging methods and methods in which a long and narrow focus line is guided over the surface are to be found. Representative of a related art is on the US 2005/0035103 A1 in which a considerable compilation is to be taken by means of the so-called ELA method (excimer laser annealing), in which an excimer laser beam is subjected to optical beam shaping in such a way that ultimately the silicon layer to be fused with a line-shaped light beam with a line length of up to about 500 mm and with a typical line width in the range of several 100 microns, down to a few microns is applied. The typically used excimer lasers are capable of emitting light at wavelengths of 157 nm, 193 nm, 248 nm, 308 nm, 351 nm. However, diode-pumped solid-state lasers, so-called DPSS lasers (diode pumped solid state), are also increasingly finding use, for example. Yb: YAG, Nd: YAG with visible wavelengths (515, 532 nm).

Further possible concrete applications for optical systems for surface finishing, in which the ones discussed below Procedures could be applied, for example the activation or annealing of dopant atoms in surface layers ("Doping activation"), the ablation of surface structures, or the "drilling" of very small holes, z. B. in films for inkjet printers.

Generally play the properties of the radiation source and their skillful Use and influence due to the growing demands to the quality of crystallization, the process speed and thus the optical image an ever larger Role.

In particular, for the production of very narrow lines, as explained above in connection with the production of TFT displays, the SPP of the radiation source limits the line width practically achievable in the process. To achieve, for example, an effective line width of approximately 5 μm required for the so-called SLS process (Sequential Lateral Solidification) or similar processes with process-suitable depth of focus, an SPP is required in the so-called short axis, ie perpendicular to the line extension of M ² = 3 ... 8. In the direction of the line axis, on the other hand, the so-called long axis, the highest possible SPP is desired in order to avoid interfering interference and speckle.

The Product of SPP components, SPPx · SPPy, is decreasing when passing through an optical system not, but remains equal or higher. This applies to conventional optical systems equally for the individual components of the SPP itself. Without higher losses, for example by trimming of the beam cross section, can be the SPP or one of their Do not reduce components.

Lots Laser types, z. As DPSS laser, have a relatively isotropic SPP and allow the fulfillment of the requirements at most in one of the two axes of the cross section. For this reason, done currently developments of high power lasers with the goal of a strongly anisotropic SPPs.

With respect to laser beam shaping optical devices, prior art references are known, from which, using cylindrical lens array arrangements, the merging of a plurality of individual beams into a total laser beam results. Such measures are typically taken when using diode laser bars which emit a plurality of parallel juxtaposed single laser beams, each with anisotropic beam divergence, which is to be unified for purposes of a particular beam usage into a uniform, mostly low-divergent single beam. For example, reference is made to the following documents: EP 1 006 382 B1 . DE 10 2004 034 253 A1 . EP 1 075 719 B1 . DE 198 19 333 A1 . US 2007/0024979 A1 ,

Presentation of the invention

Of the Invention is based on the object, a device and a Method for generating a laser beam with a demand-based SPP, with the measures to be taken on the one hand based on the emitted from a laser beam source Light output as lossless as possible and without realize big financial and equipment expenditure should be.

The Solution of the problem underlying the invention is specified in claim 1. The subject of claim 10 is a solution according to Contraption. The solution idea advantageous further education Features are the subclaims and the further description, in particular with reference to the illustrated embodiments refer to.

• 1. SPP₁ ≤ SPP₂,
• 2. SPP_1x ≠ SPP_2x,
• 3. SPP_1y ≠ SPP_2y und

SPP_1x und SPP_1y bzw. SPP_2x und SPP_2y jeweils zueinander orthogonale Komponenten sind.According to the solution, the method for producing a laser beam with a predeterminable beam parameter product, SPP _{2 for} short, is characterized in that a laser beam is emitted from a laser beam _source with a first SPP ₁ = SPP _1x SPP _1y and via at least one optical unit into a laser beam with the presettable laser beam SPP ₂ = SPP _2x · SPP _2y , where:

• 1. SPP ₁ ≤ SPP ₂ ,
• 2. SPP _1x ≠ SPP _2x ,
• 3. SPP _1y ≠ SPP _2y and

SPP _1x and SPP _1y and SPP _2x and SPP _2y are mutually orthogonal components.

The approaches approach according to the solution the problem of generating a laser beam with a possible anisotropic SPP, which can be suitably dimensioned depending on the application is, not from the influence of the laser beam source, as is the case in the prior art, but it is the Path follows, the predetermined by the radiation source SPP means redistribute suitable optical devices, in particular to from the radiation of an isotropic laser an anisotropic distribution to produce with appropriate orientation. Depending on the given laser type and concrete application is also the opposite problem conceivable, so the requirement of an isotropic SPP given Anisotropic radiating laser. In principle, the same considerations apply here as in the previous one.

• 1. SPP₁ ≤ SPP₂,
• 2. SPP_1x ≠ SPP_2x,
• 3. SPP_1y ≠ SPP_2y und

SPP_1x und SPP_1y bzw. SPP_2x und SPP_2y jeweils zueinander orthogonale Komponenten sind.A device _{according to the invention} for generating a laser beam with a predefinable beam parameter product, SPP _{2 for} short, has a laser beam _source from which a laser beam emerges with a first SPP ₁ with SPP ₁ = SPP _1x SPP _1y , which is medium or direct with at least one optical imaging unit interacts under formation of the laser beam with the prescribable SPP ₂ , with SPP ₂ = SPP ₂ × SPP _2y , where:

• 1. SPP ₁ ≤ SPP ₂ ,
• 2. SPP _1x ≠ SPP _2x ,
• 3. SPP _1y ≠ SPP _2y and

SPP _1x and SPP _1y and SPP _2x and SPP _2y are mutually orthogonal components.

The term "laser beam", in particular with regard to the transformed beam, in addition to egg A single beam with a contiguous beam cross section also means a bundle of partial beams which has emerged from a single laser beam and corresponds functionally to a single beam with respect to the specific application.

The solution principle used in the optical Redistribution of the beam parameter product of an output laser beam, which is emitted from a laser beam source is to be based on in the figures closer to illustrative embodiments described, but by their concrete form of realization do not restrict the general idea of solution should.

Is there a desire for a laser beam with a possible anisotropic beam distribution, ie a laser beam with the smallest possible value of My ² in one axis and the largest possible value Mx ² in the coordinate direction perpendicular thereto, and are available for this purpose only radiation sources available, but all Thus, a solution to the problem of redistributing the SPP, that is, reducing the SPP in one axis while increasing in the second axis, by means of a transformation with simple optical functional elements, ie, lenses, diffractive, is a solution to the problem of generating an appropriate anisotropic beam distribution optical elements, short DOEs, curved and plane mirrors or prisms.

basis of the method is the fact that the jet parameter product Although SPP is a conserved quantity in lossless beam propagation, or at least can not be reduced, that his Components SPPx, SPPy, however, can be changed according to SPPx · SPPy = SPP = const, or larger becoming.

In 1 Two basic procedures for SPP redistribution are shown using the example of a theoretical output laser beam 1 with a square beam cross section with an isotropic beam distribution, of M ² = 3 × 3. Here Mx ² = My ² = 3 is chosen concretely, which can be illustrated by a division into 3 × 3 coherence cells, or independent coherent modes. This representation is, of course, of a purely theoretical nature and generalizes and simplifies the model of coherent mode decomposition in Gaussian rays with the purpose of illustrating the fundamental relationships. A redistribution can be accomplished by dividing the beam cross section with 3 × 3 modes into 3 strips of 1 × 3 modes, and then reassembling them in a different arrangement to produce an anisotropic distribution. In the case shown, the strongest redistribution results, Mx ² = 9 and My ² = 1, if one ranks the three strips close to each other on their narrow sides (see case a)).

Other arrangements lead to My ² > 1, see case b, which will be returned below.

• 1. Segmentierung des Ausgangsstrahlquerschnitts in N Segmente,
• 2. Separation der Segmente (optional),
• 3. Transformation der Segmente und
• 4. Zusammensetzen der transformierten Segmente zu einer topologisch neuen Struktur.

A process for redistributing the SPP typically consists of the following steps:

• 1. Segmentation of the output beam cross-section into N segments,
• 2. Separation of segments (optional),
• 3. Transformation of segments and
• 4. Assemble the transformed segments into a topologically new structure.

A Redistribution without segmentation is conceivable in principle, appears technically, however, much more difficult and expensive than with segmentation.

• – Lateraler Versatz, z. B. unter Verwendung von Prismen. Dieser Fall ist in der linken Darstellung in 1 im Fall a) gezeigt.
• – Spiegelung, z. B. mittels gedrehter Zylinderlinsen sowie
• – Drehung, äquivalent zu einer geradzahligen Folge von Spiegelungen,

In principle, the following three optical options are available for the transformation of the individual segments, if necessary also further combinations of the three possibilities:

• Lateral offset, e.g. B. using prisms. This case is in the left illustration in 1 in case a).
• - reflection, z. B. by means of rotated cylindrical lenses as well
• Rotation, equivalent to an even sequence of reflections,

As in case a) in 1 For a given segmentation, there is an arrangement of the transformed segments with maximum anisotropy, ie maximum SPP or M ² in a lateral direction (here x-direction) and minimal SPP or M ² in the complementary direction (here y-direction). , In the illustrated example of the segmentation in strips, such an arrangement is obtained by linear arrangement of the transformed segments on their narrow sides, that is to say the edges perpendicular to the cut edges.

Case b) according to 1 shows a non-minimal arrangement of the segments. Such an arrangement may be advantageous in several ways. For example, the angle α (see right-hand representation in case b)), which the transformed segments enclose with their common connection axis, can be used to set the effective SPPe in the direction of the connection axis and perpendicular to it.

In the left case b) arise against (x, y) tilted axes (X, Y). This intensity distribution can be z. B. homogenize again and z. B. use for mask imaging or z. B. transform into a narrow focus line. A variable setting of the SPP allows z. As in the generation of a narrow line in a simple manner, on the one hand, variations in the divergence of the radiation source Compensate, as well as to selectively adjust the width or depth of focus of the line.

A another possibility is to use the transformed ones Segments or the corresponding sub-beams, so to speak separately apply. So you could each of the sub-rays own a optical system or subsystem. For example exist surface treatment systems in which to increase the output, the radiation of a laser evenly on several equally constructed and juxtaposed projection systems is split. Instead of a conventional division could one change the SPP at the same time?

For Systems with mask imaging, it is also conceivable the design the mask and the arrangement of the transformed partial beams on each other vote. It may be necessary to realize the entire procedure or expedient, the Strahiquerschnitt ago or after transformation laterally in one or two directions to stretch or to compress. This changes the topology not the beam and its segments and therefore not the SPP.

2 schematically shows a possible construction of a system for producing a long narrow focus line 3 For example, for the crystallization of thin silicon layers on glass by means of the so-called annealing process. As a radiation source is z. B. a Nd: YAG laser conceivable. Essential components of such a system are at least one component except the laser (not shown) 5 for redistribution of the SPP, at least one component for beam homogenization 10 and at least one component for focusing 6 of the beam in a line. Typically, there is a homogenizer 10 from one or two lens arrays and a Fourier lens. For homogenization in only one lateral direction, cylindrical lenses are used. In principle, however, other implementations are possible, for. Using mirrors or diffractive elements (DOEs). In the in 2 shown system, the laser beam is first in the long axis (LA, see the upper illustration)) with a cylindrical lens telescope 2 expanded in order to be separated in the long axis LA in individual beams. An arrangement suitable for beam separation 7 from cylindrical lens arrays shows 3 , which will be discussed below. In the optical component following the beam separation, the spatial coherence of the partial beams relative to one another is reduced or eliminated. This can be z. B. implement by introducing different optical path lengths for the partial beams, such as with the aid of a step plate, so different thickness plane-parallel plates in transmission, or according to a step mirror arrangement. A decorrelator 8th is to bring particularly favorable at the described position in the beam path, since there the sub-beams separated and run collimated, or if necessary. The SPP transformation 5 hereafter or, if applicable, as its integral part. In the component called SPP transformation 5 the redistribution of the SPP takes place. To achieve the narrowest possible line in the specific case, the SPP in the short axis (SA) must be sufficiently reduced, while in the long axis (LA), an enlargement of the SPP is advantageous. The following component 4 mediates a smoothing of the intensity profile in the long axis to prepare for LA homogenization 10 , Subsequently, the beam is in the short axis by means of a cylindrical lens telescope 9 (sketched with three lenses) expanded to an appropriate height. In essence, this serves to adjust the numerical aperture (NA) of the final focusing, and thus the adjustment of the line width. If the telescope generates an intermediate focus (not shown), a slit diaphragm can additionally be set to this position with which an additional shaping of the line profile can be achieved. The telescope can also be elsewhere in the optical chain, z. B. just before the final focus.

On component 9 following the beam is homogenized in the long axis 10 and simultaneously expanded to the desired line length. A lens 6 , in 2 represented by two lenses, finally focuses the beam to a line.

3 illustrates an arrangement for beam separation consisting of two cylindrical lens arrays A1, A2, in which each realized a successive pair of lenses a miniaturizing telescope. The cross section of the beam impinging on the first array A1 (in FIG 3 by a rectangular cross-section 11 shown) is divided by this into strips and these by the reduction of the horizontally acting telescopes into separate strips 12 separated. When using two arrays A1, A2 with N lenses each, separate partial beams emerge from the second array A2.

An alternative arrangement for beam separation by means of laterally juxtaposed prisms P1, P2, P3 shows 4 , The entrance surface EF of each prism is perpendicular to the longitudinal axis of the prism, so that axially parallel incident light L enters without refraction and passes through the respective prism. Only at the exit surface AF tilted in relation to the entry surface EF is the jet L broken and deflected. The projection of the beam cross-sectional area in the plane defined by the incidence and the exit direction of the beam has a width D at the entrance. The projection of the outgoing beam, on the other hand, has a width D '<D, where D' becomes smaller with increasing deflection angle, so the jet is compressed in the associated dimension. Depending on the mutual offset of the exit surfaces AF with respect to the longitudinal axis of the prisms P1, P2, P3, a separation into partial beams T1, T2, T3 can be achieved with arbitrary lateral distances, as in FIG 4 shown (Compare variant a with b or c).

Instead of of prisms can also be appropriate arrangements of the direction of incidence mutually offset mirrors or DOEs (Diffractive Optical Elements) use.

An embodiment for redistributing the SPP of a laser beam is shown in FIG 5 shown. Here, assume that a beam separated into three parallel sub-beams T1, T2, T3 is deflected by three stacked prisms P1, P2, P3 in three different directions in (x, z), and returned to in by means of two prisms P4, P5 three output beams T1 ', T2', T3 'is parallelized. The middle sub-beam T2 is not deflected in the example shown. Therefore, the corrective prism between P4, and P5 can be omitted. Correspondingly, P2, of course, as a plane-parallel plate could also be omitted.

ever according to desired arrangements of the intensity distributions at the exit and the required beam deflection directions the active surfaces of the prisms are suitable for each other to orient.

shown are two variants: In variant A (middle representation) are the beam cross sections T1, T2, T3 only in the x-axis against each other shifted, so that a staircase-like pattern is generated.

at Variant B (lower illustration), the rays are at the same time in x and y direction deflected and the output cross sections at their strung together short edges. The device leaves of course analogous to any Number of sub-beams or shifted segments realize.

Furthermore, it is not necessary for the function that separate input beams are used, but z. B. a single beam can be used, which is separated by the input prisms in sub-beams. The advantage of pre-separation of the beam, however, is the avoidance of diffraction and scattering at the prism edges and surfaces. Such a pre-separation can be z. B. realize by a series of two cylindrical lens arrays A1, A2 or a prism array, as in 3 respectively. 4 shown.

Furthermore, the device of 5 also realize that the beam deflections is realized by mirrors instead of suitable by mirrors. Furthermore, it may be advantageous, instead of individual prism elements, the entire device monolithically through a block B with correspondingly shaped entrance surfaces EF and exit surfaces AF, as in 6 represented, or realized by DOEs. An advantageous feature is that, in addition, by choosing different thicknesses of the prisms in a simple manner can make the traversed by the partial beams optical paths of different lengths, and so reduce the reduction of the temporal coherence and the mutual spatial coherence and possibly. Eliminate can.

A variant of the device of 5 respectively. 6 arises, if you like in 5 , Variant A, the partial beams T1, T2, T3 initially deflects only in one plane and offset, this analogous to the solution of 6 but realized for each sub-beam T1, T2, T3 separately, each with its own monolithic block, as is apparent from 7 can be seen, showing an arrangement of three cuboid glass plates GP1, GP2, GP3 in plan view (a), side view (b) and front view (c), ie along the beam axis. The respectively opposite entry and exit surfaces EF, AF of each glass plate are oriented plane-parallel to one another, so that the refraction angles at EF and AF compensate each other and thus the entrance and exit beams are each on parallel lines. The lateral offset Δx ₁ , Δx ₂ , Δx _{3 of} the beams can be realized analogously as before by individual rotation of the plates in the plane of the beam offset or by different plate thicknesses or a combination of both measures.

• 1) Man kippt eine Anordnung gemäß dem in 20 dargestellten Ausführungsbeispiel um 45° um die z-Achse für einen Strahlverlauf. 20a zeigt eine perspektivische Draufsicht auf drei Platten P2, P3, P4, die von den Teilstrahlen T1, T2, T3, T4 durchsetzt werden, 20b stellt eine diesbezügliche Draufsicht und 20c eine Seitenansicht in Strahlebene der Teilstrahlen dar, wobei die eingetragenen Rechtecke jeweils die Strahlquerschnitte der einzelnen Teilstrahlen T1, T2, T3, T4 darstellen.
• 2) Die zweite Vorgehensweise sieht vor, einen gegenüber dem in 7 dargestellten Prismensatz (P2, P3, P3) um 90° um die z-Achse gedrehten zweiten Prismensatz (Q2, Q3, Q4) vorzusehen, der die Teilstrahlen in der (y, z)-Ebene versetzt. Ein derartiges Ausführungsbeispiel ist in 21 dargestellt. Gleichsam zeigt 21a eine perspektivische Draufsicht auf beide Prismensätze P2, P3, P4 und Q2, Q3, Q4, die von den Teilstrahlen T1, T2, T3, T4 durchsetzt werden, 21b und c stellen jeweils eine diesbezügliche Drauf- und Seitenansicht dar. Diese Anordnung hat den Vorteil, dass die Teilstrahlen unterschiedliche optische Wege zurücklegen, wodurch sich ihre wechselseitige Kohärenz der Teilstrahlen reduzieren bzw. eliminieren lässt. Nachteilig ist allerdings, dass sich bei einer eventuellen nachfolgenden Fokussierung des Bündels der Teilstrahlen für jeden Teilstrahl prinzipiell ein anderer Abstand des Fokus hinter einer fokussierenden Linse ergibt. Dieser Effekt ist vernachlässigbar, sofern die SPP-Transformation innerhalb des Rayleigh-Bereiches, also im Bereich der Strahltaille des Laserstrahles, stattfindet.

The described arrangement generates a staircase-like beam cross-section according to 7c , To generate a lateral arrangement of the cross-sectional areas according to variant B in 5 In principle, the following two approaches are conceivable:

• 1) One tilts an arrangement according to the in 20 illustrated embodiment by 45 ° about the z-axis for a beam path. 20a shows a perspective top view of three plates P2, P3, P4, which are penetrated by the partial beams T1, T2, T3, T4, 20b represents a plan view and 20c a side view in the beam plane of the partial beams, wherein the registered rectangles each represent the beam cross sections of the individual partial beams T1, T2, T3, T4.
• 2) The second procedure provides, one compared to the in 7 represented prism set (P2, P3, P3) by 90 ° about the z-axis rotated second prism set (Q2, Q3, Q4), which sets the partial beams in the (y, z) plane. Such an embodiment is in 21 shown. As it shows 21a a perspective top view of both prism sets P2, P3, P4 and Q2, Q3, Q4, which are penetrated by the partial beams T1, T2, T3, T4, 21b and c respectively show a respective top and side view thereof. This arrangement has the advantage that the partial beams cover different optical paths, whereby their mutual coherence of the partial beams can be reduced or eliminated. The disadvantage, however, is that in a possible subsequent focusing of the beam of sub-beams for each partial beam in principle results in a different distance of the focus behind a focusing lens. This effect is negligible as long as the SPP transformation takes place within the Rayleigh region, ie in the region of the beam waist of the laser beam.

A further alternative embodiment to the second approach explained above is in 22 shown, which does not have the above disadvantage, since in this case the optical paths of all partial beams are the same length. However, this also eliminates the decorrelating effect of 21 shown embodiment.

Show another possibility of realization 8th and 9 A cylindrical lens homogenizer H consisting of the lens arrays A1 and A2 and the condenser lens L (= Fourier lens) is supplemented by a prism stack P1, P2, P3. In its conventional function without prisms, the homogenizer forms the intensity distributions detected by the lenses of A1 in the rear focal plane BR of the Fourier lens L in the region 2 ' and bring them there to the overlay. Inhomogeneities of the input intensity are compensated in this way. This corresponds to the illustration in 6 , Middle. By a prism with suitably selected wedge angle in the beam path behind or preferably in front of the condenser lens L, as in 8th shown above and below, the beam path can be redirected and laterally shift the intensity profile in the image plane. Stacking several prisms with different wedge angles, as in 9 shown, one above the other, so with each prism of his circumcised part of the beam path can be deflected by the associated angle and finally achieve a lateral distribution of the intensity distribution in the image plane. Again, one can imagine using mirrors or DOEs instead of prisms, or at least monolithically executing portions of the device (eg, from the second array to the condenser lens).

In 10 A further exemplary embodiment of an SPP transformation is shown in which it is exploited that an input beam E, which enters a telescope in parallel but offset laterally from the optical axis, is offset by a proportional distance in the same or the opposite direction from it exits again. The direction of the offset depends on the type of telescope. For example, a bowtie telescope imparts an opposing offset, a Galilean telescope an equidistant offset. Instead of the beam, the telescope can also be displaced while the beam is held. In the lower part of 10 this is illustrated by the example of a telescope made up of two cylindrical lenses L, L: A steel entering the center of the telescope is not displaced, see case a). If the lenses are displaced laterally by a distance dx from the symmetry position, see case b), then the beam is displaced laterally proportional to the output. If one stacks according to N such telescopes of cylindrical lenses with a mutual offset dx perpendicular to the stacking direction on top of each other and sets the stack as before in the beam, so that the (sufficiently high) beam hits the sub-telescopes by the corresponding offset dx off-axis, the beam is lateral split into N partial beams. This situation is in the upper part of 10 for N = 3 and a uniform offset outlined.

If one wants to reduce the offset of the partial beams in the y-direction or even arrange the partial beams in the same height y, correspondingly 1 , Case a), this can be achieved, for example, by providing the plan sides of the lenses L and L 'with suitable prism angles, so that L, together with the deflection and focusing in x, also results in a deflection in corresponding selected prismatic front surface of the lenses L 'is recovered.

Another option is to have one in 10 described device M1, see 11 a second, similar device M2 nachzuschalten, but which is operated in reverse order, ie, the input and output are reversed, and was additionally rotated by 90 ° about the z-axis. In 11 this is exemplified for the case N = 3 partial beams.

Incidentally, the method can be used in the same way for. B. on a device with prisms or mirrors, etc. after 1 apply, ie whenever a device by purely lateral displacement of the beam cross sections results in a diagonal offset of the cross sections relative to each other.

12 shows an arrangement for rearranging the SPP by means of lateral offset by reflections on the example of four partial beams T0, T1, T2, T3 and three mirror pairs Mn, Mn ', n = 1, 2, 3. In the illustrated arrangement, the partial beams previously from the common beam to be separated. In the upper two partial images, the mirrors are represented by squares. The orientation of the part rays T0, T1, T2, T3 is indicated at the top left by rectangular beam cross sections on the mirrors. It can be seen that only the arrangement of the partial beams changes, but not the beam orientation. Such a mirror arrangement can be realized with individual mirrors or preferably with prisms. It may be advantageous to produce the overall arrangement monolithically, or as a micro-optic. Sub-beams T0, T1, T2, T3 travel different distances. This can be advantageous with regard to the reduction or elimination of interferences by mutual coherence of the partial beams: If the differences in distance are greater than the temporal coherence length, the partial beams are uncorrelated.

13 describes a further arrangement for rearrangement of the SPP by means of lateral offset, similar to the device in FIG 12 , As in this case, the orientation of the beam cross sections is retained, and only their relative arrangement changes. Here, the steel deflection takes place by 4 reflections. The four in 13 illustrated beam paths represent views of one and the same beam path from different directions, as shown 13 removable. The partial beams experience identical optical paths, and the mutual coherence of the partial beams is maintained. On the other hand, identical optical paths enable use in a non-collimated beam path. For example, you could put in front of the entrance and behind the exit apertures lenses with suitably chosen focal length, z. B. in an arrangement as miniaturizing telescope with intermediate focus. Positive input lenses that adjoin one another laterally can be used for beam segmentation and separation when focal lengths and distance to the mirror array are chosen so that the focused beams no longer touch the aperture edges of the subsequent elements.

Also Here, the device can be advantageous by Realize prisms, if necessary monolithic.

14 shows a variation of in 13 shown device. In contrast to there, the optical paths are different in length, so that one can not superimpose a uniform optical image with convergent or divergent partial beams. However, the different path lengths can advantageously be used to reduce or eliminate the mutual coherence of the partial beams.

15 describes a further realization possibility for a device with reflective surfaces. Shown is a concrete realization with prisms. Each prism can be understood as composed of two 45 ° angle prisms (roof halves of a cube), which are rotated at their common connecting surface by 90 ° relative to each other. The illustrated prism represents a kind of minimal realization (see three-page representation A). Drawing B shows a 3D view of a single prism. In C, a three-way representation of a combination of three prisms for SPP transformation is shown. In principle, other angles than 45 ° for deflection and 90 ° for mutual rotation could be used. However, the beam segments will then i. A. no longer rotated by 90 ° to each other. The arrangement is obviously not limited to the use of the illustrated minimal prisms. It can, for example, just as well stretched prisms, a monolithic design or z. B. also use individual mirror to realize the specified beam deflections.

With the arrangement shown, all sub-beams experience the same optical paths, and they already follow 13 specified properties (embedding in a non-collimated beam path, etc.). By a simple rearrangement of the prisms can be analogous to the transition from 13 to 14 realize an arrangement in which the optical paths of the individual sub-beams have different lengths. From that follow already 14 specified properties (coherence reduction, etc.).

At the in 16 sketched arrangement, the rearrangement of the beam or the SPP by means of two cylindrical lens arrays A1, A2. Two arrays of N cylindrical lenses each are arranged one behind the other so that two consecutive pairs of lenses form a telescope, ie that the vertices of the lenses are parallel and that the spacing of the arrays is adjusted accordingly. In practice, it is advisable to make the telescopes smaller. This makes it possible to ensure that the sub-beams can pass through the respective subsequent lenses of the second array without trimming. This is particularly important in the event that no separate sub-beams strike the first array A1, but the input beam is also segmented by the first array A1 at the same time and separated in the sequence, as shown in FIG 16 is shown. When passing through the arrangement, the cross sections of the partial beams are mirrored, so to speak, on the lens axes. A tilting of the lenses by an angle α out of the (y, z) plane results in a rotation of the cross sections or their orientation by 2α. In the case shown, the partial cross sections are rotated by 90 ° with α = 45 °. With this rotation, a maximum redistribution of the SPP for this arrangement can be achieved. To illustrate the topological connections are in 17 the sub-beam cross-sections drawn rectangular and the upper left corner of each rectangle provided with a marker, which in the transformation into the lower right corner of the transformed Einzelstrah walks. It can be seen that the beam cross sections, strictly speaking, do not undergo any rotation, but a reflection at the projections of the lens vertices in the (x, y) plane). Moreover, in 17 provided that the partial beams have already undergone separation before entering the cylindrical lens arrangement.

The general case α ≠ 45 ° is in 18 shown. By varying the angle α, the distribution of the SPP, ie the anisotropy, can be adjusted in a targeted manner. This was discussed in detail at the beginning of the general considerations.

If no separated partial beams strike the first array A1, but the total beam which is then segmented by this array at the same time, diffraction occurs due to the trimming lens edges. This effect may be critical with regard to the case where very small values are to be obtained for the beam product in one axis, since then the lens width B is only a small factor larger than the lateral coherence length of the light. However, diffraction is ultimately noticeable by a deformation (oscillations) of the desired intensity profile in the image plane. To reduce the effect, for example, in 19 sketched implementation variants for the cylindrical lens arrays conceivable. A) shows a conventional realization of separate cylindrical lenses. The arrays A1, A2 in case B) are monolithic without the abutting side surfaces of the lenses present in case A). With regard to a further improvement, it is conceivable to remove or smooth the crease-like transition between two adjacent lenses that is still present in B). This is shown for array A1 in C) with a change of convex cylinder segments with radius R1 and concave cylinder segments with radius R2.

Instead of cylinder segments and other curve functions may be useful, eg. B. a sinusoidal shape, or a composite parabolic course. The period of the surface function (analogous to the width of a single lens) is in 19C shown as constant. In fact, varying periods as well as varying radii R1, R2 may be beneficial. Case D) shows the case that both arrays A1 and A2 have a kink-free surface function.

QUOTES INCLUDE IN THE DESCRIPTION

This list The documents listed by the applicant have been automated generated and is solely for better information recorded by the reader. The list is not part of the German Patent or utility model application. The DPMA takes over no liability for any errors or omissions.

Cited patent literature

• US 2005/0035103 A1 [0008]
• - EP 1006382 B1 [0014]
• DE 102004034253 A1 [0014]
• - EP 1075719 B1 [0014]
• - DE 19819333 A1 [0014]
• US 2007/0024979 A1 [0014]

Claims (32)

Method for producing a laser beam with a prescribable beam parameter product, abbreviated SPP ₂ , in which the laser beam is emitted by a laser beam source with a first SPP ₁ and converted via at least one optical unit into a laser beam with the predeterminable SPP ₂ , where: 1. SPP ₁ ≤ SPP ₂ , 2. SPP _1x ≠ SPP _2x , 3. SPP _1y ≠ SPP _2y and SPP _1x and SPP _1y as well as SPP _2x and SPP _2y are mutually orthogonal components respectively. Method according to claim 1, characterized in that that the laser beam emitted by the laser beam source has a Beam cross section which is segmented into at least two segments. Method according to claim 2, characterized in that that the at least two segments are rectangular or, are diamond-shaped or disjoint faces of a Represent ellipse with straight edges. Method according to claim 2 or 3, characterized in that the at least two segments correspond to the beam cross sections of two obtained from the laser beam emitted from the laser beam source Partial rays correspond, which spatially separated from each other become. Method according to one of claims 2 to 4, characterized in that the at least two segments of a the spatial relative position, shape and / or size subject to the segment-related transformation. Method according to claim 5, characterized in that that transformation as part of a mirroring, rotation and / or lateral displacement is performed. Method according to claim 5 or 6, characterized that the at least two transformed segments to the single deformed laser beam are joined together. Method according to one of claims 5 to 7, characterized in that the transformation by means of optical Functional elements of the following type is performed: optical lenses, diffractive optical elements, curved or plane mirrors or prisms. Method according to one of Claims 1 to 8, characterized in that the laser beam emitted by the laser beam source and / or the laser beam having the prescribable beam parameter product SPP _{2 is} subjected to optical stretching or compression in one or two orthogonal directions with respect to the respective beam cross section. Method according to one of claims 1 to 9 for producing a laser beam with an elongated Beam cross section, characterized, that on the part the laser source emitted laser beam in at least two sub-beams, each having an elongated beam cross-section with a Longitudinal extension and two the oblong beam cross-section on both sides have limiting side edge regions, is segmented, and that the at least two partial beams along each one of them Side flank areas forming a uniform, elongated Laser beam can be put together or that at least two partial beams forming a staircase-shaped in the beam cross-section Laser beam are joined, in which the beam cross sections the partial beams with respect to their respective longitudinal extents are directed in parallel, or that the at least two partial beams be arranged side by side so that each one of the individual Longitudinal beam cross-sections assignable cross-sectional center of gravity lie on a common connection axis, opposite the oblong beam cross sections along their Longitudinal extension by a uniform angle α, with 0 ° less than or equal to α less than or equal to 90 °, are inclined. Apparatus for generating a laser beam with a prescribable beam parameter product, abbreviated SPP ₂ , with a laser beam source from which a laser beam emerges with a first SPP ₁ , which passes through at least one optical unit mediumly or directly forming the laser beam with the prescribable SPP ₂ , where : 1. SPP ₁ ≤ SPP ₂ , 2. SPP _1x ≠ SPP _2x , 3. SPP _1y ≠ SPP _2y and SPP _1x and SPP _1y as well as SPP _2x and SPP _2y are mutually orthogonal components respectively. Device according to claim 11, characterized in that that in the beam path between the laser beam source and the at least an optical unit is arranged an optical separation unit is the laser beam emitted from the laser beam source in N divides parallel partial beams. Apparatus according to claim 12, characterized in that the optical separation unit is designed in the form of one of the following optical arrangements: - two Zy spaced along the beam path Linderlinsenarrays, - at least one array of prisms, - at least one array of mirrors, - at least one DOE or at least one array of DOEs. Device according to one of claims 11 to 13, characterized in that the at least one optical unit a first set of at least N - 1 in the beam path adjacent prisms (P ₁ , P ₂ , ..., P _N-1 ) with different wedge angles wherein each prism transmits part of the laser beam by changing the spatial relative positions of the individual parts of the laser beam, and that a second set of at least N-1 prisms (P ₁ ', P ₂ ', ..., P _N-1 ' ) is provided, which merge the individual parts of the laser beam into a laser beam with the predetermined SPP ₂ . Device according to claim 14, characterized in that that the first and second set of prisms in a monolithic, for the parts of the laser beam transmitting unit are integrated. Device according to claim 14 or 15, characterized that the prisms of at least one sentence differ about Thickness and thus over different length trained have optical paths. Device according to one of claims 11 to 13, characterized in that the at least one optical unit provides a first set of at least N - 1 mirrors (S ₁ , S ₂ , ..., S _N-1 , ...), wherein each mirror reflects a portion of the laser beam changing the spatial relative positions of the individual portions of the laser beam, and that at least a second set of at least N - 1 mirrors (S ₁ ', S ₂ ', ..., S _N-1 ',. ..) is provided, which merges the individual parts of the laser beam into a laser beam with the predetermined SPP ₂ . Device according to claim 17, characterized in that that at least one of the two sets of mirrors is monolithic is executed. Device according to claim 17 or 18, characterized that the respective optical paths on the associated Mirrors are different lengths. Device according to one of claims 11 to 13, characterized in that the beam path of the laser beam of the laser beam source following a cylindrical lens homogenizer is provided which has one or two cylindrical lens arrays and these in the beam path downstream of a Fourier lens, and that the at least one optical unit is formed As a set of at least N-1 prisms (P ₁ , P ₂ ,..., P _N-1 ,...) Arranged side by side in the beam path with different wedge angles or, as a set of at least N, arranged adjacent to one another in the beam path. 1 mirror with different deflection angles or - as a DOE or a set of in the beam path juxtaposed DOEs, and which is arranged in the beam path two cylinder lens arrays downstream and before or after the Fourier lens. Device according to one of claims 11 to 13, characterized in that the at least one optical Unit a first telescope arrangement consisting of two sets of cylindrical lenses, of which each set is at least N - 1 orthogonal to the telescopic beam path side by side arranged cylindrical lenses with mutually parallel cylindrical lens axes provides, wherein the cylinder axis of each cylinder lens per set at least the circumscribing cylinder of the immediately laterally adjacent Cylindrical lens penetrates and two immediately adjacent Cylindrical lenses a mutual offset δx orthogonal have to the cylinder axes. Apparatus according to claim 21, characterized that in the beam path of the first telescope arrangement below is provided for the first identical second telescopic arrangement, and that the second telescopic arrangement in the reverse transmission direction with respect to the first telescope arrangement in the beam path and with respect to the first telescope arrangement by 90 ° the telescopic axis is inserted twisted. Device according to one of claims 11 until 13, characterized in that the at least one optical Unit two in the beam path arranged one behind the other cylindrical lens arrays each having N cylindrical lenses and in the manner of a Telescope arrangement are arranged, that all Cylindrical lenses parallel to each other cylindrical lens axes to provide, and that the cylinder axes are opposite to an angle α one lying in a plane orthogonal to the beam path vertical reference axis are inclined. Device according to claim 23, characterized in that each pair faces each other in both arrays and associated lenses each have a telescope with a uniform Make magnification M less than or equal to 1. Device according to claim 23, characterized in that that the angle α = 45 °. Device according to one of claims 11 to 13, characterized in that the at least one optical unit at least a first set of at least N - 1 cuboid flat plates (P ₁ , P ₂ , ..., P _N-1 , ...) provides that transmits at least a portion of the laser beam by changing the spatial relative positions of the individual partial beams of the laser beam that the individual plates have exactly coplanar to each other oriented entrance and exit surfaces, that the plates have coplanar mutually oriented side surfaces that the plates on these side surfaces together are aligned, and that at least one plate is rotated about an axis orthogonal to its side surfaces, so that the incoming beam does not meet at least this plate is not perpendicular to the entrance surface. Device according to claim 26, characterized, that with at least N - 1 of the N plates the distances between the respective inlet and outlet surfaces the same are big, and that every two of the at least N - 1 Plates around an orthogonal axis to their side surfaces are rotated relative to each other, so that the levels of their inputs or Exit surfaces per each other a predetermined angle greater than zero. Device according to claim 26, characterized, that two of the at least N - 1 plates have a different one Distance between inlet and outlet surface have, and that coplanar the entrance and exit surfaces of two plates lie to each other. Device according to claim 28, characterized, that a set of at least N - 1 plates of two similar ones Sets of about N / 2 plates, and that these two sets of approximately N / 2 plates relative to each other about an axis orthogonal to their side surfaces are turned. Device according to one of claims 26 until 29, characterized, that the cross sections the separated partial beams entering the first set of plates or the beam cross section in its longitudinal extent of through the plates first to be segmented and separated beam lie in a plane which is orthogonal to the side surfaces the plates is standing, and that the offset of the spatial Relative positions of the partial beams in an implementation of the linear / stretched Distribution of the beam segments of the incoming beam in a staircase-shaped Distribution of parallel partial beams at the output consists. Device according to one of claims 27 to 30, characterized in that at least a second set of minimally N - 1 plates (P ₁ ', P ₂ ', ..., P _N-1 , ...) is provided, which in a reverse operation is operated in comparison to the first set and merges the individual parts of the laser beam into a laser beam with the predetermined SPP ₂ . Device according to one of claims 26 until 29, characterized in that the cross sections of in the first plate set entering separated partial beams or the beam cross section in its longitudinal extent of the plates first to be segmented and separated beam in a plane lying with the side surfaces of the plates includes an angle of 45 °, and that the offset of the spatial relative positions of the partial beams in an implementation of the linear distribution of the beam segments the incoming beam to a line-like distribution to each other parallel partial beams at the output, the axes of these Distributions are orthogonal to each other.