DE102008023763A1 - Illumination system for use in microlithographic-projection illumination system during production of semiconductor component, has Fourier optics system including ratio of overall length to bandwidth less than specific value - Google Patents

Abstract

The system (190) has a pupil formation unit (150) receiving light of a primary light source (102) and producing variably adjustable 2-dimensional intensity distribution in a pupil formation surface (110). The unit includes a Fourier optics system converting radiation bundle entering through an entry plane into radiation bundle exiting from an exit plane of the optics system. The optics system includes a ratio of overall length to bandwidth less than 1/6. The length is measured between an entry side system surface and an exit side last system surface along an optical axis (103). An independent claim is also included for a microlithographic-projection illumination system comprising a projection lens.

Description

BACKGROUND OF THE INVENTION

Field of the invention

The The invention relates to a lighting system for a Microlithography projection exposure apparatus for illuminating a Illumination field with the light of a primary light source as well as on a microlithography projection exposure machine with such a lighting system. Furthermore, the invention relates on a Fourieroptiksystem, which z. B. as part of a lighting system a microlithography projection exposure system can.

Description of the Related State of technology

to Production of semiconductor devices and other fine-structured Today, components are predominantly microlithographic Projection exposure method used. This will be masks (Reticle) uses the pattern of a structure to be imaged wear, z. B. a line pattern of a layer of a semiconductor device. A mask is placed in a projection exposure machine between lighting system and projection lens in the area of the object surface of the Projection lens positioned and with one of the illumination system illuminated illumination radiation illuminated. The by the Mask and the pattern changed radiation is running as projection radiation through the projection lens, which the pattern of the mask is imaged on the substrate to be exposed, the usually a radiation-sensitive layer (photoresist, photoresist) wearing.

at The projection microlithography is the mask with the help of a Illuminated lighting system that emanates from the light of a primary Light source, in particular a laser, directed to the mask Illuminating radiation is formed by certain lighting parameters is defined. The illumination radiation hits within one Illumination field (area of defined shape and size, z. Rectangle field or curved ring field) on the mask on, taking shape and size of the illumination field usually constant, d. H. not variable. Usually will be within of the illumination field as uniform as possible Intensity distribution sought, including within the lighting system Homogenizing devices, such as light mixing elements as honeycomb condensers and / or rod integrators, be provided can.

Furthermore become common depending on the type of structures to be imaged different lighting modes (so-called "lighting settings") needed by different local Intensity distributions of the illumination radiation in one Pupil surface of the illumination system characterized can be. Sometimes one speaks in this context from "structured lighting" or from a "structuring the illumination pupil "or a structuring the secondary light source. The pupil surface of the Lighting system in which certain definable two-dimensional Intensity distributions (the secondary light sources) are also referred to in this application as "pupil shaping surface", because essential properties of the illumination radiation with the help this intensity distribution "shaped". The lighting settings include, for example the conventional lighting settings round to the optical Axis of the illumination system centered illumination spots different Diameter (usually defined by the degree of coherence σ of the Lighting) and in non-conventional, d. H. off-axis Types of illumination the ring illumination (or annulare illumination) as well as polar intensity distributions, for example dipole illumination or quadrupole illumination. The non-conventional lighting settings for generating an off-axis (oblique) illumination Among other things, the increase in depth of field serve by two-beam interference and increasing the resolution.

The "pupil shaping surface" of the illumination system in which the desired two-dimensional intensity distribution (secondary light source) is to be located may be at or near a position optically conjugate to a pupil plane of a subsequent projection lens in a lighting system incorporated in a microlithography projection exposure apparatus. In general, the pupil shaping surface may correspond to or lie in the vicinity of a pupil surface of the illumination system. If the intermediate optical components do not change the beam angle distribution, ie operate at angle preserving, the angular distribution of the illumination radiation striking the pattern of the mask is determined by the spatial intensity distribution in the pupil shaping surface of the illumination system. In addition, as far as the intermediate optical components operate to maintain the angular, the spatial Intensi distribution in the pupil of the projection lens is determined by the spatial intensity distribution (spatial distribution) in the pupil shaping surface of the illumination system.

Those optical components and assemblies of the lighting system, the are intended to light a primary light source, z. As a laser or a mercury vapor lamp, and from this a desired two-dimensional intensity distribution (secondary light source) in the "pupil forming surface" of the To produce illumination system, together form a pupil shaping unit, which should be adjustable in the rule.

From the US 2007/0165202 A1 (corresponding WO 2005/026843 A2 The applicant discloses illumination systems in which a pupil-shaping unit for receiving light from a primary light source and for generating a variably adjustable two-dimensional intensity distribution in a pupil-shaping surface of the illumination system comprises a multi-mirror array (MMA) with individually controllable individual mirrors which Angle distribution of the radiation falling on the mirror elements can selectively change so that results in the pupil shaping surface, the desired illumination intensity distribution.

Methods for calculating optimal structurings of the intensity distribution in the pupil shaping surface of an illumination system as a function of mask structures to be imaged are, for example, made US 6,563,556 or US 2004/0265707 known.

SUMMARY OF THE INVENTION

It It is an object of the invention to provide a lighting system for to provide a microlithography projection exposure apparatus, which allows a quick change between different lighting modes.

It Another object of the invention is to be integrated into one Illumination system for a microlithography projection exposure apparatus to provide a suitable, compact light mixing system, which is capable of a light mixture with a small geometric light conductance essentially without introduction of geometric light conductance to effect.

to Solving these and other objects is the invention an illumination system with the features of claim 1 and a Fourier optical system having the features of claim 24 ready. Continue to be a light mixing system with the features of claim 26 and a A projection exposure apparatus provided with the features of claim 28.

advantageous Further developments are in the dependent claims specified. The wording of all claims becomes by reference to the content of the description.

According to one aspect of the invention, a lighting system for a microlithography projection exposure apparatus is provided for illuminating a lighting field with the light of a primary light source. The illumination system has a variably adjustable pupil-shaping unit for receiving light from the primary light source and generating a variably adjustable two-dimensional intensity distribution in a pupil-shaping surface of the illumination system. The pupil shaping unit has a Fourier optical system for converting an entrance beam entering through an entrance plane of the Fourier optical system into an exit beam emanating from an exit plane of the Fourier optical system, the Fourier optical system having a focal length f _FOS and a first system surface between an entrance side and an exit side last system along an optical axis measured length has L and the condition (L / f _FOS ) <1/6 applies.

The term "Fourier optical system" here stands for an optical system which transforms a radiation power distribution present in the entrance plane of the Fourier optical system into the exit plane while maintaining the optical conductivity (etendue, geometrical flux) of the transmitted radiation. The exit plane is a Fourier-transformed plane to the entry plane. A passing beam defines an entrance surface of a specific shape and size, for example a circular entry surface or a square or otherwise rectangular entry surface, in the entry plane. In the Fourier-transformed exit plane, the beam defines an exit surface whose shape and size is determined by the angular distribution of the radiation in the entry plane. The geometry of the entrance surface is defined by the beam heights of the passing beams. The geometry of the entrance surface is determined by the Fourier optical system in a corresponding angular distribution (distribution of beam angles) in the exit plane transferred. The "entrance surface" and the "exit surface" are defined here as the cut surfaces of a penetrating beam with the entrance plane or the exit plane and thus each have a certain surface area. In the case of the Fourier transformation taking place between the entrance level and the exit plane, the power distribution of each individual area element in the entrance area is distributed over the entire exit area according to the local divergence on the exit area. All surface elements received on the outlet side are thereby superimposed additively in at least one dimension.

A beam within a real optical system contains a plurality of beams of different propagation directions. The angular distribution of the beams of a beam can be described by the divergence DIV of the beam, which describes the largest angular difference between beams within the beam. Alternatively, the description by the numerical aperture NA of the beam is possible, which corresponds to the sine of half divergence angle in this application. In paraxial optics, ie at small beam angles relative to the optical axis of an optical system, the numerical aperture NA thus corresponds to half the divergence, ie NA = DIV / 2. The effect of a Fourier optical system on a penetrating beam with a given input divergence (divergence on the entrance side) can be described in simplified terms, that each beam angle RA _{E of} a beam on the input side is assigned to the beam angle proportional beam height RH _A on the exit side. The beam height is defined here as the perpendicular distance of a beam at a given axial location to the optical axis. The proportionality between the beam angles on the entrance side and the beam heights on the exit side is given by the focal length f _{FOS of} the Fourier optical system according to RH _A = f _FOS * sin (RA _E ).

There a Fourier optics system accordingly beam angle on its entrance side in beam heights on its exit side as specified the focal length of the Fourieroptiksystems is a Fourieroptiksystem with a large focal length z. B. capable of an input beam with given small input divergence an exit beam with correspondingly larger cross-sectional area with a given focal length of the Fourier optical system the size ratio between the entrance surface and exit area depends on the input divergence and smaller, the larger the input divergence is.

In a lighting system, which with a laser as a primary Light source works, lies according to the spatial Coherence of laser radiation usually primary radiation with very little divergence in beams with relative small beam cross section before. On the other hand be stands in lighting systems the need, within the lighting system at least one area in which the radiation passing through a relatively large Beam cross section has. For example, in the field of large Beam cross-section uses a light modulation device to the angular distribution of within an impinging beam variably set existing radiation, so the spatial resolution the variable setting can be improved when the light modulation device sits in a range of relatively large beam diameter and a field with many individually controllable individual elements contains, each a sub-beam of the impinging Beam angle influence angle. ever greater the beam diameter at the location of the light modulation device is, the easier it is, a sufficiently large number to provide controllable individual elements of the light modulation device, to allow a high spatial resolution of the angle setting.

One Fourier optics system with relatively long focal length can do so be used, despite relatively low divergence of an incoming Beam a beam with a relatively large Beam cross section to produce. On the other hand, stands in a lighting system usually only limited space for optical subsystems a pupil forming unit available. By the Use of a Fourieroptiksystems invention the opposing demands for effective beam expansion an input beam with small divergence on the one hand and relatively low space requirements on the other hand agree.

Preferably, the condition (L / f _FOS ) <0.166 applies to the telefax TF = L / f _FOS . The Telefaktor can z. B. 0.125 or less or 0.1 or less or 0.075 or less.

In some embodiments, the focal length f _{FOS of} the Fourier optical system is more than 10 m (eg 15 m or more or 20 m or more or 50 m or more) and the length L is less than 4 m, z. B 3.5 m or less or 3 m or less.

The Fourier optics system causes an odd number of Fourier transforms and can z. B. 3 or 5 Fourier transform cause. Preferably finds only a single Fourier transformation between the entrance surface and exit surface instead, resulting in a short overall length is favored.

A Fourier optical system with a relatively short length compared to the focal length usually has at least 3 lenses. In some embodiments, the Fourier optical system has a first lens group having an entrance-side first lens and an exit-side second lens and a second lens group downstream of the first lens group having an entrance-side first lens and an exit-side second lens, between an exit-side last system surface of the first lens group and an entrance-side first system area of the second lens group is a group distance d _G. In this embodiment, at least 4 lenses are thus provided. It may be single lenses, one or more of the lenses may also be configured as a split lens or lens group. The group spacing is usually greater than the lengths of the first and the second lens group.

In some embodiments, the group distance d _{G is subject to} the condition d _G > 0.66 · L. The group spacing can thus make up a substantial proportion of the total construction length L. The condition d _G > 0.7 · L can also apply. The mutually facing lenses of the first and second lens group should thus have a relatively large distance, which z. B. in terms of the energetic load of these lenses is advantageous.

Compared to the focal length, the group spacing can be relatively small. In some embodiments, the condition d _G <0.1 × f _FOS applies. In particular, d _G <0.08 · f _FOS or d _G <0.06 · f _FOS .

design principles for one with regard to the radiation load of the lens elements Optimized construction of a Fourier optical system are related explained in detail with the embodiments.

If the Fourier optical system is designed to transmit a radiation energy E per unit of time at a light conductance H, P _{A is} a predeterminable maximum energetic loading of the exit-side second lens of the first lens group and P _{B is} a predeterminable maximum energetic loading of the entrance-side first lens of the second lens group, then some Embodiments provided that a group distance d _G between an exit-side last system surface of the first lens group and an entrance-side first system surface of the second lens group is not smaller than a minimum group distance d _G ^min , wherein for the minimum group distance applies: d G min = n / H · E / (P a P b ) 1.2

Becomes complied with this condition, it can be achieved by the Radiation exposure particularly at risk lenses not excessive be charged, so that a continuous operation without Linsendegradation is possible.

On the other hand, in order to keep the overall construction length L moderate, it may be provided that the group distance d _{G is} between d _G ^min and 3 · d _G ^min .

at In some embodiments, the pupil shaping unit a light mixing device upstream of the Fourier optical system on. This light mixing device is thus between the primary Light source and the Fourieroptiksystem arranged. When the light mixing device the incoming radiation mixes so that a substantially homogeneous Distribution in the angular space is present, this is the downstream Fourier optical system in a homogeneous light distribution in the spatial area converted in the area of the exit surface, ie in one largely uniform illumination of the exit surface. The light mixing device may comprise at least one honeycomb condenser. Its rear, the Fourieroptiksystem facing focal plane can essentially with the entrance surface of the Fourier optical system coincide or slightly opposite this surface be postponed. By combining a homogenizing in the angle space effective light mixing device with a downstream Fourieroptiksystem is it possible to input light with relatively low geometric Conductance, for example, the light of a laser beam, substantially without the introduction of optical conductivity to mix or homogenize.

at In one variant, the Fourier optical system has at least one pair of crossed cylindrical lens systems, with a pair of crossed Cylinder lens systems, a first cylindrical lens system with at least a curved in a first plane of curvature first cylindrical surface and a second cylindrical lens system with at least one in a second curvature surface has curved second cylindrical surface, wherein the first and the second plane of curvature perpendicular to each other stand. Under certain stress conditions, when using Crossed cylindrical lens systems the length of a load optimized Fourieroptiksystems be lower than when using rotationally symmetric lenses.

cylindrical lenses differently oriented curvature levels can nested, d. H. be arranged in alternating sequence. It is also possible, the differently oriented cylindrical lenses into "pure" subsystems. At a Variant, the Fourier optical system has a first cylindrical lens group with several first cylindrical lenses and a downstream second Cylindrical lens group with several second cylindrical lenses with orthogonal Orientation of the curvature plane.

The invention also relates to a Fourier optical system for converting an entrance beam entering through an entrance plane of the Fourier optical system into an exit beam emanating from an exit plane of the Fourier optical system, wherein the Fourier optical system has a focal length f _FOS and a final system surface measured along an entrance side last system surface along an optical axis Length L has and the condition (L / f _FOS ) <1/6 applies.

The Fourier optical system can be used in a lighting system of a projection exposure machine for microlithography as described or elsewhere Place to be used. Alternatively, it is also radiative in others Systems can be used, for. B. in a laser processing machine.

The The invention also relates to a light mixing system for receiving light a primary light source and for generating a substantially homogeneous two-dimensional intensity distribution in one Illuminating surface, wherein the light mixing system is a Fourier optical system of the type mentioned and the Fourieroptiksystem one in the angular space effective light mixing device is connected upstream. This can be a compact Lichtmischsystem be provided with moderate space requirements, which is capable of a light mixture in a small geometric Optical conductivity essentially without the introduction of geometric To cause the light conductance.

The The invention also relates to a microlithography projection exposure apparatus for exposing one in the area of a picture surface of a Projection lens arranged radiation-sensitive substrate with at least one image of one in the area of an object surface of the projection lens pattern of a mask with: a primary light source; a lighting system for reception the light of the primary light source and for the formation of on the pattern of the mask directed illumination radiation; and a projection lens for imaging the structure of the mask a photosensitive substrate, wherein the illumination system is at least a Fourier optical system of the type described in this application contains.

The Terms "radiation" and "light" in the The meaning of this application is to be interpreted broadly and is intended in particular electromagnetic radiation from the deep ultraviolet range include, for example at wavelengths of about 365 nm, 248 nm, 193 nm, 157 nm or 126 nm.

The The foregoing and other features are excluded from the claims also from the description and the drawings, wherein the individual features each alone or to several in the form of subcombinations in embodiments of the Invention and other fields be realized and advantageous as well as for protectable versions can represent.

BRIEF DESCRIPTION OF THE DRAWINGS

1 shows a schematic overview of a microlithography projection exposure apparatus with a pupil-forming unit;

2 schematically shows essential components of an embodiment of a pupil-forming unit, wherein 2A an overview is and 2 B . 2C schematically show a multi-mirror arrangement used in the pupil forming unit;

3 shows in 3A and 3B a light mixing system with a honeycomb condenser and a downstream Fourier optical system;

4 shows in 4A a schematic representation of a pupil forming unit with a honeycomb condenser and a downstream folded Fourieroptiksystem and in 4B a pupil forming unit having another light mixing device and a downstream folded Fourier optical system;

5 shows a meridional lens section through an embodiment of a Fourier optical system which can be used in the pupil shaping unit,

6 shows schematically the paraxial beam path through a Fourieroptiksystem;

7 shows schematically the representation of a paraxial beam path through an optical system and the associated Delano diagram;

8th shows a representation of a distance d between two points in the Delano diagram;

9 illustrates an energetic loading model for optical elements of an optical system;

10 shows a Delano diagram for a Fourier optical system with only one lens or refractive power b;

11 shows a Delano diagram for a Fourier optical system with three powers;

12 shows a Delano diagram of a Fourier optical system with four lenses or refractive powers;

13 Fig. 15 shows a semi-quantitative plot of areal radiation power density S for excellent areas of a Fourier optical system;

14 shows simplified Delano diagrams for Fourier optical systems with 4 individual lenses and different power sequences, namely pppp in 14A , pnpp in 14B and ppnp in 14C ;

15 shows a schematic perspective view of a Fourieroptiksystems with cylindrical lenses;

16 shows a schematic representation of a Fourieroptiksystems with associated input field and output field;

17 shows a simplified Delano diagram similar 12 to illustrate the load-optimized arrangement of lenses;

18 shows a Delano diagram of a rotationally symmetric Fourieroptiksystems with four lenses and lens sequence pnnp.

19 1 shows a Delano diagram of a first cylindrical lens group with a long output slice width, which in the direction of irradiation in front of a second cylindrical lens group (FIG. 20 ) is arranged;

20 shows a Delano diagram of a second cylindrical lens group, which in the direction of irradiation behind a first cylindrical lens group ( 19 ) is arranged.

DETAILED DESCRIPTION THE PREFERRED EMBODIMENTS

In 1 is an example of a microlithography projection exposure apparatus 100 which can be used in the manufacture of semiconductor devices and other finely-structured components and works to achieve resolutions down to fractions of a micron with light or electromagnetic ultraviolet radiation (DUV). As a primary light source 102 serves an ArF excimer laser with a working wavelength of about 193 nm, the linearly polarized laser beam coaxial with the optical axis 103 of the lighting system 190 is coupled into the lighting system. Other UV light sources, for example, F ₂ laser with 157 nm working wavelength, ArF excimer laser with 248 nm operating wavelength or mercury vapor lamps, z. B. with 368 nm or 436 nm operating wavelength, and primary light sources with wavelengths below 157 nm are also possible.

The polarized light of the light source 102 first enters a beam expander 104 a, for example, as a mirror arrangement according to the US 5,343,489 may be formed and used to reduce the coherence and increase the beam cross section.

The expanded laser beam has a certain cross-sectional area with an area, for example, in the range between 100 mm ² and 1,000 mm ² and a certain cross-sectional shape, for example, a square cross-sectional shape. The divergence of the expanded laser beam is usually smaller than the very small divergence of the laser beam before beam expansion. The divergence can z. B. between about 1 mrad and about 3 mrad lie.

The expanded laser beam enters a pupil-forming unit 150 including a plurality of optical components and groups and configured to be in a subsequent pupil shaping surface 110 of the lighting system 190 to generate a defined, local (two-dimensional) illumination intensity distribution, sometimes referred to as a secondary light source or as an "illumination pupil". The pupil shaping surface 110 is a pupil surface of the illumination system.

The pupil shaping unit 150 is variably adjustable, so that depending on the control of the pupil shaping unit different local illumination intensity distributions (ie differently structured secondary light sources) can be adjusted. In 1 For example, various illuminations of the circular illumination pupil are shown schematically by way of example, namely a centered circular illumination spot convention setting CON, a dipole illumination DIP or a quadrupole illumination QUAD.

In the immediate vicinity of the pupil shaping surface 110 is an optical raster element 109 arranged. A coupling optics arranged behind it 125 transfers the light to an intermediate field level 121 in which a reticle / masking system (REMA) 122 is arranged, which serves as an adjustable field stop.

The optical raster element 109 has a two-dimensional array of diffractive or refractive optical elements and has several functions. On the one hand, the incoming radiation is shaped by the raster element so that, after passing through the subsequent coupling optics 125 in the area of the field level 121 a rectangular illumination field illuminates. The grid element, also known as field-defining element (FDE) 109 with rectangular radiation characteristic generates the majority of the light conductance and adapts it to the desired field size and field shape in the mask plane 165 optically conjugate field level 121 , The grid element 109 can be designed as a prism array, in which arranged in a two-dimensional field prisms locally introduce certain angles to the field level 121 to illuminate as desired. The through the coupling optics 125 The generated Fourier transform causes each specific angle at the exit of the raster element to a location in the field plane 121 corresponds, while the location of the grid element, ie its position with respect to the optical axis 103 , the illumination angle in the field level 121 certainly. The outgoing of the individual raster elements beam bundles are superimposed in the field level 121 , It is also possible to design the field-defining element in the manner of a multi-stage honeycomb condenser with microcylinder lenses and lenses. By suitable design of the grid element 109 or its individual elements can be achieved that the rectangle field at the field level 121 is illuminated substantially homogeneously. The grid element 109 thus serves as Feldformungs- and homogenizing element and the homogenization of the field illumination, so that can be dispensed with a geson dertes light mixing element, for example acting on multiple internal reflection integrator rod or a honeycomb condenser. As a result, the optical structure in this area is particularly compact axially.

The following imaging lens 140 (also called REMA lens) forms the intermediate field plane 121 with the field stop 122 on the reticle 160 (Mask, lithographic master) from a scale, the z. B. between 2: 1 and 1: 5 and in the embodiment is about 1: 1. The image is taken without an intermediate image, so that between the intermediate field level 121 representing the object plane of the imaging lens 140 corresponds, and the image plane optically conjugate to this object plane 165 of the imaging lens, the exit plane of the illumination system and at the same time the object plane of a subsequent projection objective 170 corresponds to exactly one pupil surface 145 which is a Fourier transformed surface to the exit plane 165 of the lighting system. In other embodiments, at least one intermediate image is generated in the imaging lens. One between this pupil surface 145 and the image surface arranged at 45 ° to the optical axis 103 inclined deflecting mirror 146 makes it possible to install the relatively large lighting system (several meters in length) horizontally and the reticle 160 store horizontally. Between the intermediate field level 121 and the picture plane 165 of the imaging objective, radiation-influencing elements can be arranged, for example polarization-influencing elements for setting a defined polarization state of the illumination radiation.

Those optical components that the light of the laser 102 receive and form from the light illuminating radiation that is on the reticle 160 directed, belong to the lighting system 190 the projection exposure system. Behind the lighting system is a device 171 for holding and manipulating the reticle 160 arranged so that the pattern arranged on the reticle in the object plane 165 of the projection lens 170 is located and in this plane for scanner operation in a scan direction (y-direction) perpendicular to the optical axis 103 (Z direction) is movable by means of a scan drive.

Behind the reticle plane 165 follows the projection lens 170 , which acts as a reduction lens and an image of the mask 160 arranged pattern on a reduced scale, for example, on a scale of 1: 4 or 1: 5, on a wafer coated with a photoresist layer or photoresist layer 180 whose photosensitive surface is in the image plane 175 of the projection lens 170 lies. Refractive, catadioptric or catoptric projection lenses are possible. Other reduction measures, for example, greater reductions to 1:20 or 1: 200, are possible.

The substrate to be exposed, which in the example is a semiconductor wafer 180 is through a facility 181 held, which includes a scanner drive to synchronize the wafer with the reticle 160 to move perpendicular to the optical axis. Depending on the design of the projection lens 170 (For example, refractive catadioptric or catoptric, without intermediate image or with intermediate image, folded or unfolded) these movements can be parallel to each other or counterparallel. The device 181 which is also referred to as Waferstage, as well as the facility 171 , also referred to as "reticle days", are part of a scanner device which is controlled via a scan control device.

The pupil shaping surface 110 is at or near a position that is optically conjugate to the next succeeding pupil surface 145 as well as the image-side pupil surface 172 of the projection lens 170 is. Thus, the spatial (local) light distribution in the pupil 172 of the projection lens by the spatial light distribution (spatial distribution) in the pupil shaping surface 110 of the lighting system. Between the pupil surfaces 110 . 145 . 172 In each case there are field surfaces in the optical beam path which are Fourier-transformed surfaces to the respective pupil surfaces. This means in particular that a defined spatial distribution of illumination intensity in the pupil shaping surface 110 a certain angular distribution of the illumination radiation in the area of the subsequent field surface 121 which in turn gives a certain angular distribution to the reticle 160 falling illumination radiation corresponds.

In 2 are schematically essential components of an embodiment of a pupil shaping unit 150 shown. The entering, expanded laser beam 105 is through a plane deflecting mirror 151 towards a honeycomb condenser (fly eyes lens) 152 deflecting the incoming radiation beam into sub-illumination beams which are subsequently separated by a Fourier optical system 500 on a lens array 155 ie transferred to a two-dimensional array of lens systems. The lens array 155 concentrates the part-illumination beam 156 on individually controllable mirror elements of a multi-mirror arrangement 300 (multi-mirror-array, MMA), which is related to 2 B and 2C will be explained in more detail. The multiple mirror arrangement is here operated as a reflective light modulation device for the controllable change of the angular distribution of the radiation beam incident on the light modulation device and ensures by the alignment of its individual mirrors 302 for a definable by means of the multi-mirror arrangement illumination angle distribution, which is in the Pupillenformungsfläche 110 superimposed on an intensity distribution in this pupil surface. The individual mirrors 302 the multi-mirror arrangement, which on a common carrier element 301 are attached are about one or more axes for varying the propagation angle of the incident partial illumination beam 156 tiltable. The of the individual mirrors 302 outgoing partial illumination beam are through a lens 157 guided and by means of a subsequent condenser optics 158 into the pupil shaping surface 110 displayed. The lens array 155 and / or the micromirror arrangement 300 can essentially be constructed as it is in the US 2007/0165202 A1 the applicant is described. The related disclosure of this patent application is incorporated herein by reference. Also transmissive light modulation devices are possible.

The 3A . 3B show schematically some assemblies of the pupil forming unit 150 extending between the optional deflecting mirror 151 and the multi-mirror assembly 300 are located. On the representation of the optional lens array 155 was waived. 3A shows as a single, through a channel of the honeycomb condenser 152 guided partial illumination beam using the honeycomb condenser and the subsequent Fourieroptiksystems 500 on the multiple mirror arrangement 300 is shown. The Fourier optical system is operated here as a condenser and arranged so that the second honeycomb channel plate 152A the honeycomb condenser in the entrance (front) focal plane and the multi-mirror arrangement 300 in the exit (rear) focal plane of the Fourier optical system 500 located. For clarity, the beam paths of selected beams of the sub-illumination beam are shown in the form of solid and dashed lines, the optical axis 103 is dot-dashed. The solid lines represent those rays of the sub-illumination beam that make up the corresponding honeycomb channel of the honeycomb condenser 152 and hit the highest possible angle. On the other hand, the dashed lines represent those rays which strike the individual honeycomb channel parallel to the optical axis and thus at the smallest possible angle. Thus, the divergence of the sub-illumination beam is present the honeycomb condenser given by the full aperture angle between the imaging beam paths of the solid lines. The entry-side divergence DIV _E is in 3A symbolically represented by a solid circle whose filled area is to be a measure of the divergence of the part-Bleuchtungsungsstrahlbündels.

In the direction of transmission behind the honeycomb condenser 152 That is, after the honeycomb condenser, it is the imaging beam paths of the dashed lines that tune the divergence of the sub-illumination beam. This exit-side divergence DIV _A is again represented symbolically in the form of a filled circle whose area is greater than that of the circle representing the entry-side divergence, as a result of which the divergence-increasing effect of the honeycomb condenser 152 is clarified.

In 3B shows in contrast to 3A the representation of two partial illumination beams, which are guided by different honeycomb channels of the honeycomb condenser. Both partial illumination beam paths represent imaging beam paths of illumination beams that run parallel to the optical axis and thus meet perpendicular to the honeycomb condenser. It can be seen that the sub-exposure beam guided through different honeycomb channels with the aid of the Fourier optical system 500 in the area of the multiple mirror arrangement 300 be superimposed. The imaging beam paths are superimposed at the same location on the multiple mirror array 300 although they come from two different honeycomb channels.

If the two partial illumination beam bundles shown have a spatial coherence with one another, this can lead, with high spatial coherence, to overlapping periodic intensity fluctuations over the multiple mirror arrangement when superposed onto the multiple mirror arrangement, which is schematically represented by an intensity function 310 are shown. One or more coherence-reducing elements can therefore be inserted in the beam path, for example suitable phase elements 153 in the area of the honeycomb condenser. A phase element can be designed such that it influences the relative phases of different partial exposure beams differently in the region of the honeycomb condenser and therefore shifts them in phase relative to one another, so that in the region of the superimposition in the multiple mirror arrangement an overlapping of many periodic functions results overall lead to a significant reduction in the magnitude of the intensity variations in the area of the multi-mirror array.

Of the Use of phase elements is, for example, in the December 21st US Provisional Application filed in 2007 with serial no. 61/015918 the applicant, the disclosure of which by reference is made to the content of the description.

In order to optimize lithography processes, it is usually desirable to set the two-dimensional intensity distribution in the pupil shaping surface of the illumination system with high accuracy and high spatial resolution. If a light modulation device with a two-dimensional field arrangement of individually controllable individual elements is used for structuring the illumination pupil, with the aid of which the angular distribution of the incident radiation can be changed (such as, for example, multiple mirror arrangement 300 ), the spatial resolution can be achieved by a correspondingly large number of individual elements with adapted effect characteristics. For example, the multi-mirror arrangement may contain more than 500 or more than 1,000 or more than 2,000 or more than 4,000 controllable individual elements. On the other hand, the design becomes more complex and the size becomes larger as more individual elements are to be accommodated, so that the upper limit of the number of individual elements is practically some tens of thousands of individual elements for practical reasons, for example less than 80,000 or less than 60,000 or less 40,000 individual elements. Light modulation devices with high spatial resolution therefore generally have a surface area which is at least one or more square centimeters and may for example be between about 2 cm ² and 80 cm ² to 100 m ² or more.

In order to achieve complete illumination of all individual elements of the light modulation device intended for use, it is thus necessary for the cross section of the illumination beam to be adapted to the cross-sectional area of the light modulation device to be illuminated. In the case of a laser beam bundle entering the pupil shaping unit with typical cross-sectional areas of, for example, 10 to 100 mm ² , an adaptation of the beam cross section to the size of the light modulator device to be illuminated is thus required. This object is achieved in the described embodiments by the Fourier optical system 500 taken over, which is explained in more detail below.

4 shows in 4A a schematic representation of the pupil forming unit with honeycomb condenser 152 , Fourier optics system 500 , Lens array 155 and micromirror arrangement 300 , The Fourieroptiksystem is folded for reasons of space Z-shaped, including two deflection prisms 501 . 502 are arranged with planar reflecting Umlenkspiegelflächen in the beam path. 4B shows a variant in which the honeycomb condenser by another light mixing device 152A has been replaced, which may contain, for example, a rod integrator or an optical fiber or a corresponding fiber optic bundle. The Fourier optical system 500 is designed so that it is able to transform an input beam with small divergence into a beam of relatively large cross-section. For this purpose, the relatively small input-side beam angles are converted into beam heights on the outlet side of the Fourier optical system.

With reference to the embodiment of a Fourieroptiksystems 500 in 5 In the following, the structure and function of a Fourier optical system will be described 500 which is designed for use as a beam adjustment system in the illumination system of a microlithography projection exposure apparatus. The Fourier optical system transforms the radiation power distribution, which is present with very low optical conductivity and high power, in the entrance plane 510 or transmitter level of the Fourier optical system is present, while maintaining this Lichtleitwertes in the entry plane to the Fourier-transformed exit plane 520 , which can also be referred to as the recipient level. The receiver level 520 can in the built-in lighting system Fourieroptiksystem z. B. in the vicinity of the lens array 155 lie. The penetrating beam defines an entrance surface in the entry plane 511 and in the exit plane an exit surface 521 which is related to the entrance surface via a Fourier transformation. The local radiation power distribution of each individual surface element in the entrance surface 511 In this case, it is distributed over the entire surface of the exit surface, so that the radiation power originating from the entry-side surface elements is superimposed additively in each case in the exit surface. This results in a homogenization of the local radiation power.

The Fourier optical system 500 , hereinafter also referred to as "F-optics" for short, has five lenses and no further optical elements with refractive power, so it is constructed purely refractive. The lenses are arranged in two spaced-apart lens groups LG1, LG2. The entrance-side first lens group LG1 has an entrance-side first lens L1-1 with positive refractive power (positive lens, p-lens) in the form of a biconical lens and an exit-side second lens L1-2 with negative refractive power (negative lens, n-lens) in the form of an exit side concave negative meniscus lens. The second lens group LG2 arranged at a large distance d has an entrance-side first negative-power lens L2-1 in the form of a biconcave lens, and two exit-side second lenses, namely an entrance-side concave positive meniscus lens L2-2 and a subsequent biconvex positive lens L2 -3, which forms the exit side last lens of the Fourieroptiksystems.

The system data is given in Table A. The column "radius" indicates the radius of curvature of the respective surfaces, the column "thickness" the center thickness on the optical axis. Table A Distance d area Radius r group element [Mm] [Mm] shape Thickness [mm] material LG1 L1-1 S1 276 convex 10 SiO2 S2 3910 convex 300 LG1 L1-2 S3 154 convex 3 S4 67.66 concave SiO2 1254 LG2 L2-1 S5 73.45 concave 4 SiO2 S6 21.76 concave 182 LG2 L2-2 S7 403.8 concave 12 SiO2 S8 133.6 convex 56 LG2 L2-3 S9 771.4 convex 15 SiO2 S10 307.5 convex

There is a distance L between the vertexes of the entry-side first system surface S1 of the weakly positive first lens L1-1 and the convex, exit-side last system surface S10 of the exit-side positive lens, which indicates the physical length of the Fourier optical system. Between the exit Side last system area S4 of the first lens group LG-1 and the entry-side first system surface S5 of the second lens group is a group distance d _G , which is many times greater than the correspondingly defined group lengths of the lens group LG-1 and the second lens group LG-2.

In the Fourier optical system 500 is the entry level or sender level 510 in the front focal point of the Fourier optical transmitter 500 while the exit level or receiver level 520 in the rear focal point of the transformer 500 located. Thus, the distance A between the entrance plane and the exit plane corresponds to the focal plane distance of the optical system

The embodiment has a focal plane distance A = 1,750 mm and a length L = 1665 mm. The group distance d _G between the lens groups LG1, LG2 is 1.254 mm. The Fourier optical system has a focal length f _FOS of 25,000 mm. The following power distribution it is in diopters ^[m-1]: L1-1: 2.0; L1-2: -4.0; L2-1: -30; L2-2: 2.50; L2-3: 2:50; Total refractive power: 0.040.

These values exemplify a first particularity of this type of Fourier optical system. As is known, a thin positive lens performs a Fourier transformation between its front focal plane and its rear focal plane with a focal length f, in which case the distance between the front and rear focal plane corresponds to twice the focal length, ie a focal plane distance 2f. With a focal length of 25,000 mm (as in the embodiment of the Fourier optical system 500 ) would thus result in a focal plane distance of 50,000 mm. In contrast, the focal plane distance A = 1.750 mm of the exemplary embodiment is smaller by a multiple, so that the Fourier optical system is built axially very compact compared to its long focal length. This is one of the prerequisites that makes it possible to integrate such a Fourier optical system into a lighting system of a projection exposure apparatus.

Further boundary conditions to be taken into account in the design of a Fourier optical system result from the fact that the Fourier optical system within the illumination system of a microlithography projection exposure apparatus is to operate as a beam guidance system for high-energy laser radiation. Within the Fourier optical system, the power to be transmitted is guided by relatively small area optical elements, resulting in high local radiated power. However, the optical material used for the lenses can degrade under exposure to material-specific limits under radiation exposure. In order to avoid a degradation of the lenses, the energetic load of the lenses, for example measured as energy density in [mJ / cm ² ], should be kept as low as possible or be below a material-specific threshold for each lens. On the other hand, components that are permanently exposed to higher levels of radiation exposure might need to be replaced during the life of the optical system.

Around a radiation-induced degradation of the lenses could avoid the system can be constructed so that the optical elements not at the narrowest bottlenecks of the energetic flow, which can also be formulated as a requirement that the irradiated Surface on the respective lenses not too small may. In principle, it would be possible with increasing distance between transmitter and receiver level, So with a larger space, the irradiated cross sections bigger in order to over that way avoiding the thresholds. However, as stated above, the distance between transmitter and Receiver level, and thus needed by the Fourier optical system Space to be kept as small as possible, which in turn increases the risk of strong radiation exposure of individual lenses. Thus, the length of the Fourieroptiksystems and the lifetime of this Fourier optical transmitter Operate with high energetic radiation conflicting Requirements. It therefore requires special consideration to form orders of powers in a limited Space between a front and a rear focal point of a Fourieroptiksystems come which at the same time the local radiation power limited to values below material-specific limits.

A method for determining the spatial arrangement of powers taking into account the system-specific thresholds for the radiation load is explained in more detail below. In this case, the optical system is described by means of parameters in the so-called diagram of Delano (Delano diagram) and this description is linked to a loading model taking into account the radiation load. The description of optical systems using Delano diagrams is known per se and will be described below in connection with the 6 to 8th explained, as far as this seems useful for the present application. Fundamentals of Delano diagrams, for example, the professional article "First Order Design and the y-ybar diagram" by E. Delano, Applied Optics, Vol. 12, December 1963 be removed.

In Delano charts generally become the beam heights or Beam angle of two selected rays, namely one Aperturstrahls and a field beam shown. The aperture ray For example, the aperture edge beam (short: edge beam) and For example, the field beam may be the aperture center beam. Lies the aperture diaphragm on the main surface, so corresponds to the Aperture center beam to the main beam (chief ray, principal ray), the main surfaces are then simultaneously entrance and exit pupil. If, on the other hand, the aperture diaphragm lies in the front focal plane, then the aperture center beam the focus beam. The exit pupil is located then at infinity.

6 schematically shows the paraxial beam path of a reference beam RR and a marginal ray MR of a beam through a lens L with focal length f between the entrance pupil EP and the receiver plane IM. The reference beam used here is a beam which extends from an edge point of the object field or the optically conjugate receiver field and intersects the optical axis in the area of the entrance pupil or a pupil plane optically conjugate thereto. A marginal ray leads from the intersection of the optical axis with the object plane or the image plane optically conjugate to the outer edge of the aperture diaphragm or, in the representation of 6 , to the outer edge of the entrance pupil. The vertical distance of these rays from the optical axis gives the corresponding steel heights, which are referred to as the reference ray height or marginal ray height. In a Delano diagram, the (paraxial) principal ray height is plotted against the (paraxial) marginal ray height in a planar diagram, with the principal ray height in the x-direction (abscissa) and the marginal ray height in the y-direction (ordinate). In 7A and B is explained in more detail. Conventionally, the ordinate (edge beam height) is designated by parameter "y" and the abscissa (x direction) by parameter "ybar". The dashed line ray A corresponds to the vector addition of the reference ray and the edge ray and is referred to herein as the "Delano ray". This ray corresponds in each plane perpendicular to the optical axis to a point in the plane of the Delano diagram. The projection of this beam into the y-ybar plane corresponds to the Delano diagram ( 7B ).

From the Delano diagram some properties of the optical system can be read directly or calculated relatively easily. The object plane or the image plane corresponds to the intersection with ybar, since there the reference beam height is maximum and the marginal beam height is equal to zero. A pupil plane corresponds to the intersection with y, since here the reference beam height is zero. A lens diameter, ie the optically free diameter of a lens (or a mirror) corresponds to the amount sum | y | + | ybar |. The refractive power of an optical surface of a lens or a mirror corresponds to a change of direction (cf. 7A ). The axial distance d between two lenses corresponds to the area of a triangle spanned between the origin of the Delano diagram and the points defined by the lenses (see 8th ).

In addition to these properties known per se, the radiation load of lenses or optical elements in the Delano diagram can be represented, which is in connection with 9ff will be explained in more detail. For the load-optimized arrangement of lenses within an optical system, the axial distance between the lenses and the energetic loading of the lenses are essential parameters, the linking of which will be explained in more detail below.

8th illustrates that the geometric axial distance d between two points (y _1, ybar ₁₎ and (y _2, ybar _2), which are connected by a straight line in the Delano diagram, is proportional to the triangular area between these points and the Origin is spanned. The axial distance d can be determined as follows:

In this equation, H is the Lagrange invariant corresponding to the geometric waveguide value LLW (geometrical flux, etendue), n is the refractive index between the points, and y _i and ybar _i are the coordinates of the corresponding surfaces in the Delano diagram. The determinant of the matrix corresponds to the triangular area.

From the beam heights for reference beam and boundary beam, assuming homogeneous energy sources, a load model for the optical elements of the system was derived. This is based on the 9ff explained in more detail. First, the derivation in a one-dimensional case (1D) is explained, the 1D case, for example, corresponds to a system with cylindrical lenses (curvature in only one plane). The integral power at any z position, ie at any position along the optical axis se, is proportional to the convolution of the principal ray height with the marginal ray height. Starting from two tophat loads, the convolution results in a trapezoidal stress whose 50% width is given by the maximum Max (| y |, | ybar |). Because of the assumed energy conservation within the system, which can also be described as conservation of the geometric light conductance LLW, the trapezoidal area is independent of the z position, ie A1 = A2 = .... = A _n . From the 50% width B of the trapezoidal surfaces and the respective trapezoidal height, the trapezoidal area is calculated to B · h. With B = Max | y |, | ybar |) and h = P, P · Max (| y |, | ybar |) = constant for all z positions. Here h = P corresponds to the peak load of the corresponding area. In the two-dimensional case, ie for system surfaces with bends in multiple directions (as in the spherical optics) one obtains P · (Max (| y |, | ybar |)) ² = constant, since the two dimensions separate.

In the Delano diagram, constant loads are represented as a square around the origin ( 9B ). Within the square the loads are higher, outside smaller. The size of the square, which can be parameterized, for example, over the edge length or half the edge length, is thus a measure of the load threshold to be taken into account in the load model, which should not be exceeded in any of the system areas. The load model is derived here for a conventional lighting setting, but can also be extended to other settings. Likewise, the model can also be applied to other intensity profiles, such. As a Gauss distribution to be extended.

The influence of the radiation load on the overall length and the number of refractive powers is illustrated by means of the following examples. For example, a Fourier optics system with a focal length f = 25,000 mm, an entrance pupil diameter EPD = 36 mm, an entry-side numerical aperture NA _o = 0.0018, a length L = 1,800 mm and a load threshold determined by the lens materials with a radiation load <20 mJ / cm ^{2 are} sought , The Fourier optical system is thus a geometric light conductance H = EPD / 2 * NA _o · n _o = 0.033 mm designed, where H is the product of the radius EPD / 2 of the entrance pupil, the entrance-side numerical aperture NA _o and the entrance-side refractive index n _o is ,

10 shows the corresponding Delano diagram of this F optics with only one refractive power (Bk). The entrance pupil is at position a, the refractive power Bk at position b and the image plane at position d. The outer square shows the load threshold for 20 mJ / cm ² for the 2D model, the inner square for a 1D model. The length of this Fourier optical system with only one refractive power is 50,000 mm, which is twice the focal length.

As related to 8th As already explained, the area inscribed by the Delano beam towards the origin should be as small as possible if the overall length of a system is to be as short as possible. A Fourier optical system with two powers, ie with lenses at positions a and d in 10 , would only halve the length.

11 explains that at least three refractive powers are required to shorten the overall length significantly. This shows 11 the Delano diagram of a Fourier optics system with three refractive powers (3-Bk optics) in the sequence positive refractive power - negative refractive power - positive refractive power (pnp). The Delano diagram shows that a short overall length and a low peak load (peak load) interfere with each other. Because none of the lens surfaces are inside the outer square 2D should be arranged, if the peak load should not be exceeded, results in a minimum distance, which is proportional to the area of the hatched triangle.

12 schematically shows the Delano diagram of a Fourier optical system with four lenses (4-Bk optics) with the refractive power sequence positive-negative-negative-positive between the entrance pupil and the image plane. This pnnp system has only a length of 1,600 mm, which is recognizable by the fact that the hatched triangular area between the origin and the projection of the Delanostrahls is much lower than that in 11 or 10 ,

From in connection with 8th explained distance formula for the distance d between lenses in combination with the load model can be derived, which minimum distance between the near-pupil refractive power a and the near field refractive power b must be at least when the Fourieroptiksystem is designed for transmitting a radiation energy E per unit time at a Lichtleitwert and an energetic peak load P on the lens surfaces should not be exceeded. In this description, let n be the refractive index in the medium between the lenses, H the Lagrange invariant indicating the geometric waveguide according to H = EPD / 2 · NA _o · n _o [mm], E the energy (in [J] given by the optical system is to be transported and P is the peak load on the lenses (in [J / mm ² ] for the two-dimensional case (2D) and in [J / mm] for the one-dimensional case (1D) onale load model: d = n / H · E / (P a P b ) 1.2 (A2) and for the one-dimensional load model (eg for cylindrical lenses) d = n / H · E 2 / (X a P a x b P b ) (A3)

In (A3), the parameters x _a and x _b respectively indicate the beam extension in the unfolded direction.

The distance d may correspond to the above-defined group distance d _G between the exit-side last system area of the first lens group and the entry-side first system area of the second lens group (cf. 5 ).

A load-optimized Fourier optical system can be designed, for example, so that the load on these energetically particularly endangered lenses (lens a corresponds to the last lens of the entrance-side lens group LG1 and lens b corresponds to the first lens of the exit-side lens group LG2) is approximately evenly distributed, so that P _a is about P _b . Uneven loading of the two lenses is also possible as long as the individual loads are not above the threshold value. Taking into account that the group distance d _{G describes} the axial distance between the near-field powers of the first lens group LG1 and the near-pupil powers of the second lens group LG2, a Fourier optics system with a very long focal length and a very short overall length can be created by essentially using the og least possible group spacing between the lens groups is set. A shorter distance will always result in a higher load on at least one of the two hochbelaste th lenses, whereby the risk of excessive radiation exposure of the corresponding lens and an associated degradation of the lens increases.

at the design of Fourier optical systems for use in lighting systems of microlithography projection exposure equipment into the o. g. Parameters are, for example, in the following areas. The energy E transported by the Fourier optics system should, for example, be in the range between 2 mJ and 20 mJ, in particular in the range between 5 mJ and 10 mJ, for example at about 7 mJ to 8 mJ. The Lagrangian invariant H can be, for example between 0.01 mm and 0.2 mm, in particular in the range between 0.02 mm and 0.1 mm, for example in the range of 0.03 mm to 0.05 mm.

in the Regard to the very low divergence of the transmitted Radiation at the entrance side of the Fourier optical system can Lagrange invariant also using the marginal ray or reference ray heights and boundary beam or reference beam angle can be parameterized. Then H = n * (y * sin (ubar) -ybar * sin (u)), where n is the refractive index, y the marginal beam height, ybar the reference beam height, u the marginal beam angle and ubar the reference beam angle is.

The material-specific peak loads P when using calcium fluoride (CaF ₂ ) as a lens material, for example, in the range between 5 mJ / cm ² , in particular in the range of about 10 mJ / cm ² , while using fused silica, for example be in the range between 0.2 mJ / cm ² and 1.5 mJ / cm ² , in particular in the range of about 0.5 mJ / cm ² . The peak load may in some cases, e.g. B. in improving the current material properties, also higher. For example, it may be possible to shift the range of allowable peak load of calcium fluoride to higher levels, e.g. B to the area around

20 mJ / cm ² or by about 30 mJ / cm ² or by about 40 mJ / cm ² .

The method just described using Delano diagrams has been used to simplify the load-optimized design of the embodiment in FIG 5 to obtain. Important data of this embodiment may be summarized as follows: transmitter plane in the front focal point of the Fourier optical system Diameter of the entrance surface: 35mm Distance to the first element L1-1: 75 mm Receiver plane in the rear focal point of the Fourier optical system Diameter of the exit surface: 100 mm Distance to last element L2-3: 10 mm Distance A between transmitter and receiver (corresponds to focal plane distance 1750 mm Group distance d _G : 1254 mm Focal length f _FOS : 25000 mm Ratio length L to focal plane distance A: 0950 Group distance / focal length ratio (d _G / f _FOS ): 0050 Ratio length / focal length (L / f _FOS ): 0075

Table B shows the area ratios and loading conditions: Table B Diameter [mm] Area [mm ² ] Load ratio normalized to load in the entry surface entry surface 35 1225 1 Area of maximum load (smallest system surface) 12 144 <9 times exit area 100 10000 0.1225 times

To illustrate shows 13 a semi-quantitative diagram in which the area-related radiation power density S in [W / m ² ] is plotted in a bar chart for various excellent areas of the system. Among other things, it can be seen from the diagram that the radiation power density in the entry plane (transmitter plane) 510 is higher by a factor of 2 than in the Fourier-transformed exit plane (receiver plane) 520 because the irradiated area is correspondingly larger due to the beam expansion than in the entrance plane. The highest radiation load occurs at the entrance-side negative lens L2-1 of the second lens group LG2, onto which the radiation beam is focused by the lenses of the first lens group. However, the radiation power density at the highest loaded lens L2-1, despite the very short length is less than 9 times as high as the area-related radiation power density in the entrance surface, and it is approximately 0.5 mJ / cm ² well below the material-specific damage threshold of approx. 0.6 mJ / cm ^{2 of} the synthetic quartz glass used.

The refractive power sequence pnnp is not the only way to construct a Fourier optical system with a similar Telefaktor TF = (L / f _FOS ). Other variants with a similar length are possible with the refractive power sequences pppp, pnpp or ppnp.

14 shows corresponding Delano diagrams. All examples are calculated under the boundary conditions f _FOS = 25,000 mm, EPD = 36 mm, NA _o = 0.0018 and a length of the second lens group of less than 250 mm and designed for a maximum load of 20 mJ / cm ² . 14A shows the Delano diagram of an embodiment of a 4-lens system with refractive power pppp. The total construction length is 1,526 mm. 14B shows the Delano diagram for a 4-lenser with refractive power pnpp and length of 1.576 mm. 14C shows the Delano diagram for a 4-lenser with refractive power ppnp and length of 1.576 mm. The lenses are each individual lenses.

As is known, in some cases it may be advantageous to divide a single lens into two or more lenses, in which case the refractive power of the resulting multi-lens lens group may correspond substantially to the refractive power of the single lens. As a result, additional degrees of freedom, for example, in the correction of aberrations are possible. Other variants with 5, 6 or more lenses and corresponding power combinations are also possible accordingly. However, the four single-lens systems are representative basic forms for building large-focal-length, axially compact systems with a relatively small telefactor L / f _FOS , for example, with a telecoctor less than 1/6 or 1/8 or 1/10.

The Fourier optical system 500 is telecentric on its input side (transmitter side) and on its output side (receiver side). Specifically, in order to allow the radiation to pass substantially perpendicularly through the exit plane (receiver plane), three lenses are provided in the second lens group LG2, the exit-side last lens L2-3 essentially providing the exit-side telecentricity.

at In the embodiment, all lens surfaces spherical. In other embodiments, at least one of the lenses at least one aspherical lens surface. In particular, the exit plane next, exit side Lens surface S10 have an aspherical shape, in particular, to contribute effectively to the exit-side telecentricity.

Around a uniform overlay of the local To obtain radiation powers in the exit plane, the Abbe sine condition at least approximately fulfilled be. With noticeable deviations from the sine condition it can to the variance of local irradiance and thus to an incomplete homogenization in the exit plane come. It has been shown that compliance with the sine condition can be simplified when the refractive power of the inlet side Lens element of the second lens group, d. H. the lens L2-1, is distributed to two or more lens elements. In such embodiments Accordingly, the second lens group may be four or five Have lens elements.

In the following it will be explained with reference to a further embodiment that it is possible to further reduce the above-described minimum length limit by using at least one pair of two mutually orthogonally oriented cylindrical lens systems. 15 shows an example of a perspective view of a Fourieroptiksystems 1500 , which consists of two successive cylinder optics Z1 and Z2. That of the entrance area 1510 Immediately following first cylinder optical system Z1 consists exclusively of cylindrical lenses which have a finite radius of curvature (and therefore a refractive power in the xz-plane) in the x-direction, but are not curved in the orthogonal y-direction and therefore in the y-direction (ie in the yz plane) have no refractive power. These cylindrical lenses are also referred to herein as "x-lenses". The downstream second cylindrical lens system Z2 has exclusively lenses with finite radius of curvature in the y-direction and infinite radius of curvature in the x-direction (y-lens). The first cylindrical lens system constructed with x-lenses has two lens groups LGX1, LGX2, which are spaced apart from one another, and between which there is a spacing which is greater than the sub-length of the lens groups LGX1 or LGX2. The entrance-side first lens group LGX1 has an entrance-side first lens L1X-1 with positive refractive power and an exit-side second cylinder lens L1X-2 with negative refractive power. The second lens group LGX2 located far behind has an entrance-side first lens L2X-1 whose refractive power is divided into two lens elements, and an exit-side second lens L2X-2 whose refractive power is divided into two directly successive positive-phase cylinder meniscus lenses , Instead of the split lenses, individual lenses may be provided in each case. The second cylindrical lens system Z2 acting in the y direction has a corresponding structure with an entrance-side first lens group LGY1 and an exit-side second lens group LGY2 and refractive power sequence pnnp. The first cylindrical lens system Z1 causes a constriction of the radiation power of the incoming beam exclusively in the x-direction, while the downstream second cylindrical lens system acts exclusively in the orthogonal y-direction. The one-dimensional refractive powers are adapted to one another such that, in the case of an entry-side quadratic input field, the output field is also square again. In the following, it is explained why such a construction with a large number of cylindrical lenses can nevertheless lead to a more axially compacted Fourier optical system than to a structure with rotationally symmetrical lenses.

16 shows for this purpose a schematic representation of an input field shown on the left with a square cross-sectional area, which is transmitted through a Fourier optical system FOS into an output field shown on the right. The input beam has a square cross section with edge length 2a and a divergence DIV which is twice the numerical aperture NA of the beam (DIV = 2NA). As explained above, it is possible to determine the minimum length permissible under radiation load aspects using Delano diagrams. 5 shows a simple example of such a Fourier optics system with pnnp structure and five lenses. In this case, p corresponds to a lens with positive refractive power and n to a lens with negative refractive power. (For cases in which the telecentricity of the output field is irrelevant, as may be the case with photo lenses, the last lens with positive refractive power on the output side can also be dispensed with, so that a pnn structure results).

17 shows a simplified Delano diagram for a Fourier optical system with the requirements 15 with four lenses (pnnp). Lenses within the hatched area "see" an energetic load that is greater than or equal to oI _o . The parameter ξ is therefore a parameter that is the material-specific Exposure upper limit of the lens material describes. The total power P transported through the system is constant, I ₀ = P / 4a ²

For simplification reasons, an isotropic input radiation field with a homogeneous intensity distribution I _{0 is} assumed below, ie each surface element of the entrance surface "sees" the same intensity I ₀ . Furthermore, assume a top-hat-shaped angular distribution, which essentially means that there is a uniform beam angle distribution between a minimum beam angle and a maximum beam angle. Under these assumptions, for rotationally symmetric optical systems, the maximum intensity in each plane along the optical axis can be calculated according to the following equation (B1):

wobei der Index i über die Knoten im Delano-Diagramm und somit über alle Linsen läuft. Daher hat beispielsweise ein Fourieroptiksystem mit vier Linsen mit der geringsten Baulänge seine Knotenpunkte im Delano-Diagramm bei (0, a/√ξ) und (a/√ξ, 0). Ein solches idealisiertes Optiksystem nötigt daher nur Bauraum zwischen der austrittsseitigen letzten Linse der ersten Linsengruppe und der eintrittsseitigen Linse der zweiten Linsengruppe, wobei sich dieser Bauraum gemäß Gleichung (B2) wie folgt ergibt:

Here, max (y, y) is the maximum of the two coordinates in the Delano diagram and represents a measure of the beam expansion. If the upper load limit expressed by the intensity of the input beam given by I _max = ξI ₀ , then results after insertion into equation (B1) in the Delanodiagram, a region forbidden for lenses with a square shape and half the edge length a / √ξ around the origin (see hatched area in 17 ). The length of the optics shown in the Delano diagram can then be represented by the following equation (B2):

where the index i runs over the nodes in the Delano diagram and thus across all lenses. Therefore, for example, a Fourier lens system with four lenses of the shortest length has its nodes in the Delano diagram at (0, a / √ξ) and (a / √ξ, 0). Such an idealized optical system therefore only requires installation space between the exit-side last lens of the first lens group and the entrance-side lens of the second lens group, this installation space resulting according to equation (B2) as follows:

This distance d corresponds to the minimum group distance d _G. In the case of an optical system of cylindrical optics, as shown schematically in 15 is shown, the beam path is constricted in a system with four lenses according to the above example only in one dimension. This has the consequence that the radiation load only increases linearly with the jet constriction. Expressed in Delano coordinates, the maximum energy load along the optical axis can then be calculated as follows:

A comparison with equation (B1) shows that here the edge length a is not quadratic, but only linear. This results in a "forbidden" range for lenses in the Delano diagram with half the edge length σ / ξ and thus a distance lower limit of

However, in order to form a beam bundle by means of cylinder optics in two mutually orthogonal directions, successively arranged cylinder optics with orthogonally oriented curvature surfaces can be provided. Since the radiation beam has already been formed in one direction, the load limit in the Delano diagram for the second cylinder optics changes to a ² / bξ. This results in a minimum length according to:

This results in a ratio of the lengths of the conventional rotationally symmetric optics (at given by the minimum group distance d or d _G ) and that with two mutually orthogonal cylinder optics (indicated by d (1) cylinder + d (2) cylinder from:

from that This results in the following result: Although in the case of using Cylinder optics two Fourier optics systems for the two Spatial directions (x and y direction) can be arranged one after the other need, the overall length is large Values of ξ, d. H. for large load limits, shorter than in the case of a rotationally symmetric Fourier optical system. For example, have the input field and the output field the same Size (a = b), then the system is with cylinder optics for ξ> 2 shorter.

in the Generally, the overall length is slightly larger be as the relationships shown here under simplifying assumptions. On the one hand it plays a role, that also a certain one Space between the lenses within a lens group (i.e. z. Between lenses L1-1 and L1-2 and between L2-1 and L2-2) is necessary. It is also to be considered that each cylinder optical system one of the length of its orthogonal optical systems corresponding input slice size or output slice size needed. Nevertheless, a Fourier optical system with a Sequence of at least two pairs of orthogonally oriented Cylinder optics systems with the same total focal length under circumstances be designed axially shorter than a round optical system (rotationally symmetric System) of the same focal length.

To demonstrate the circumstances is in connection with the 18 to 20 a comparative example is explained quantitatively. For this example, a = 18 mm, NA = 0.0018, b = 18 mm and ξ = 3. 18 shows the corresponding Delano diagram of a rotationally symmetric Fourier optical system with four lenses and lens sequence pnnp. The system has a total length of 3692 mm. In a total construction with Zylinderopti ken results in the following picture. 19 1 shows a Delano diagram of a first cylinder optics with a long output slice width, which in the direction of irradiation in front of a second cylindrical optics (FIG. 20 ) is arranged. The first cylinder optic has a length of 1383 mm and an initial cutting width of 1469 mm. The second cylinder optics, whose Delano diagram in 20 has a length of 1386 mm and an input section of 1466 mm. This results in a total construction length of the built-up with cylindrical optics Fourieroptiksystems of 2852 mm, which is thus approximately a factor of 1.3 shorter than the rotationally symmetric optics same optical performance.

The Divided into two consecutively arranged "pure" cylindrical lens systems (ie systems built only with x-lenses or only with y-lenses) is not mandatory. The cylindrical lenses can also be arranged nested be so that x-lenses and y-lenses may alternate several times can. This may possibly further the length be shortened. For example, a structure with refractive power sequence is p_x / n_x / n_x / p_y / p_x / n_y / n_y / p_y possible, with z. B. p_y a y lens with positive power designated.

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 2007/0165202 A1 [0007, 0070]
• WO 2005/026843 A2 [0007]
• - US 6563556 [0008]
• US 2004/0265707 [0008]
• US 5343489 [0059]

Cited non-patent literature

• "First Order Design and the y-ybar diagram" by E. Delano, Applied Optics, Vol. 12, December 1963 [0088]

Claims (28)

A lighting system for a microlithography projection exposure apparatus for illuminating an illumination field with the light of a primary light source, comprising: a variably adjustable pupil shaping unit for receiving light from the primary light source and producing a variably adjustable two-dimensional intensity distribution in a pupil shaping surface of the illumination system, the pupil shaping unit being a Fourier optical system for conversion an entrance ray bundle entering through an entrance plane of the Fourier optical system into an exit beam emerging from an exit plane of the Fourier optical system, the Fourier optics system having a focal length f _FOS and a length L measured between an entrance side first system surface and an exit end last system surface along an optical axis; L / f _FOS ) <1/6 applies. The illumination system of claim 1, wherein the condition (L / f _FOS ) <0.1. Illumination system according to claim 1 or 2, wherein the focal length f _{FOS of} the Fourier optical system is more than 10 m and the length L is less than 4 m. The illumination system according to claim 1, wherein the Fourier optical system comprises a first lens group having an entrance-side first lens and an exit-side second lens and a second lens group downstream of the first lens group having an entrance-side first lens and an exit-side second lens, between an exit-side last system surface the first lens group and an entrance-side first system surface of the second lens group is a group distance d _G. The illumination system according to claim 4, wherein between the exit-side last system area of the first lens group and the entrance-side first system area of the second lens group is a group distance d _G for which the condition d _G > 0.66 · L holds. An illumination system according to claim 4 or 5, wherein between the exit-side last system area of the first lens group and the entrance-side first system area of the second lens group there is a group distance d _G for which the condition d _G <0.1 · f _FOS holds. Lighting system according to one of the preceding claims, wherein the Fourieroptiksystem is designed to transmit a radiation energy E per unit time at a H, P _A a predetermined maximum energetic load of the exit side second lens of the first lens group and P _B a predetermined maximum energetic load of the entrance side first lens of the second Lens group, and a group distance d _G between an exit-side last system surface of the first lens group and an entrance-side first system surface of the second lens group is not smaller than a minimum group distance d _G ^min , where the minimum group distance is: d G min = n / H · E / (P a P b ) 1.2 The illumination system of claim 7, wherein the group spacing d _{G is} between d _G ^min and 3 · d _G ^min . An illumination system according to any one of the preceding claims, wherein the Fourier optical system is designed to have a geometric light guide of 0.01 mm ≤ H ≤ 0.2 mm, where H is the product of the entrance pupil radius EPD / 2, the entrance numerical aperture NA _o and the entrance refractive index n _o , Lighting system according to one of the preceding Claims in which the Fourier optical system has four, five or has six lenses. Lighting system according to one of the claims 4 to 10, wherein the entrance-side first lens group is exactly two Has lenses. Lighting system according to one of the preceding Claims in which the Fourier optical system has at least one Lens with at least one aspherical lens surface Has. Lighting system according to one of the preceding Claims in which the next level of exit, exit-side lens surface of the Fourier optical system a has aspheric shape. Lighting system according to one of the preceding Claims in which the Fourier optical system on the input side and is telecentric on the output side. Lighting system according to one of the preceding Claims in which the Fourier optical system has at least one plan to include deflecting mirrors. Lighting system according to one of the preceding Claims in which the Fourier optical system is Z-shaped folded. Lighting system according to one of the preceding Claims in which the Fourier optical system is at least one Pair of crossed cylindrical lens systems, wherein a pair of crossed cylindrical lens systems a first cylindrical lens system with at least one curved in a first plane of curvature first cylindrical surface and a second cylindrical lens system with at least one in a second curvature surface has curved second cylindrical surface, wherein the first and the second plane of curvature perpendicular to each other stand. The illumination system of claim 17, wherein the Fourier optical system a first cylindrical lens group with a plurality of first cylindrical lenses and a downstream second cylinder lens group having a plurality having second cylindrical lenses. Lighting system according to one of the preceding Claims wherein the pupil shaping unit is a Fourier optical system having upstream light mixing device. A lighting system according to claim 19, wherein the light mixing means a honeycomb condenser. Lighting system according to one of the preceding Claims, wherein the pupil shaping unit is a light modulation device for controllably changing an angular distribution of on the light modulation device incident light beam and the Fourier optical system between the primary light source and the light modulation device is arranged. An illumination system according to claim 21, wherein the light modulation means a two-dimensional array of individually controllable individual elements includes, with the help of which the angular distribution of the impinging Radiation is changeable. Lighting system according to claim 21 or 22, wherein the light modulation device a multi-mirror array (multi-mirror array, MMA) with a large number of individually controllable individual mirrors includes. Fourier optical system for converting an entrance beam entering through an entrance plane of the Fourier optical system into an exit beam emanating from an exit plane of the Fourier optical system, the Fourier optical system having a focal length f _FOS and a length L measured along an optical axis between an entrance side first system surface and an exit side first system surface; Condition (L / f _FOS ) <1/6 applies. Fourier optical system according to claim 24, characterized by the features of the characterizing part of one of the claims 2 to 18. Light mixing system for receiving light from a primary Light source and to produce a substantially homogeneous two-dimensional Intensity distribution in a lighting area, wherein the light mixing system is a Fourier optical system according to any one of Claims 24 or 25 and the Fourieroptiksystem preceded by an effective in the angular space light mixing device is. A light mixing system according to claim 26, wherein the light mixing means a honeycomb condenser. A microlithography projection exposure apparatus for exposing a radiation-sensitive substrate arranged in the region of an image surface of a projection lens to at least one image of an image in Be a pattern of a mask rich in an object surface of the projection lens, comprising: a primary light source; an illumination system for receiving the light of the primary light source and for forming illumination radiation directed to the pattern of the mask; and a projection lens for imaging the structure of the mask on a photosensitive substrate, wherein the illumination system is constructed according to one of claims 1 to 23.