WO2016188810A1

WO2016188810A1 - Pupil facet mirror

Info

Publication number: WO2016188810A1
Application number: PCT/EP2016/061092
Authority: WO
Inventors: Ralf GEHRKE; Thomas Fischer; Martin Endres
Original assignee: Carl Zeiss Smt Gmbh
Priority date: 2015-05-22
Filing date: 2016-05-18
Publication date: 2016-12-01
Also published as: DE102015209453A1; TW201641960A

Abstract

A pupil facet mirror (10) comprises facets (11) having different radii of curvature.

Description

Pupil facet mirror

The content of the German Patent Application DE 10 2015 209 453.7 is incorporated by reference herein.

The invention relates to a pupil facet mirror for an illumination optical unit of a projection exposure apparatus, and also an illumination optical unit and an illumination system for a projection exposure apparatus comprising such a pupil facet mirror. In addition, the invention relates to a method for determining the radii of curvature of facets of a pupil facet mirror.

Furthermore, the invention relates to an optical system and a projection exposure apparatus for microlithography. Finally, the invention relates to a method for producing a microstructured or nanostructured component, and a component produced according to the method.

WO 2009/07421 1 Al discloses a correction device by means of which a uniform intensity distribution in an illumination field is settable by way of a transverse coordinate transversely with respect to a displacement direction of an object displaced during the projection exposure within specific tolerance limits. Such a correction device is also referred to as UNICOM. UNICOMs are additionally known from EP 0 952 491 A2 and

DE 10 2012 205 886 Al .

It is an object of the present invention to develop an illumination optical unit with a corresponding UNICOM. This object is achieved by means of a pupil facet mirror whose facets have different radii of curvature.

According to the invention, it has been recognized that it is thereby possible to reduce dose fluctuations in the object field which can be attributed to shifts in the illumination field caused by source fluctuations in the region of the UNICOM.

It has been recognized in particular that source fluctuations, that is to say spatial variations of the plasma and/or other spatial fluctuations of the radiation source, lead to a shift in the illumination field in the UNICOM plane and thus to fluctuations of the radiation dose that reaches the object field. These fluctuations are not detected by energy sensors arranged in the optical path upstream of the UNICOM. Dose errors can thus occur.

It has further been recognized that the dose errors caused by such source fluctuations can be reduced, in particular minimized, by providing the facets of the pupil facet mirrors with different radii of curvature. It is possible, in particular, to choose the radii of curvature of the pupil facets in such a way that the field shift resulting from the sum of the shifts of all the individual channels in the UNICOM plane is reduced, in particular minimized.

The radii of curvature of the pupil facets can be chosen in particular in such a way that they at least partly compensate for the dose fluctuations in the object field which are caused by a shift of the individual illumination channels on account of source fluctuations.

The facets of the pupil facet mirrors are embodied monolithically, in parti- cular.

The facets of the pupil facet mirrors are embodied in particular in a rigid fashion, that is to say in a non-displaceable fashion. This facilitates the structural layout of the pupil facet mirrors. The pupil facet mirror according to the invention has in particular a predetermined design, in particular a design optimized according to a method that will be described in even greater detail below, that is to say an arran- gement of a multiplicity of pupil facets having predefined, different radii of curvature. In this case, the design of the pupil facet mirrors can be adapted in particular to the details of the provided radiation source or the provided radiation sources of the illumination system and/or to the design data of the illumination optical unit, in particular of the first facet mirror, which images the image of the radiation source from an intermediate focus onto the pupil facets.

In accordance with one aspect of the invention, at least two radii of curvature of the facets differ by at least 1%. The ratio of the largest radius of cur- vature _max to the smallest radius of curvature R_min of all the facets of the pupil facet mirrors Rmax ^'■ Rmin is in particular at least 1.01, in particular at least 1.02, in particular at least 1.03, in particular at least 1.04.

Rmax ^'■ Rmin is in particular at most 1.2, in particular at most 1.1.

The radii R of curvature can be in particular at least 500 mm, in particular at least 1000 mm, in particular at least 2000 mm.

It has been found that an extensive compensation of the field shifts caused by source fluctuations in the UNICOM plane is possible with such differences in the radii of curvature of the facets.

In accordance with a further aspect of the invention, in each case at least 100, in particular at least in each case 200, in particular at least in each case 300, in particular at least in each case 400, in particular at least in each case 500 facets have the same radius of curvature.

This facilitates the production of the pupil facet mirror.

The pupil facet mirror can comprise overall at least 1000, in particular at least 1200, in particular at least 1400, in particular at least 1600, in particular at least 1800, in particular at least 2000 pupil facets. The total number of pupil facets is in particular at most 10 000, in particular at most 8000, in particular at most 6000, in particular at most 4000, in particular at most 2000.

Such a large number of pupil facets makes it possible to set a multiplicity of different illumination settings, that is to say to choose the angle-of- incidence distribution of the illumination radiation on the object field flexibly as required.

In accordance with a further aspect of the invention, the facets form a plurality of different groups, wherein the facets of the different groups have different radii of curvature. The facets of the same group have in each case the same radius of curvature.

Firstly, this facilitates the production of the pupil facets. In addition, this facilitates the optimization of the assignment of the pupil facets to a spe- cific radius of curvature.

The number of groups of facets having different radii of curvature is in particular two, three, four, five or more. It is in particular at most twenty, in particular at most ten. The individual facets have in particular in each case a single radius of curvature. The radius of curvature is in particular constant in each case over the entire reflection surface of a facet. The facet has in particular a spheri- cal reflection surface.

Different radii of curvature lead to different focal lengths of the facets and hence different image distances or image positions.

The object according to the invention is additionally achieved by means of an illumination optical unit comprising a corresponding pupil facet mirror. The advantages are evident from those of the pupil facet mirrors.

In accordance with a further aspect of the invention, the radii of curvature of the second facets are chosen in such a way that a subset of the second facets has a focal length, which is also referred to hereinafter as image distance, which is less than its distance from the correction plane (UNICOM plane).

In accordance with a further aspect of the invention, a subset of the second facets has an image distance which is greater than its distance from the UNICOM plane.

In accordance with a further aspect of the invention, a subset of the second facets has an image distance which substantially corresponds to its distance from the UNICOM plane.

In other words, the radii of curvature of the pupil facets are chosen such that a portion thereof images the field facets into a plane upstream of the UNICOM plane. It is also possible for a portion of the pupil facets to image the first facets into a plane downstream of the UNICOM plane. An imaging of the field facets into the UNICOM plane is likewise possible. Preferably, the radii of curvature of the second facets are chosen in such a way that a first subset of the second facets has an image distance which is less than its distance from the UNICOM plane, and a second subset of the second facets has an image distance which is greater than its distance from the UNICOM plane. This enables a mutual compensation of field shifts caused by source fluctuations in or relative to the UNICOM plane. In addition thereto, a third subset of the second facets can have an image distance which corresponds approximately to its distance from the

UNICOM plane.

On account of a finite extent of the pupil facet mirrors and/or a tilted arran- gement thereof relative to the UNICOM plane, an exact correspondence of the image distance to the distance from the UNICOM plane is not possible or at least not possible for all facets of the same group. Hereinafter, correspondence of the image distance of the facets of a group to its distance from the UNICOM plane should be understood to mean that there exists at least one facet in said group whose image distance deviates by less than 1%, in particular less than 0.5%, in particular less than 0.3%, in particular less than 0.2%, in particular less than 0.1%, from its distance from the

UNICOM plane. The advantages of the illumination system according to the invention are likewise evident from those of the pupil facet mirror.

The radiation source is in particular an EUV radiation source, in particular a plasma source. Alternative radiation sources are likewise possible. For the sake of simplicity, hereinafter mention is nevertheless made in part of plasma sources or of the plasma of the radiation source. This should not be understood to be restrictive. The concept according to the invention is applicable to alternative radiation sources without any problems.

In accordance with one aspect of the invention, the radii of curvature of the second facets are chosen in such a way that for at least two illumination channels dose fluctuations in the object field which are caused by source fluctuations at least partly compensate for one another. It was able to be shown that this is possible by choosing the radii of curvature of the second facets in such a way that one subset of the second facets has an image distance which is less than its distance from the correction plane, and one subset of the second facets has an image distance which is greater than its distance from the correction plane.

The radii of curvature of the second facets are chosen in particular in such a way that in the case of at least one predefined illumination setting the dose fluctuations averaged over all the illumination channels at least partly compensate for one another. In other words, they are chosen in such a way that a reduction, in particular a minimization, of the total dose fluctuation in the object field occurs. The total dose fluctuations in the object field can be reduced in particular to values below a predefined permissible maximum value of 10%, in particular of 5%, in particular of 3%, in particular of 1%.

This can be achieved, as described in even greater detail below, by minimizing a merit function. This involves an optimization problem, in particular a discrete optimization problem. The radii of curvature of the second facets are chosen in particular in such a way that the total dose fluctuation is reduced, in particular minimized, for a predefined selection of illumination settings. It can be ensured, in particular, that the total dose fluctuation in the object field for all of the predefined illumination settings is at most of the same magnitude as a predefined maximum value.

These indications relate in each case to mechanical fluctuations of the radiation source within a predefined maximum fluctuation range. The fluctua- tion range of the radiation source proceeding from a nominal position can be for example up to 1 mm, up to 2 mm or up to 3 mm in each direction.

The fluctuation range of the radiation source can have the same extent in each spatial direction. It can also be anisotropic, that is to say have diffe- rent extents in different spatial directions.

A further object of the invention is to specify a method for designing a pupil facet mirror, in particular a method for determining the radii of curvature of facets of a pupil facet mirror.

This object is achieved by means of a method comprising the following steps:

Predefining an illumination system according to the preceding descrip- tion,

Predefining a number of different illumination settings with illumination channels,

Predefining a discrete number of different radii of curvature of the second facets, Determining a relationship of a displacement of images of the first facets in the region of a correction plane with fluctuations of the radiation source depending on the radii of curvature of the second facets, Determining a field shift averaged over all the illumination channels in the region of and/or relative to the correction plane for the different illumination settings,

Minimizing a merit function, comprising weighted portions of the field shifts on account of fluctuations of the radiation source in different spatial directions, depending on an assignment of the second facets to different radii of curvature .

The method according to the invention makes it possible to determine the radii of curvature of the facets of the pupil facet mirror in such a way that the illumination of the object field is as insensitive as possible with regard to fluctuations of the radiation source. In particular the stability of the illumination of the object field can be improved as a result.

Regarding the different radii of curvature, it is possible firstly merely to predefine how many discrete, different radii of curvature are actually pro- vided. In principle, it is also possible to predefine concrete values for the radii of curvature. In the first case, the concrete values of the radii of curvature are determined in the minimization of the merit function.

In the second case, only the assignment of the individual facets to the al- ready predefined radii of curvature is determined.

When determining the field shift averaged over all the illumination channels, provision can be made for weighting the individual illumination channels of different illumination settings with predefined weighting fac- tors. This makes it possible, in particular, for a selection of illumination channels to be weighted to a greater or lesser extent.

In accordance with one aspect of the invention, the merit function compri- ses in each case portions of the maximum field shifts depending on the different illumination settings, wherein different spatial directions of the field shifts are weighted in each case with a weighting factor.

According to the invention, it has been recognized that the dose stability in the object field can be further improved by a different weighting of the field shifts in the different spatial directions.

In accordance with a further aspect of the invention, when minimizing the merit function it is predefined as a boundary condition that in each case at least 20%, in particular at least 25%, in particular at least 30%, of the second facets have the same radius of curvature. This can be advantageous for the production of the pupil facet mirror.

Further objects of the invention are to improve an optical system for a pro- jection exposure apparatus, and a projection exposure apparatus for micro- lithography.

These objects are also achieved by providing a pupil facet mirror according to the preceding description. The advantages are evident from those of the pupil facet mirror.

Further objects of the invention are to improve a method for producing a microstructured or nanostructured component, and a component produced according to the method. The advantages are evident from those already explained above.

Further advantages, details and features of the invention are evident from the description of exemplary embodiments with reference to the figures. In the figures:

Figure 1 shows a projection exposure apparatus for microlithography schematically and with regard to an illumination optical unit in meridional section,

Figure 2 shows schematically an enlarged excerpt from the beam path of the illumination optical unit for elucidating the invention, and

Figure 3 shows schematically a diagram for clarifying the dependence of the field shift in the correction plane depending on a radius of curvature of the pupil facets for the case where all the pupil facets have the same radius of curvature.

Firstly the general component parts of a projection exposure apparatus 1 are described below. For further details of the projection exposure apparatus 1 , in particular for further details of a correction device 24, also referred to as UNICOM, reference should be made to DE 10 2012 205 886 Al and EP 2 240 830 Bl, which are fully integrated into the present invention as part thereof.

A projection exposure apparatus 1 for microlithography serves for producing a micro structured or nanostructured electronic semiconductor compo- nent. A radiation source 2 emits EUV radiation used for illumination in the wavelength range of, for example, between 5 nm and 30 nm, in particular in the range of 13.5 nm or less. The radiation source 2 can be in particular a plasma source, for example a GDPP (gas discharge produced plasma) source or an LPP (laser produced plasma) source. A radiation source based on a synchrotron can also be used for the radiation source 2. Information concerning such a radiation source can be found by the person skilled in the art in US 6 859 515 B2, for example. EUV illumination light or illumination radiation in the form of an imaging light beam 3 is used for illumi- nation and imaging within the projection exposure apparatus 1. The imaging light beam 3 downstream of the radiation source 2, firstly passes through a collector 4, which can be, for example, a nested collector having a multi-shell construction known from the prior art, or alternatively an el- lipsoidally shaped collector then arranged downstream of the radiation source 2. A corresponding collector is known from EP 1 225 481 A.

Downstream of the collector 4, the EUV illumination light 3 firstly passes through an intermediate focal plane 5, which can be used for separating the imaging light beam 3 from undesirable radiation or particle portions. After passing through the intermediate focal plane 5, the imaging light beam 3 firstly impinges on a field facet mirror 6 comprising field facets 7.

In order to facilitate the description of positional relationships, a Cartesian global xyz-coordinate system is in each case depicted in the drawing. In Figure 1 , the x-axis runs perpendicularly to the plane of the drawing and out of the latter. The y-axis runs towards the right in figure 1. The z-axis runs upwards in Figure 1.

In order to facilitate the description of positional relationships for individual optical components of the projection exposure apparatus 1, a Cartesian local xyz- or xy-coordinate system is in each case also used in the following figures. The respective local xy-coordinates span, unless described otherwise, a respective principal arrangement plane of the optical component, for example a reflection plane. The x-axes of the global xyz- coordinate system and of the local xyz- or xy-coordinate systems run parallel to one another. The respective y-axes of the local xyz- or xy-coordinate systems are at an angle with respect to the y-axis of the global xyz- coordinate system which corresponds to a tilting angle of the respective optical component about the x-axis.

The field facets 7 are rectangular and have in each case the same x/y aspect ratio. The x/y aspect ratio can be for example 12/5, can be 25/4, can be 104/8, can be 20/1 or can be 30/1.

After reflection at the field facet mirror 6, the imaging light beam 3 split into imaging light partial beams assigned to the individual field facets 7 impinges on a pupil facet mirror 10 comprising pupil facets 1 1. The respective imaging light partial beam of the entire imaging light beam 3 is guided along a respective imaging light channel.

In accordance with one exemplary facet arrangement, the pupil facets 1 1 are arranged around a centre in facet rings lying one inside another. A pupil facet 1 1 is assigned to each imaging light partial beam of the EUV illumi- nation light 3 which is reflected by one of the field facets 7, such that a respective facet pair impinged upon and comprising one of the field facets 7 and one of the pupil facets 1 1 predefines the imaging light channel for the associated imaging light partial beam of the EUV illumination light 3. The channel-by-channel assignment of the pupil facets 1 1 to the field facets 7 is effected depending on a desired illumination by the projection exposure apparatus 1. The imaging light channels are also referred to as illumination channels. The totality of the illumination channels defines an illumination setting.

Via the pupil facet mirror 10 and a downstream transfer optical unit 15 consisting three mirrors 12, 13, 14, the field facets 7 are imaged into an object plane 16 of the projection exposure apparatus 1. The mirror 14 is embodied as a mirror for grazing incidence (grazing incidence mirror). Ar- ranged in the object plane 16 is a reticle 17, from which, with the illumination radiation 3, an illumination region is illuminated which coincides with an object field 18 of a downstream projection optical unit 19 of the projection exposure apparatus 1. The illumination region is also referred to as an illumination field. The object field 18 is rectangular or arcuate depending on the concrete embodiment of an illumination optical unit of the projection exposure apparatus 1. The imaging light channels are superimposed in the object field 18. The EUV illumination light 3 is reflected from the reticle 17. The reticle 17 is held by an object holder 17a, which is displaceable in a driven manner along the displacement direction y with the aid of an object displacement drive 17b indicated schematically.

The projection optical unit 19 images the object field 18 in the object plane 16 into an image field 20 in an image plane 21. Arranged in said image plane 21 is a wafer 22 bearing a light-sensitive layer, which is exposed during the projection exposure by means of the projection exposure apparatus 1. The wafer 22, that is to say the substrate onto which imaging is effected, is held by a wafer or substrate holder 22a, which is displaceable along the displacement direction y synchronously with the displacement of the object holder 17a with the aid of a wafer displacement drive 22b likewise indicated schematically. During the projection exposure, both the reticle 17 and the wafer 22 are scanned in a synchronized manner in the y-direction. The projection exposure apparatus 1 is embodied as a scanner. The scanning direction y is the object displacement direction.

An illumination intensity correction device 24 is arranged in a correction plane 23, said illumination intensity correction device being explained in even greater detail below. The correction plane 23 is spaced apart from the object plane 16 by not more than 20 mm, for example by 6 mm, 8 mm, 10 mm or 16 mm. The correction device 24, which is also referred to as UNICOM, serves inter alia for setting a scan-integrated intensity distribution, that is to say intensity distribution integrated in the y-direction, of the illumination light over the object field 18. The correction device 24 is driven by a control unit 25. Examples of a field correction device are known from WO 2009/074 21 1 A 1 , EP 0 952 491 A2 and from

DE 10 2008 013 229 Al, to which reference is hereby made.

The field facet mirror 6, the pupil facet mirror 10, the mirrors 12 to 14 of the transfer optical unit 15 and the correction device 24 are parts of an il- lumination optical unit 26 of the projection exposure apparatus 1. Together with the radiation source 2 the illumination optical unit 26 forms an illumination system of the projection exposure apparatus 1. The illumination optical unit 26 together with the projection optical unit 19 forms an optical system.

During the projection exposure, firstly the reticle 17 and the wafer 22, which bears a coating that is light-sensitive to the illumination light 3, are provided. Afterwards, a section of the reticle 17 is projected onto the wafer 22 with the aid of the projection exposure apparatus 1. Finally, the light- sensitive layer on the wafer 22 that has been exposed with the illumination light 3 is developed. A microstructured or nanostructured component, for example a semiconductor chip, is produced in this way. Further details of the illumination system of the projection exposure apparatus 1, in particular of the pupil facet mirror 10, are described below.

The pupil facets 1 1 of the pupil facet mirror 10 image the field facets 7 of the field facet mirror 6 into the object field 18 with the reticle 17. In this case, the UNICOM 24 serves for setting and correcting intensity field profiles in the object field 18. The correction plane 23, which is also referred to as UNICOM plane, is situated upstream of the object field 18 with the reticle 17 in the direction of the beam path of the illumination radiation 3. Fluctuations of the radiation source 2 can occur during operation of the projection exposure apparatus 1. Said fluctuations are also referred to as source fluctuations. Such source fluctuations lead to a shift in the illumination field in the UNICOM plane, in particular in a direction parallel to the scanning direction, in particular into the UNICOM 24. This can have the effect that a part of the illumination radiation 3 provided for illuminating the reticle 17 does not pass to the reticle 17, but rather is blocked out by the UNICOM 24. Since this is not detected by energy sensors arranged in the beam path upstream of the correction plane 23, such a field shift can lead to dose errors in the exposure of the reticle 17. In particular a shift of the il- lumination field into the UNICOM 24, that is to say in a direction parallel to the scanning direction and parallel to the UNICOM plane 23, is relevant to the dose error. The fluctuations of the radiation source 2 may be attributable to different causes. For the sake of simplicity, they are also referred to as plasma fluctuations. Their extent is in the region of at most 3 mm, in particular at most 2 mm, in particular at most 1 mm.

In order to reduce, in particular to minimize, the dose error caused by the source fluctuations, the invention provides for determining the radii of curvature of the pupil facets 1 1 in the design phase in such a way that the field shift resulting from the totality of all the illumination channels in the cor- rection plane 23 becomes minimal. In this case, it is not necessary to optimize the imaging of each individual illumination channel independently of one another.

As a boundary condition it can be predefined that the pupil facets 1 1 in each case form groups, wherein the pupil facets 1 1 of the same group have the same radius of curvature and pupil facets 1 1 of different groups have different radii of curvature. The number of groups can be two, three, four, five or more. It is preferably at most twenty, in particular at most fifteen, in particular at most ten, in particular at most five.

The determination of the radii of curvature of the pupil facets 1 1 or the assignment of the pupil facets 1 1 to different radii of curvature can be ascertained as follows: Firstly, for each of the spatial directions x, y and z the field shift (i UNI) in a direction parallel to the scanning direction in the correction plane 23 on account of fluctuations of the radiation source 2 in the respective spatial direction (i Plasma_x, i Plasma_y and i Plasma_z) depending on the radii of curvature of the pupil facets 1 1 is determined according to the following formula:

The highest value over a number of predefined illumination settings max

upon averaging over all the illumination channels is defined as x, Setting

the resulting field shift into the UNICOM 24, that is to say in the y- direction. In this case, the individual illumination channels can be weighted with different weights (gi,settin_g). As weights

(gi, setting) use is made, in particular, of the intensities of the respective illumination channels.

A merit function comprises portions of the field shifts which result from source fluctuations in the different spatial directions x, y and z. Said portions can be weighted with weighting factors α, β, γ. For determining the optimum radii of curvature of the pupil facets 1 1 or the assignment of the pupil facets 1 1 to the groups having different radii of curvature, the following merit function is thus minimized: . UNI _ UNI UNI

Ment = a + β + γ

<iPlasma_x <iPlasma_Y <iPlasma_z

This involves a discrete optimization problem.

The invention can provide for predefining one or a plurality of the radii RPF of curvature of the pupil facets 1 1. Said radii RPF of curvature can be predefined as a boundary condition in the optimization in particular already before the minimization of the merit function. This makes it possible, for example, to use already existing pupil facets 1 1. For other reasons, too, it may be desirable to predefine a radius RPF of curvature beforehand at least for individual pupil facets from among the pupil facets 1 1.

Figure 3 illustrates by way of example how the field shift, that is to say the shift of the images of the field facets in the UNICOM plane 23 in the y- direction, is dependent on a radius RPF of curvature of the pupil facets 1 1. The illustration shows a case in which all of the pupil facets 1 1 have the same radius RPF of curvature. The illustration shows separate regions which in each case indicate the effect of a mechanical shift of the radiation source 2 by 1 mm in the direction identified by the symbol. The regions in each case clarify the variation of the respective variable over the illumination settings and field points. As can be gathered qualitatively from Figure 3, the field shift in the case of a source fluctuation in the z-direction is substantially insensitive to the radius RPF of curvature of the pupil facets 1 1. In the case of a source fluctuation in the x-direction, said field shift exhibits a weak dependence on the radius RPF of curvature of the pupil facets 1 1.

In the case of a source fluctuation in the y-direction, however, said field shift exhibits a significant dependence on the radius RPF of curvature of the pupil facets 1 1. As can be seen qualitatively, assuming that all the pupil facets 1 1 have the same radius R of curvature, only a source fluctuation in one direction (y- direction) can be compensated for, while the other two directions are substantially insensitive to a variation of the radius R of curvature and therefore cannot be optimized. These values were able to be considerably improved when the boundary condition that all of the pupil facets 1 1 must have the same radius of curvature was relinquished. In other words, additional radii of curvature of the pupil facets 1 1 enable compensations of additional degrees of freedom of the fluctuations of the radiation source 2.

Restriction to two, three, four or five different radii of curvature has proved to be a particularly advantageous compromise.

In this case, as a boundary condition it can be predefined that in each case a minimum number of in each case at least 100 pupil facets 1 1, in particular at least 200 pupil facets 1 1, in particular at least 300 pupil facets 1 1, in particular at least 400 pupil facets 1 1 , have the same radius of curvature. In particular as a boundary condition when minimizing the merit function it can be predefined that in each case at least 20%, in particular at least 25% of the pupil facets 1 1 have the same radius of curvature.

Given a number of two groups of pupil facets 1 1 having different radii of curvature, it was possible to reduce the maximum sensitivity of the field shift with regard to source fluctuations in the three different spatial directions by a factor of approximately three from approximately 180 μηι/mm to approximately 60 μιη/mm. Even when one of the two radii of curvature was taken over from an existing design of the pupil facet mirror 10 and fixedly predefined, the sensitivity of the field displacement relative to source fluctuations could still be reduced by more than a factor of 2. With addition of a third group of pupil facets 1 1 , that is to say a third radius of curvature, the sensitivity was able to be reduced further. However, the additional benefit with addition of a third radius of curvature was no longer as great as the difference upon the transition from a single radius of curvature to two different radii of curvature. The reduction of the sensitivity upon the transition from two to three radii of curvature was in the range of 10% to 30%.

Different aspects of the pupil facet mirror 10 according to the invention are explained again below with reference to Figure 2. The pupil facet mirror 10 comprises a multiplicity of pupil facets 1 1 having different radii of cur- vature. Figure 2 illustrates by way of example three pupil facets 1 11 , l b, 1 h having three different radii Ri, R₂, R3 of curvature.

The different radii Ri of curvature of the pupil facets 1 1 lead to different focal lengths and hence different image distances or image positions there - of. The pupil facet 1 11 has for example an image distance which indeed corresponds approximately to its distance from the correction plane 23. The pupil facet 1 l₂ has an image distance which is less than its distance from the correction plane 23. The pupil facet 1 h has an image distance which is greater than its distance from the correction plane 23.

The distance between the pupil facets 1 li and the correction plane 23 is measured here in each case along a central ray of the beam comprising illumination radiation 3 impinging on them. At least two of the radii of curvature of the pupil facets 1 1 differ by at least 1%, in particular at least 2%, in particular at least 3%, in particular at least 4%. The radii of curvature of the pupil facets 1 1 preferably differ at most by 10%, in particular at most 9%, in particular at most 8%, in particular at most 7%, in particular at most 6%. Only three of the pupil facets 1 1 are illustrated in Figure 2, for reasons of clarity. In actual fact the number of pupil facets 1 1 of the pupil facet mirror 10 is significantly higher. It can be in particular at least 1000, in particular at least 1200, in particular at least 1400, in particular at least 1600, in particular at least 1800, in particular at least 2000. It is preferably at most 10 000, in particular at most 8000, in particular at most 6000, in particular at most 4000, in particular at most 2000. In each case at least 100 pupil facets 1 1, in particular at least in each case 100 pupil facets 1 1, in particular at least in each case 300 pupil facets 1 1, in particular at least in each case 400 pupil facets 1 1 have the same radius of curvature. The pupil facet mirror 10 comprises one, two, three, four, five or more groups of pupil facets 1 1, wherein pupil facets 1 1 of the different groups have different radii Ri of curvature. Pupil facets 1 1 of the same group have the same radius of curvature.

Claims

Patent Claims:

Pupil facet mirror (10) for an illumination optical unit (26) of a projection exposure apparatus (1) comprising facets (1 1),

1.1. wherein the facets (1 1) form a plurality of different groups,

1.1.1. wherein the facets (1 1) of the different groups have different radii of curvature, and

1.1.2. wherein the facets (1 1) of the same group have in each case the same radius of curvature.

Pupil facet mirror (10) according to Claim 1, characterized in that at least two radii of curvature of the facets (1 1) differ by at least 1%.

Pupil facet mirror (10) according to either of the preceding claims, characterized in that at least in each case 100 facets (1 1) have the same radius of curvature.

Pupil facet mirror (10) according to any of the preceding claims, characterized in that the number of groups of facets (1 1) having different radii of curvature is at least two.

Illumination optical unit (26) for a projection exposure apparatus (1) for microlithography comprising

5.1. a first facet mirror (6) comprising a plurality of first facets (7), and

5.2. a second facet mirror comprising a plurality of second facets (1 1),

5.3. wherein the second facet mirror is formed by a pupil facet mirror (10) according to any of the preceding claims. Illumination optical unit (26) according to Claim 5, characterized in that it has a device (24) for correcting an illumination intensity distribution in a correction plane (23), and in that the radii of curvature of the second facets (1 1) are chosen in such a way that a subset of the second facets (1 1) has an image distance which is less than its distance from the correction plane (23).

Illumination system for a projection exposure apparatus (1) comprising

7.1. an illumination optical unit (26) according to either of Claims 5 and 6, and

7.2. a radiation source (2) for generating illumination radiation (3).

Illumination system according to Claim 7, characterized in that the radii of curvature of the second facets (1 1) are chosen in such a way that for at least two illumination channels dose fluctuations in the object field (18) which are caused by source fluctuations at least partly compensate for one another.

Method for determining the radii of curvature of facets (1 1) of a pupil facet mirror (10) comprising the following steps:

9.1. predefining an illumination system according to either of Claims 7 and 8,

9.2. predefining a number of different illumination settings with illumination channels,

9.3. predefining a discrete number of different radii of curvature of the second facets (1 1),

9.4. determining a relationship of a displacement of images of the first facets (7) in the region of a correction plane (23) with fluctuations of the radiation source (2) depending on the radii of curvature of the second facets (1 1),

9.5. determining a field shift averaged over all the illumination channels in the region of and/or relative to the correction plane for the different illumination settings,

9.6. minimizing a merit function, comprising weighted portions of the field shifts on account of fluctuations of the radiation source (2) in different spatial directions, depending on an assignment of the second facets (1 1) to different radii of curvature.

10. Method according to Claim 9, characterized in that the merit function comprises in each case portions of the maximum field shifts depending on the different illumination settings, wherein the field shifts in different spatial directions are weighted in each case with an inde- pendent weighting factor.

1 1. Method according to either of Claims 9 and 10, characterized in that as a boundary condition in the minimization of the merit function it is predefined that in each case at least 20% of the second facets (1 1) have the same radius of curvature.

12. Optical system comprising an illumination optical unit (26) according to either of Claims 5 and 6 and

12.1. a projection optical unit (19) for imaging an object field (18) into an image field (20).

13. Projection exposure apparatus (1) for microlithography, comprising 13.1. an illumination optical unit (26) according to either of

Claims 5 and 6,

13.2. a projection optical unit (19) for imaging an object field (18) into an image field (20), and

13.3. a radiation source (2) for generating illumination radiation (3). 14. Method for producing a microstmctured or nanostmctured component comprising the following steps:

providing a wafer (22), to which a layer made of a light-sensitive material is at least partly applied,

providing a reticle (17) which has stmctures to be imaged,

- providing a projection exposure apparatus (1) according to

Claim 13,

projecting at least one part of the reticle (17) onto a region of the layer of the wafer (22) with the aid of the projection exposure apparatus (1).

15. Component produced according to a method according to Claim 14.