OPTICAL ELEMENT WITH MULTIPLE PRIMARY LIGHT SOURCES

Field of the Invention

The invention relates to an illumination system having an optical element with multiple primary light sources. The illumination system can be used in microlithography. The invention further relates to a microlithography projection system.

Description of the State of the Art

In order to be able to reduce the structure widths in electronic components still further, particularly into the submicron range, it is necessary to reduce the wavelength of the light that is being used for the microlithography process. A conceivable solution is offered for example by lithography processes with wavelengths < 193 nm, for example using soft X-rays, a concept which is referred to as EUV lithography.

An illumination system that is suitable for lithography with wavelengths smaller than 193 nm, particularly wavelengths in the EUV range, should use as few reflections as possible to form a field that is defined for the lithographic process, corresponding to the object field of a projection objective, and to illuminate this field as uniformly as possible. Furthermore, the exit pupil of the illumination system, which coincides with the entry pupil of the projection objective which follows next in the light path, should be illuminated for each field point and filled with light as uniformly as possible while maintaining fixed pupil positions. In contrast to VUV systems for 248 nm and 193 nm, mirrors of low reflectivity in the range from 60 to 70 percent have to be used in systems that operate with wavelengths smaller than 193 nm, particularly in systems that operate with soft

X-rays. Consequently, to achieve a very high effectiveness in such systems requires that the object field and the entry pupil can be formed and uniformly illuminated, i.e. filled with light, with as few reflections as possible.

A first embodiment of an illumination system which is suitable for lithography with soft X-rays, i.e. for EUV lithography, has been disclosed for example in US 6,438,199, wherein the light of an EUV light source, for example the light of a plasma light source, which can be produced through gas discharge as well as by means of laser excitation, is collected by means of a collector. The collected light is received by a double-facetted optical system consisting of a first optical element which contains multiple first raster elements, so-called field facets, and a second facetted element arranged downstream in the light path, which contains multiple second raster element, so-called pupil facets. By way of the double-facetted element, the field as well as the pupil are pre-shaped out of the light received by the collector from the light source. In the embodiment disclosed in US 6,438, 199, the shaping of the field into a ring field segment occurs with the so-called field mirror which is arranged in the light path downstream of the facetted optical elements.

As an alternative to this concept, the field facets can already have the shape of the field. Reference is made in this regard to US 6,861 ,656.

An illumination system with a optical element comprising field facets and a plurality of light sources is disclosed in US 6,570,168 as well as in US 2006/0119824.

Also WO2007/002170 or WO2006/074812 shows a illumination system and a projection exposure apparatus with a multiple of light sources.

An illumination system that is intended in particular for EUV lithography has been disclosed in US 2006/0132747.

Description of the Invention

The invention provides for a highly efficient illumination system for microlithography with soft X-rays.

The illumination system according to the invention comprises an optical element having multiple primary light sources. With the illumination system a field in a field plane having a field contour is illuminated. The illumination system is configured in such a way, that each of the primary light sources only illuminates an area in said field plane having a dimension smaller than the field contour. This means that the size of the area, which is illuminated by each of the light sources is smaller than the area, which is encircled by the field contour in the field plane. If the area in the field plane which is encircled by the field contour is for example rectangular and has a size or a dimension of 26 mm x 2 mm = 52 mm ² and for example the illuminated area of the light source in the field plane is a circle with a radius r = 0,5 mm, then the size of the illuminated area by the primary light source in the field plane is ττr ² = 0,785 mm ² , which is about 1/60 of the size of the field encircled by the field contour in the field plane

The area, which is encircled by the field contour is also denoted as a field which is illuminated in the field plane. The form of the field can be of any shape, for example rectangular as described above or alternatively arc-shaped.

Also the form of the area illuminated by each of the light sources can be of any shape. For example the area can be circular. Alternatively the area can be of elliptical shape.

The area illuminated by different light sources of the multiple of light sources in the field plane can also be of different size. For example one light source of the multiple of light sources can illuminate a circular area having a first size. Another light source of the multiple of light sources can illuminate an elliptical area having a second size, which is different from the first size.

The size of the area illuminated by one of the light sources in the field plane is smaller than the size of the field defined by the field contour. For example the size of the area illuminated by each of the light sources in the field plane is smaller than 1/10, alternatively smaller than 1 /50 of the size of the field plane. Alternatively the size of the area illuminated by one of the multiple of light sources can be smaller than 1/100 or 1/1000 of the size of the field defined by the field contour.

In the optical element with multiple primary light sources, the arrangement of the primary light sources and the respective radiation output characteristics of each light source can be such that each of the multitude of primary light sources illuminates an area in the field plane of the illumination system directly or with the help of further optical components, whereby through the combined areas of each light source in the field plane, the field, which is also referred to as object field, is illuminated in a specific shape, for example a segment of a ring.

According to the invention by using the light of multiple primary light sources with a small number of reflections of two or fewer a field in a field plane with a specific field contour can be illuminated. Furthermore a pupil illumination can for example be set up by simply switching on or switching off specific primary light sources of the optical element with multiple primary light sources.

It can be envisioned that the radiation output characteristic, meaning in particular the angle of deflection of the light rays emitted by the primary light source, is set by means of reflective optical components, in particular reflective surfaces, which are assigned to each primary light source. This means that one reflective surface, i.e., one micro-collector, is assigned to each light source of the multi-source array. Depending on how the reflective surface is configured, specifically how it is positioned relative to the primary light source, different discrete areas in the field plane can be illuminated.

The aforementioned multi-source array with multiple primary light sources can have an arrangement of the primary light sources in regular columns and rows. The cross-correlation between one primary light source and one reflective area, specifically a collector, can also be realized in a fixed unit, which is referred to as a light source unit.

However, it is also possible that the multiple light sources are held for example in columns and rows in a first carrier element and the respective reflective areas or collectors assigned to the light sources are held in a second carrier element. By shifting, specifically tilting, the first carrier element with multiple light sources and/or the second carrier element with multiple collectors, it is possible to change the direction of the light bundles which originate from a current combination of primary light sources and collectors, and thereby to illuminate specific areas in the field plane. This makes it possible for example to influence the uniformity of the illumination in the field plane.

However, it is also possible by tilting or shifting the first carrier element with a multiple of light sources and/or the second carrier element with a multiple of collectors relative to each other to influence the pupil illumination in the exit pupil of the illumination system. By tilting and shifting the direction of the light bundle originating from a combination of a light source and a collector can be changed. As an alternative to shifting or tilting, it is also possible that a larger number of light sources are present than are required for a specific pupil illumination. It is preferred in such a case if at least a part of the multiple primary light sources are responsive to direct control to allow the setting of different pupil illuminations, also referred to as pupil settings, for example a dipolar, quadrupolar or circular illumination or an annular illumination. In this way, if a sufficient quantity of primary light sources is available in the multi-source array, the setting can be selected simply by turning the respective primary light sources on or off which are assigned to the selected setting. Another possibility for selecting the setting presents itself, if the adjustment of the setting is not made by switching the primary light sources on and off, but by tilting or shifting of the reflective areas or collectors

assigned to the respective primary light source unit. The tilting or shifting allows not only the shape of the illumination, i.e. the setting, to be influenced, but also for example the telecenthcity of the pupil illumination. The so-called telecentricity error of the pupil illumination means the deviation of the energy-weighted central ray from the geometric principal ray in a field point. The geometric principal ray in a field point in the object plane of the projection objective is defined by the line connecting the field point to the center of the entry pupil of the projection objective. It is of course possible to combine the switching-on and switching-off of specific primary light sources with the shifting or tilting.

In an embodiment of the primary light sources, the latter can be configured for example as semiconductor laser diodes. Such laser diodes (LEDs) emit light radiation of a wavelength of for example 400 nm. Laser diodes for use in microlithography have been disclosed for example in US 6,233,039.

For an illumination with light of a usable wavelength smaller than 100 nm, for example EUV radiation with wavelengths in the range from 8 to 20 nm, preferably 13.5 nm, so-called pinch plasma sources can be used as primary light sources. Sources of this type have been shown in WO 2007/002170.

The primary light sources can emit a usable radiation in the range of wavelengths shorter than or equal to 100 nm, particularly in the range of so-called EUV radiation with wavelengths between 5 nm and 20 nm.

For the formation of multiple images of a primary light source in a plane, the reflective surface can have positive optical power and therefore a collecting effect. The size of the illumination spots or the images of the primary light sources in case of a critical illumination in the plane in which the field is formed is small in relation to the field to be illuminated. For the special case of a critical illumination the light source images are formed in a field plane of the illumination system.

In the field plane, then the so-called illumination spots or light source images are superimposed on each other, which produce a ring-shaped illumination of the

desired area. By changing the orientation of the reflective surfaces with optical power in relation to the respective light sources from which the light is received, it is possible to illuminate different areas in the field plane by changing the position of the reflective surfaces in relation to the primary light sources.

In order to illuminate different areas in the field plane by the images of the primary light sources it is possible to change the position of the reflective surfaces in relation to the primary light sources by moving and/or tilting the light source. Alternatively the reflective surface could be moved and/or tilted. In a further alternative the reflective surface as well as the light source could be moved and/or tilted.

Besides different areas in the field plane, an area of the exit pupil of the illumination system is also being illuminated by the light of the primary light source. In the exit pupil also different areas for different light sources are illuminated. The area illuminated by a light source in the exit pupil is smaller than the area encircled by an exit pupil contour.

The exit pupil of the illumination system is defined by the crossing point of the chief-rays associated to each field point in the field plane. In a projection exposure system the exit pupil of the illumination system corresponds to the entrance pupil of a subsequent projection objective. If the projection objective is a rotational symmetric system around an axis of rotation then the crossing point of the chief ray to the central field point with the optical axis of the illumination system determines the exit pupil plane. The exit pupil plane in such a system is the plane on which the optical axis of the projection objective stands perpendicular. For the special case of a Koehler illumination images of the light sources are formed in the exit pupil plane.

The individual primary light sources and the reflective surfaces, specifically the micro-collectors, assigned to them are arranged so that substantially a ring field segment is being illuminated in the field plane. With an appropriate arrangement

of the primary light sources, the illumination of the exit pupil can take on any desired shape, preferably a circular or a bipolar or quadrupolar outside contour, i.e. a circular or a bipolar or quadrupolar setting.

It is possible with the present invention to change the illumination of the pupil, i.e. the setting in the pupil, without the need to mechanically change the positions of reflective surfaces or collector surfaces. This is possible for example by providing appropriate light sources for each setting. To change the setting only requires switching individual primary light sources on or off as needed for the desired setting.

The invention thus shows a way to achieve a significantly faster change of the pupil illumination.

In an alternative an illumination system with an optical element having a plurality of facets, to each facet a light source of the light source array could be associated. For example in case of an illumination system shown in US 2006/0132747 each individual facet of the facetted optical element could be illuminated by separate light source.

The light source array contains between 100 and 100,000 primary light sources set up in a regular arrangement, which provide the radiation, specifically in the EUV range of wavelengths. Sources of this kind have a source size in the range from 0.1 μm to 40,000 μm and are complemented by an array of reflective surfaces, so-called micro-collectors, which can be produced for example through microstructure techniques and which are individually assigned to each primary light source.

As the primary light sources are set up in a regular arrangement, but on the other hand the aim is to achieve a ring-shaped illumination in the field plane, one form of the invention may include provisions whereby the light sources which are arranged

in a regular array are controlled in such a way that a kidney-shaped outside contour is obtained as a result.

In an embodiment an illumination system is realized whereby the field in the field plane as well as the exit pupil are illuminated by the optical element itself with multiple primary light sources, i.e. a so-called multi-source array, there is no reflection on a facetted mirror. The efficiency of such a system can be increased by 30 to 40 percent in comparison to a system having a facetted miror. A system of this kind now has for example only a single reflective optical element. Such a single optical element could be a collector. A reflector surface of a collector is assigned to a respective primary light source for light traveling from the light source array to the field plane of the illumination system. It is possible to design collectors with more than one reflection. An example of this is a collector which is designed as a so-called Wolter collector. In a collector of this type, two reflections occur at the reflector surfaces of the collector.

An illumination system with a multi-source array and with collectors that are assigned to the primary light sources can have no further optical component arranged in the light path from the optical element with multiple primary light sources to a field plane of the illumination system.

If the optical element with multiple primary light sources is laid out as a multi- source array with a regular arrangement of the primary light sources in columns and rows, it can be envisioned, as an alternative to the switching-off of light sources, to shape the ring field by way of an optical component that is arranged in the light path between the primary light source and the field plane. This field- shaping component can for example be a grazing-incidence mirror, preferably a convex grazing-incidence mirror.

As an alternative possibility, there can be an image-producing element in the light path from the optical element to the field plane in addition to, or instead of, the field-shaping element. This could be the case if the light path has to be folded

because of spatial considerations, or if aperture stops, for example field aperture stops, are to be used.

An illumination system with multiple primary light sources is used in a projection exposure apparatus, specifically for microlithography. In a projection exposure apparatus of this type, a mask is arranged in the area of the illuminated field in the field plane, and by means of a projection objective an image of the mask is formed on a light-sensitive carrier substrate. The preferred application for projection exposure apparatus of this type is in the manufacturing process for microelectronic components, specifically semiconductor chips. The projection objective whereby an image of the mask in the illuminated area of the field plane is projected into the image plane is preferably an objective which has at least six mirrors. As an alternative, the apparatus can have a projection objective with at least eight mirrors.

The invention will now be described through examples with references to the drawings, wherein:

Figures 1A -1 B represent the illumination of a field in a field plane

Figures 1C-1 D represent the illumination of an exit pupil in an exit pupil plane

Figure 1 E represents a circular exit pupil of the illumination system with cells of a grid;

Figure 2 represents illuminated areas in a field plane of the illumination system;

Figure 3 represents the grid structure for multiple primary light sources in the plane in which the optical element with multiple light sources is arranged;

Figure 4 represents the grid structure for multiple primary light sources in the plane in which the optical element with multiple light sources is arranged, subsequent to transformation into a grid which provides an equal radiation intensity of each illuminated area in the exit pupil;

Figure 5A A shows a first embodiment of an optical element with multiple primary light sources;

Figures 5B-5F show different arrangements relative to each other of the primary light sources and of the collectors assigned to the primary light sources;

Figure 5G shows a light source array with a regular arrangement

Figure 6 represents a embodiment of an optical element with multiple primary light sources in an illumination system with the light source images lying outside the field plane;

Figure 7A-7B shows different pupil illuminations when switching on and off different light sources;

Figure 8 shows an illumination system wherein no further optical element is arranged in the light path from the optical element with multiple primary light sources the field plane, and wherein the element with multiple primary light sources, the so-called multi-source array is not arranged in a pupil plane nor in a field plane nor in a plane that is conjugate to the field plane or to the pupil plane;

Figure 9 shows a projection exposure apparatus with an illumination system which has no further optical element between the

optical element with multiple primary light sources and the field plane, and with an illumination, wherein the images of the light sources are projected substantially into the field plane;

Figure 10 represents an illumination system which is configured as a grazing-incidence mirror in the light path from the optical element with multiple primary light sources to the field plane;

Figure 11 represent a projection exposure apparatus with an illumination system according to Figure 10;

Figure 12 represents an illumination system with an image-producing optical element in the light path from the optical element with multiple primary light sources to the field plane;

Figure 13 shows a projection exposure apparatus with an illumination having image producing optical elements

Figure 14 serves to illustrate the concepts of positive and negative back focus;

Figure 15 illustrates a projection exposure apparatus with negative back focus and with no optical element in the light path from the optical element with multiple light sources to the field plane;

Figure 16 illustrates a projection exposure apparatus with negative back focus and with a grazing-incidence mirror as the only optical element in the light path from the optical element with multiple light sources to the field plane;

Figure 17 illustrates a projection exposure apparatus with negative back focus and with a grazing-incidence mirror as the only optical

element in the light path from the optical element with multiple light sources to the field plane, wherein the optical element with multiple light sources is arranged in an intermediate image plane that is neither a pupil plane nor a field plane nor a conjugate plane relative to the pupil plane, nor a conjugate plane relative to the field plane; and

Figures 18A represent an embodiment of an optical element with multiple and 18B light sources.

With the illumination system according to the invention comprising an optical element with multiple light sources, the different primary light sources illuminate different areas in the field plane of an illumination system. Different illumination systems and microlithography apparatus comprising such an illumination system with an optical element having multiple light sources according to the invention are shown in figures 8 to 17.

In figures 1A to 1 B are described the difference of the illumination of the field plane by systems having fly eye lenses and the invention. In Figure 1C and 1 D are shown the difference of the illumination of the exit pupil according to a system with a Koehler illumination and of an inventive system.

In figure 1A the illumination in a field plane of an illumination system with two optical elements with raster elements, so called reflective fly-eye-lenses according for example to US 6,438, 139 or WO 2007/002170 is shown. Each of the raster elements 2000.1 , 2000.2, 2000.3, 2004.4, 2000.5 of a first optical element 2002 with raster elements is projected in the field plane. In figure 1 A the projection of the raster elements 2000.4 and 2000.5 in the field plane is shown. In the field plane 2004 which comprises a coordinate system in x- and y-direction the images of each of the raster elements substantially overlap. In figure 1 A only the illuminated areas 2006.4, 2000.5 of raster elements 2000.4 and 2000.5 are shown. By superposition of the images of all raster elements of the first reflective fly eye

lens a substantially homogenous illumination in the field plane is achieved. When imaging the raster elements in the field plane an arcuate shape of the illuminated area is achieved, e.g. by a field forming mirror. The illumianted area is also denoted as field, which is illumianted in the field plane. In figure 1A also shown is the central field point 2020 of the field, which is denoted with x-/y-coordiante (0,0).

In figure 1 B the illumination in the field plane according to a system of the invention is shown. In figure 1 B two light sources 2100.1 , 2100.2 out of the plurality of light sources of the optical element with a plurality of light sources are shown. Each of the light sources 2100.1 , 2100.2 is associated to a collector 2102.1 , 2102.2 forming a light source unit 2104.1 , 2104.2. According to the invention each of the light sources 2102.1 , 2102.2 illuminate a different area 2106.1 , 2106.2 in a field plane of an illumination system according to the invention. In case of a critical illumination images of each light source come to lie in the field plane. In case the focal points of the light sources and therefore their images come to lie in a plane before or after the field plane of the illumination system the illuminated areas in the field plane are not images of the light source and therefore denoted as illumination spots. It is now assumed that the images of the light sources lie outside the field plane. As shown in figure 1 B each of the illumination spots 2106.1 , 2106.2 come to lie in different areas in the field plane 2108. The total illumination in the field plane is then a superposition of the illumination spots 2106.1 , 2106.2. In a preferred embodiment of the invention the superposition leads to an illumination of a field 2110 with an arc-shaped field contour 2112. As is apparent the size of the area illuminated in the field plane by the images of the light source or the illumination spots 2106.1 , 2106.2 is smaller than the size of an area encircled by a field contour in the field plane. Also shown in figure 1 B is the central field point 2120 of the field 2110 and a x-/y-coordinate system.

The size of the illumination spots is denoted by SP(1) for the illumination spot 2106.1 and SP(2) for illumination spot 2106.2. The size of an illumination spot can be calculated by an area integral

W dxdy = SP(1) or area(1)

JJ dxdy = SP(2) area(2)

for each of the illumination spots 2106.1 , 2106.2 in the field plane.

Also the size SF of the area encircled by the field contour in the field plane can be calculated by an area integral.

W dxdy = SF area(field)

The size SP(1), SP(2) of an area illuminated by each illumination spot 2601.1 , 2106.2 in the field plane is smaller than the area SF of the field encircled by the field contour, thus SP(1) < SF and SP(2) < SF holds. The size SP(1), SP(2) could be different for different illumination spots. In an embodiment of the invention the size SP(1), SP(2) is smaller than 1/10 SF, alternatively smaller than 1/50 SF, in a further alternative smaller than 1/100 SF, in even a further alternative smaller than 1/1000 SF. The illumination of the field in the field plane encircled by the field contour is the sum of the intensities of all illumination spots lying within the field contour2112.

In figure 1C the illumination in an exit pupil of a illumination system having a Koehler illumination is shown. In a double facetted illumination system a light bundle of a primary light source is divided into a plurality of light bundles by field raster elements of a first reflective fly-eye-lens. Each field raster element provides for a secondary light source. The position of each secondary light source associated to each field raster element is determined by the shape, the orientation and the optical power of the associated field raster element. In figure 1C five raster elements 2200.1 , 2200.2, 2200.3, 2200.4 and 2200.5 and the associated five secondary light sources 2220.1 , 2220.2, 2220.3, 2220.4 and 2220.5 are shown. The secondary light sources are then projected by

further optical elements, such as pupil raster elements into an exit pupil plane. In a illumination system with a Koehler illumination the images of the of the secondary light sources are lying in the exit pupil plane. According to the position of the images of the secondary light sources, which are also denoted as tertiary light sources the exit pupil is illuminated. The exit pupil can then have an illumination for example with a circular exit pupil contour or short spoken a circular illumination. Also a annular or a multipolar illumination would be possible.

In figure 1 D the illumination in an exit pupil of a system not having a Koehler illumination with multiple primary light sources is shown. According to the invention the bundles originating from the primary light source 2300.1 , 2300.2, 2300.3, 2300.4 itself illuminate associated areas or illumination spots 2320.1 , 2320.2, 2320.3, 2320.4 in the exit pupil plane 2340. The position of the illumination spots 2320.1 , 2320.2, 2320.3, 2320.4, originating from the primary light source determine the illumination in the in the exit pupil 2340. Therefore the illumination in the exit pupil plane can be determined by the position of each primary light source and eventually the associated collector to each primary light source in a system with multiple primary light sources. In figure 1 D the field plane is denoted with reference number 2350. In figure 1 D one field point of the field illuminated is shown. The field point shown in figure 1 D is the central field point 2360, from which a x-/y-coordinate system originates.

Generally speaking in a system which has not a Koehler illumination nor a critical illumination, the primary sources are imaged in a plane which does not coincide with neither the pupil plane nor the field plane. In this way the spot size on the reticle as well as the homogeneity of the pupil can be controlled.

As described before the illuminated area in the field plane can be defined by the distribution of the emission angles of the individual primary light sources according to the invention. By an individual tilting of the respective optical axes of the light bundles emitted by the individual primary light sources, it is possible with each light bundle to illuminate different areas in the field plane. The primary light

sources are designed so that when all areas illuminated by the primary light sources are summed up, the resulting total illumination in the field plane will have a predetermined shape, for example the shape of a ring field segment. This is shown in figure 1 B. Furthermore, the arrangement of the primary light sources is selected so that not only a field in a field plane is illuminated in a predetermined shape, but also an exit pupil, for example in an annular, bipolar or quadrupolar shape as shown in figure 1 D.

In order to achieve the desired ring-shaped illumination in the field plane and the desired illumination of the exit pupil, a very complex arrangement of the primary light sources in relation to each other is required. It is possible that geometrically adjacent primary light sources illuminate image points that are far apart from each other in the exit pupil. It is possible to calculate the arrangement of the light sources as well as their distribution of emission angles by way of a so-called grid transformation. In a grid transformation, every pupil that belongs to a field point is represented by a specific grid.

The grid transformation should now be described in detail.

A grid is shown in Figure 1 E for a circular illuminated exit pupil. The exit pupil 1 is subdivided into a grid 20 with cells 22 of equal size. The equally sized cells 22 have equal power densities, i.e. equal radiation intensities. In accordance with the dimensions of the primary light sources, and taking the telecentricity requirements for the pupil illumination in the exit pupil into account, a number of discrete field points are selected with which a homogeneous illumination of the exit pupil is achieved.

These discrete field points 23 are shown in Figure 2 together with the border of the area 24 that is illuminated by a primary light source. In the case of a critical illumination, the area 24 coincides substantially with the images of the respective primary light sources. In the case the images of the primary light sources are not projected in the field plane, the area 24 corresponds to an illumination spot caused

by each of the primary light sources in the field plane. Also shown is the field contour 5 of the illuminated field 7. The illuminated field in a field plane is obtained by adding up the individual illuminated areas 24 of the primary light sources, e.g. or the illumination spots caused by the primary light sources in the field plane.

The grid of the respective exit pupil according to Figure 1 E is traced back through each area 24 of an illumination of a primary light source in the field plane into the plane in which the primary light sources are arranged. By the superposition of the grids and the respective pupils that belong to different images of light sources, a complicated network is produced at the location of the primary light sources.

This network is referred to as the facet network 26. A facet network 26 of this kind is shown in Figure 3. A transformation network is now calculated for the plane in which the multiple primary light sources are aranged, wherein as a boundary condition the requirement for equal radiation intensity of each cell in the exit pupil is imposed. The facet network is overlaid on the transformation network, and the facet network is transformed in such a way that the transformation network becomes a square grid, i.e. evenly spaced and orthogonal. A facet is now drawn around each grid point of the transformed facet network. Next, the facet network is transformed back onto the transformation network. As a result, facets of different sizes and of a different position are obtained which is defined by an associated field point. The angle of inclination of the individual primary light sources that are associated with each facet is now adjusted in such a way that the ray emitted by the primary light source into the respective facet point is steered from the center of the primary light source in the direction of the associated field point.

Figure 4 shows an example for the transformation network that is obtained under the boundary condition of equal radiation intensity for each cell. A multitude of rectangular areas 27 are shown, which are respectively associated with a primary light source, or in which a respective primary light source is arranged. By controlling the individual angles of inclination of the respective light rays emitted by the primary light sources, a multitude of field points and pupil locations can be

illuminated. In Figure 4, the field points 29 are identified which are projected back into the plane of the optical element with multiple light sources. Also indicated is the contour or field shape 30 of the illuminated field 33 in the field plane. Specially marked in Figure 4 are the primary light sources 31 , which in essence illuminate the dark-filled area 24 in the field plane. The illumination spots or illumination points of the primary light source 31 are in essence super imposed in the dark- filled area 24 of the field plane. As can be seen clearly, the dimension or size of the area 24 is significantly smaller than the dimension of the field to be illuminated which is defined by the contour 30. Under the concept of the invention, the field contour 30 of the field is illuminated by the summation of a multitude of illumination spots 24 as described with reference to figure 1 B.

Although the description of figures 1 E to 4 is made for a critical illumination the description is not limited thereto.

The optimization can be carried out not only with a view towards the most uniform possible illumination of the exit pupil but also taking into account the uniformity of the field plane. Accordingly, the primary light sources are to be placed differently.

For an improved illumination of the exit pupil of the illumination system, the arrangement can also be selected in such a way that for any integration of the portions of the illumination in the field plane which lie on a straight line that runs in the scan direction, the same integral value for the partial illumination of the exit pupil is obtained.

Figures 5A to 7 show embodiments of an optical element with multiple primary light sources in a regular arrangement as it is used in an illumination system according to the invention. Especially in the figures 5A to 7 is shown how the light beams which are emitted form each of the primary light sources of the multiple primary light sources can be directed to different positions in a field, such that different areas in a field are illuminated.

Figure 5A represents such a component with multiple primary light sources. The component consists of two parts, one of which is a light-element part 40 which contains the multiple primary light sources 42. As can be seen in Figure 5A, these primary light sources 42 are constituted for example by discharge light sources. The primary discharge light sources are arranged in a regular array of columns and rows. Possible discharge light sources include Ne (neon) or N ₂ /O ₂ discharge light sources. Discharge light sources with other inert gases such as helium or argon are likewise conceivable and possible. The diameter of an individual light source is approximately 200 μm. As another gas filling, a Xenon (Xe) filling could be used. A discharge source with Xe filling can generate light pulses with an EUV wavelength of 13.5 nm. The power output of this kind of primary light source is in the range of 5 to 200 watt. The diameters of the primary light sources are in the range from 0.1 to 40000 μm alternatively from 100 to 400 μm. The voltages at which the light sources are operated are in the range from 100 V to 3000 V, preferably 600 to 1000 V. With dimension of 200 μm for a primary light source, an array with dimensions in the range of a few centimetres can accommodate several hundred to several thousand primary EUV light sources.

The primary light sources which emit EUV radiation can be for example hollow cathode light sources, capillary light sources, or the z-pinch plasma sources described above. Also possible are laser plasma sources.

The light rays of the light source are released through a hole 50 in a hollow cathode 52. Based on the dimensions of the hollow cathode 52, it is envisioned in light sources of this type to collect the light by means of reflective surfaces 54 which are arranged in the light path from the light source to the field plane that is to be illuminated. The reflective surfaces 54 are also referred to as collector surfaces. One collector surface is assigned to each light source. A correlation of this kind is advantageous but by no means an absolute requirement.

The reflectors 54 shown here are arranged with rotary symmetry about an optical axis HA. The collectors reflect the light at a very shallow angle, i.e. with grazing

incidence. If a rotary paraboloid is selected as a collector surface, a substantially parallel alignment of the light rays is obtained downstream of the collector associated with the respective primary light source.

However, according to the invention a collector with a convergent effect on the light beams is preferred, for example a collector with a surface in the shape of a rotary ellipsoid. Combined surfaces, for example combinations of rotary ellipsoids and rotary hyperboloids, are likewise possible. As a result of the convergent effect of the rotationally symmetric collector surface, the multiple primary light sources are projected into multiple light source images. In an embodiment of the invention, these light source images lie in a plane which is neither a field plane of the illumination system nor a exit pupil plane of the illumination system nor a plane that is conjugate to the field plane or conjugate to the exit pupil plane. In such a case in the field plane only illumination spots as shown in figure 1 B come to lie, whereas the images of the primary light sources are situated in a plane in the light path before or after the field plane. Nevertheless the position of the images of the primary light sources in a plane influence the position of the illumination spots in the field plane. If the illumination system is part of a projection exposure apparatus with a projection objective, the field plane of the illumination system coincides with the object plane of the projection objective, and the exit pupil plane of the illumination system coincides with the entrance pupil plane of the projection objective. Arranged in the field plane of the projection objective is the object of which an image is to be projected, for example a structured mask, the so-called reticle into a image plane of the projection objective. Arranged in the pupil plane is for example the aperture stop of the projection objective. To each field point can be assigned an energy-weighted central ray representing the center of gravity of the bundle of rays originating from this field point. In an ideal case the complete set of all these energy-weighted central rays intersects in the pupil plane 85 or 87 resp., which coincides or is conjugated to the entrance pupil of the subsequent projection optics. The energy-weighted central ray is also denoted as chief ray.

The position of the light source images or the illumination spots in the field plane can generally be controlled, for example with uniformly aligned primary light sources, by adjusting the position, i.e. tilting and/or moving of the grazing- incidence collectors associated with the respective primary light sources. The tilting of the collectors relative to the light sources is illustrated in Figure 5B. What has been said in the context of Figure 5A regarding the possibility for position adjustment, which can include tilting and/or shifting, applies in principle for all of the embodiments illustrated in Figures 5B to 5F.

Associated with each of the primary light sources 42.1 , 42.2, 42.3 is an axis z1 , z2, z3, respectively, which represents the symmetry axis of the respective hollow cathode 52.1 , 52.2, 52.3.

The symmetry axes z1 , z2, z3 of the hollow cathodes 52.1 , 52.2, 52.3 in the illustrated example are oriented substantially parallel to each other. The individual grazing-incidence collectors 54.1 , 54.2, 54.3 assigned to the primary light sources

42.1 , 42.2, 42.3, likewise, each have a symmetry axis HA1 , HA2, HA3. The shells of the grazing-incidence collectors are surfaces running in a closed loop with rotational symmetry relative to the respective symmetry axis HA1 , HA2, HA2, i.e. so-called rotational shells. The direction of the light bundle reflected by a collector substantially coincides with the direction of the symmetry axis HA1 , HA2, HA3 of the respective collector. Thus, the direction of the reflected light ray bundle 57.1 ,

57.2, 57.3 can be set and adjusted by changing the direction of the symmetry axis HA1 , HA2, HA3 of the respective collector. Accordingly, the position of the light sources images or the illumination spots in a field plane can be adjusted by tilting of the respective grazing-incidence collectors relative to the primary light sources.

Conversely, the collectors can also take on a position where the symmetry axes HA1 , HA2, HA2 are parallel, and the symmetry axes z1 , z2, z3 of the primary light sources are being tilted in order to set the position of the light source images. This is illustrated in Figure 5C.

Of course, it is also possible that the symmetry axes z1 , z2, z3 of the light sources as well as the symmetry axes HA1 , HA2, HA3 of the respectively associated collectors are tilted, i.e. are not parallel, relative to each other, as illustrated in Figure 5D.

The grazing-incidence collectors can be rotational shells, for example ellipsoids, or composite surfaces consisting of a hyperboloid segment and, adjacent to the latter, an ellipsoid segment, as disclosed in US 7,015,489, whose content is hereby incorporated herein by reference in its entirety. Collectors with a plurality of rotational shells arranged within each other are referred to as nested collectors. Systems with composite surfaces consisting for example of a hyperboloid segment and an ellipsoid segment are referred to as Wolter systems.

Another possibility to consider are collectors, which have so-called free-form surfaces with no rotary symmetry about a unique axis. In collectors of this kind, the position of a light source image would not be adjusted by a tilting of the collector but by the shape of the optical free-form surface itself. Of course, it is also conceivable that a collector with a free-form surface is in addition being tilted.

Finally, the positions of the light source images could also be adjusted by light sources with an asymmetric emission of radiation whose directional emission characteristic can be adjusted, i.e. the position of the light source images and thus the illumination in the field plane and the pupil plane of an illumination system can be determined by the directional light-emission characteristic of the source.

As described above, the discharge-based light source itself has a preferred axis, which for example in a pinch plasma source is the z-axis in the longitudinal direction of the discharge cavity, and if the collector is configured with rotary symmetry about an axis, the symmetry axis HA represents a preferred axis.

With preference, the arrangement of the multiple light sources and collectors is selected so that the superposition of the illuminated areas in the field plane adds

up to a predetermined field shape, for example the shape of a ring field as shown in Figures 1 B. Furthermore the arrangement of the multiple light sources and collectors can be selected such, that the illuminated areas in an exit pupil plane provide for a predetermined exit pupil illumination, e.g. a circular exit pupil illumination having e.g. a certain telecentricity.

The collectors 54, 54.1 , 54.2, 54.3 are preferably arranged in a collector micro- array 56 of their own which can for example also be moved in relation to the light element part 40.

A configuration as a single unit is also possible, wherein the light sources together with the collectors form one light source micro-array.

In a configuration as a single unit, the different positions of the light source images as described above are achieved through the concept that the individual primary light sources and collectors take on specific angles relative to each other, for example that the collector axes are tilted in relation to the axes of the light sources or vice versa. It is also possible to actuate individual primary light sources with motors, or for example to actuate collectors with micro-motors and thereby to change the position of the light source in the plane.

In a two-part configuration where the multiple primary light sources are arranged in a first array and the multiple collectors in a second array, such a change of the positions of the light source images can also be achieved for example by offsetting the first array 40 with multiple light sources relative to the entire array 56 of collectors by a displacement distance y. This is illustrated in Figure 5E. Where components in Figure 5E are identical to those in Figures 5C and 5D, they are identified by the same reference symbols. As can be seen in Figure 5E, the entire first array 40 with primary light sources 42.1 , 42.2, 42.3 is shifted in its position relative to the array 56 with collectors 54.1 , 54.2, 54.3. This also allows the field illumination in the field plane to be moved.

Of course, the offsetting of the entire first array 40 with light sources relative to the entire second array 56 with collectors can also be coupled with an individual tilting of the individual primary light sources and/or collectors. This is illustrated in Figure 5F. Components in Figure 5F that are identical to those in Figure 5E are identified by the same reference symbols. In this application an array of light sources denotes the two dimensional arrangements of light sources and collectors. The two dimensional arrangement could be any two dimensional surface such as a plane but also a curved surface. A special arrangement of such an array is an arrangement in rows and columns. A part of such an array with a multiple of light sources arranged in rows and columns is shown in figure 5G. In figure 5G the primary light sources 42.1 , 42.2, 42.3 are arranged in a first row 97.1 and light sources 42.4, 42.5, 42.6 in a second row 97.2. Light sources 42.1 , 42.4 are arranged in a first column 99.1 , light sources 42.2, 42.5 in a second column 99.2 and light sources 42.3, 42.7 in a third column 99.3.

The relative offsetting of the arrays against each other as shown in Figures 5E and 5F changes the direction of a light bundle emitted by the light source. The direction of the light bundle is indicated by the chief ray CR. If the direction of the chief ray CR is changed by shifting the arrays relative to each other, this has for example the result that different areas are illuminated in a field plane which in Figure 5E is identified by the reference symbol 1000. If the two arrays are not offset relative to each other, the direction of the chief ray is CR1. By a displacement of an amount y, the direction of the chief ray is changed into the direction CR2. Now, if for example a field aperture stop 1010 is arranged in the field plane that is to be illuminated, the light bundle which has been deflected due to the offset falls with its chief ray onto the field aperture stop 1010 and therefore does not contribute to the illumination of the field plane. In this way, it is possible by way of the position- shifting to influence the illumination in the field plane, in particular the so-called uniformity in the field plane, in a specifically targeted way. It is also possible to tilt the entire array consisting of individual primary light sources and collectors by a certain angle. This also results in a change of the illumination.

Likewise, as the light source array together with the multiple collectors that are respectively assigned to each primary light source can be influenced by tilting and/or shifting, it is possible to influence not only the illumination or uniformity in the field plane, but also the shape and the characteristics of the illumination in the exit pupil, i.e. the so-called setting and i. e. the telecentricity of the illumination in the exit pupil.

Besides grazing-incidence collectors which, in the light path from the light source to the plane to be illuminated, preferably to the field plane, are arranged in front of the light source, it is possible to envision an alternative embodiment where the collectors are configured so that they also take in light emitted in the rearward direction. A system of this kind would be for example a system in which the LED diodes 69.1 , 69.2, 69.3, 69.4 are used as primary light sources. In LED diode arrays of this kind it is customary to collect the light by means of rear reflectors 73.1 , 73.2, 73.3, 73.4.

Such rear reflectors 73.1 , 73.2, 73.3, 73.4. have a preferred direction in relation to which they can have a rotationally symmetric configuration. The rear reflectors, too, can be designed in such a way that they have a light-converging effect.

In figure 6 a system with such rear reflectors 73.1 , 73.2, 73.3, 73.4 are shown. The primary light sources 69.1 , 69.2, 69.3, 69.4 are projected in a plane 105. In plane 105 images 71.1 , 71.2, 71.3, 71.4 of the primary light source are produced.

The field plane in which an object to be illuminated can be situated is denoted with reference number 108. As is clear from figure 6 the images of the light sources do not lie in field plane 108. Each light source therefore illuminates a so called illumination spot in field plane 108. The sum over all illumination spots as discussed with regard to figure 1 B provides for the field illumination, which can have e.g. an arc-shaped field contour. In order to be able to illuminate the desired contour shape in the field plane, it is necessary to project the images of the respective primary light sources into different areas in the field plane. To achieve

this, the individual light bundles 75.1 , 75.2, 75.3, 75.4 emitted by the respective primary light sources 69.1 , 69.2, 69.3, 69.4 are tilted relative to each other as shown in Figure 6. The tilt angles of the individual light channels or light bundles 75.1 , 75.2, 75.3, 75.4υelative to each other are defined essentially by the tilt angles of the chief rays CR1 , CR2, CR3, and CR4 of the respective light bundles 75.1 , 75.2, 75.3, 75.4. A tilting of the chief rays CR1 , CR2, CR3, CR4 of the light bundles 75.1 , 75.2, 75.3, 75.4 associated with the respective primary light sources 69.1 , 69.2, 69.3, 69.4 can be achieved in different ways. In a first embodiment of the invention, the reflector assigned to the respective primary light source 69.1, 69.2, 69.3, 69.4 is being tilted. The tilting of the reflector has the effect that the light bundle defined by the reflector has a tilted chief ray. By tilting the light sources and illuminating different areas in the field plane not only different field contours can be illuminated, but also the homogeneity of the illumination by mixing different illumination spots can be influenced.

In the embodiment illustrated in Figure 6, the light source 69.1 , 69.2, 69.3, 69.4 is tilted together with the reflector 73.1 , 73.2, 73.3, 73.4, while the relative position between the reflector and the light source always stays the same. The light source and the reflector together form in this case a so-called light source unit. This is a possibility to realize the tilting of the light bundles 75.1 , 75.2, 75.3, 75.4 originating, respectively, from the primary light sources 69.1 , 69.2, 69.3, 69.4.

In an alternative embodiment of the invention, all of the light sources 42 could be set up in a regular arrangement in a multi-source array in columns and rows as shown for example in Figure 5G. The collectors which are assigned to the respective primary light sources are in this case arranged in a second component. In this second component, the collectors can be tilted relative to their associated light sources, or each collector assigned to a primary light source can have a very specific reflection characteristic, so that by means of the different relative positioning between the lights sources and their associated collectors different areas of the field plane 108 are illuminated.

In exactly the same way, it would of course be possible to assemble the collectors into a multi-collector array and to tilt or position the respective primary light sources assigned to the collectors in such a way relative to the collectors that the light bundle radiates in the desired direction into the field plane.

If the collectors assigned to the individual primary light sources are designed with a tilt or with the capability of moving in relation to the primary light sources, it is possible to illuminate different areas of the field plane and different areas in the pupil plane by moving or tilting of the individual collectors. By moving the collectors and/or the light sources relative to each other, it is possible to set different illuminations in the exit pupil.

As an alternative to setting illuminations in the exit pupil with reflectors or light sources that are movable relative to each other, it is conceivable to provide a multi-source array with a larger number of primary light sources than are necessary for an illumination, and to switch the light sources on and off individually.

This is shown in figure 7A to figure 7B. In figure 7A a multiple light source array 80 with light sources 81.1 , 81.2, 81.3, 81.4, 81.5, 81.6, 81.7, 81.8 is shown. Only the light sources 81.3, 81.5 and 81.8 are turned on. Each turned on light source provides for an image of the light source 83.3, 83.5 and 83.8 in a plane 105. The rays intersecting in a specific field point in the field plane 108 form the pupil in the associated pupil plane 85 or 87 respectively. The exit pupil contours belonging to different field points in plane 108 overlap congruently in the plane 85 or 87 respectively. To each field point can be assigned an energy-weighted central ray representing the center of gravity of the bundle of rays originating from this field point. In an ideal case the complete set of all these energy-weighted central rays intersects in the pupil plane 85 or 87 resp., which coincides or is conjugated to the entrance pupil of the subsequent projection optics. In figure 7A and 7B only one field point of the field - the central field point 102 - in the field plane 108 is shown. The illumination with the light sources turned on according to figure 7A provides

for a ring shaped exit pupil 85.

In figure 7A different light sources, e.g. the light sources 81.1 , 81.2, 81.4, 81.7 and 81.6 are turned on provided in an a circular exit pupil 87 belonging to a field point - here the central field point 102 of the field - in field plane 108.

As described above a first illumination in an exit pupil could be provided by a first subset of the multiple light sources, and a second illumination in an exit pupil could be provided by a second subset of the multiple light sources. By switching off the light sources associated with the first illumination out of the totality of the multiple primary light sources and switching on the second subset of multiple primary light sources, it is possible to perform a setting change very quickly. A setting change of this kind can be realized significantly faster than with a mechanical process.

The following Figures 8 to 13 show different illumination systems in which an optical element according to the invention with multiple primary light sources is used.

Figure 8 shows a first illumination system according to the invention, wherein an optical element according to the invention with multiple primary light sources is used in the illumination system. The optical element 100 with multiple primary light sources 101.1 , 101.2, 101.3 performs the imaging of the primary light sources into a plane 105 which is not the field plane nor the pupil plane nor a plane conjugate to the field plane and/or the pupil plane, for example by means of collectors 103.1 , 103.2, 103.3 which are assigned to the respective primary light sources, and also performs the shaping of the ring field in the plane 108 in which the object to be illuminated is arranged, as well as the illumination of the exit pupil 106 of the illumination system. The angles of inclination of the light bundles 109.1 , 109.2, 109.3 emitted by the associated light sources and the arrangement of the primary light sources on the optical element can be calculated by means of the grid transformation as described previously herein. The optical element of the primary light sources as shown in Figure 8 can comprise up to several thousand individual

primary light sources. The location of the entry pupil of a projection system as illustrated, which coincides with the exit pupil of the illumination system, is defined by the point of intersection S of the optical axis HAO of the projection objective with the energy-weighted chief ray 107 of a light bundle to a field point of the field to be illuminated, for example the central field point. The energy-weighted chief ray represents the energy-weighted average of all illumination rays that pass through a field point 102. The optical element 100 with multiple primary light can be arranged for example as a multi-source array, wherein the primary light sources are arranged in rows and columns as shown for example in figure 5G. The field point 102 shown in figure 8 is the central field point of the illuminated field in the field plane 108. Furthermore shown in figure 8 is the plane 105 in which the light source images 104.1 , 104.2, 104.3 are provided. As is apparent from figure 8 the light source images 104.1 , 104.2, 104.3 are not provided in the field plane 108. Therefore each light bundle 109.1 , 109.2, 109.3 provides for an illumination spot in the field plane 108.

Figure 9 illustrates a projection exposure apparatus with an illumination system in which no further optical components are arranged in the light path from the optical element 100 with a multiple of light sources to the plane that is to be illuminated which in this case coincides with the field plane 108. If the images of the primary light sources of the optical element 100 with a multiple of light sources in the embodiment of Figure 9 are produced in the field plane, this example represents the special case of a critical illumination. As a general statement, the optical element 100 with a multiple of light sources as shown in Figure 8 can be arranged in any desired plane which is not the field plane or conjugate to thereto nor the pupil plane or a plane conjugate to the pupil plane. In the projection exposure apparatus according to Figure 9, the exit pupil 106 of the illumination system coincides with the entry pupil of the projection objective 112 that follows the illumination system. The projection objective 112 in this case is a six-mirror projection objective with mirrors 114.1 , 114.2, 114.3, 114.4, 114.5, 114.6 as disclosed for example in US 6,353,470. The optical axis of the projection objective is identified as HAO. Instead of a six-mirror projection objective, imaging systems

with more than six mirrors which project an object, e.g. a reticle 113 that is arranged in the field plane 108 onto a light-sensitive object 116 situated in an image plane 120 are also conceivable. Only as an example, an eight-mirror projection objective as disclosed in US 2002/0154395 A1 should be mentioned here.

A projection exposure apparatus as shown in Figure 9 or a projection exposure apparatus with an illumination system as shown in Figure 8 and with a six-mirror projection objective or with an eight-mirror projection objective has a high light output since the number of the reflections in the illumination system is low.

Furthermore, such a system has very good imaging properties and correction capabilities due to the large number of mirrors in the image-producing projection objective. The projection objective 112 projects an image of a mask, which is also called a reticle 113 arranged in the field plane 108, into the image plane 120 where a light-sensitive object 116 for example a wafer is arranged. The mask in the field plane as well as the light-sensitive object in the image plane can be arranged so that they are capable of being moved, for example with so-called scanning stages, which are movable along the scan direction.

In a further embodiment which is shown in Figure 10, an illumination system with an optical element 100 with multiple primary light sources 101.1 , 101.2, and 101.3 is presented, where a field-shaping element 110 is interposed in the light path from the optical element 100 to the field plane 108. Elements that are identical to those in Figure 8 are identified with the same reference symbols.

The advantage of an embodiment in accordance with Figure 10 should be seen in the fact that the optical element 100 with multiple light sources in an embodiment according to Figure 10 can be designed for a rectangular field.

The shaping into a ring field occurs through the field-shaping element 110 which in the present example is a field-shaping grazing-incidence mirror.

Figure 11 illustrates a projection exposure apparatus with an illumination system according to Figure 10. Elements that are identical to those in Figure 9 are identified with the same reference symbols.

As an alternative to the illustrated illumination system, an illumination system with an optical element with multiple primary light sources can include one or more imaging elements 111.1 , 111.2. An first embodiment of such an illumination system with one normal incidence mirror 111.1 as imaging optical element is presented in Figure 12.The embodiment shown in figure 11 is also an embodiment in which the optical element with multiple light sources 101.1 , 101.2, 101.3 is neither situated in a field plane 108 or a plane conjugated thereto nor in a exit pupil plane 106 or a plane conjugated thereto. Furthermore the system is a system in which the light source images 104.2 come to lie in plane 105, which is in the light path from the optical element 100 to the field plane 108 situated before the field plane 108. Also shown is the central field point 102 and the intersection point S of chief ray 107 to the central field point 102 with the optical axis HAO of a subsequent projection objective.

In figure 13 a second embodiment with one imaging optical element 111.2 and a field shaping element 110 is shown. The field shaping element 110 is a grazing incidence mirror. Components in figure 12 and 13 that are identical to those in Figure 10 or 11 carry the same reference symbols. The element 111.1 , 111.2 which performs an imaging function can project the optical element 100 with multiple primary light sources 101.1 , 101.2, 101.3 into the exit pupil of the illumination system. Of course it is also possible to use several optical imaging elements which may even have the capability to produce an intermediate image of the ring field segment in the field plane 108 in case that aperture stops are used. Using additional images or other mirrors can be advantageous if the ray geometry has to be additionally folded for example in order to arrange the optical element with multiple light sources in a more accessible space. This is obviously the case in Figure 13. In all of the systems shown in Figures 8 to 13 the optical axis HAO of the projection objective 112 is indicated. The exit pupil of an illumination system,

which coincides with the entry pupil of the projection exposure apparatus, is defined as the point of intersection of the energy-weighted central ray to the different field point of the optical axis HA. This can be seen with particular clarity in Figures 8 to 10, where the exit pupil is identified by the reference symbol 106. Of course, an individual of ordinary skill in the pertinent art would be able without inventive effort to envision a virtual pupil instead to the actual pupil. The principal ray angle at the field plane would in this case be negative. Such a system is presented in WO 2004/010224. The disclosure of WO 2004/010224 is hereby incorporated by reference in its entirety in the present application.

The entry back focus is negative in reflective systems with a so-called virtual entry pupil. Figure 14 graphically illustrates the concept of a negative back focus.

Figure 14 shows the principal ray CRB of an illumination light bundle to the central field point of an illuminated field as shown for example in Figure 2 or Figure 4. As shown here, the principal ray CRB of the illumination light bundle is reflected on a reflective object REFLOBJ, for example a reticle, and enters as the principal ray CRP of a projection light bundle into a projection objective, of which the first mirror S1 and the second mirror S2 are shown. The concept of negative back focus means that the principal ray CRB or CRP intersects the optical axis of the projection objective in the light path ahead of the reflective object, for example the reticle. In contrast to this, the point of intersection in systems with positive back focus lies in the light path after the reflective object REFLOBJ.

The term "back focus of the entry pupil" as used herein means the distance of the point of intersection of the optical axis HA of the projection objective with the principal ray CRB to the central field point in the object plane. In a system with negative back focus of the entry pupil, the entry pupil VE, which could also be called the virtual entry pupil VE, is mirrored into a mirrored entry pupil RE at the object plane. The mirrored entry pupil RE is formed in the light path from the light source to the image plane ahead of the reflective object REFLOBJ.

In systems with positive back focus of the entry pupil as shown in Figures 8 to 13, the entry pupil will lie, in the light path from the object plane to the image plane, in a position after the reflective object REFLOBJ in the object plane. In systems with negative back focus, the mirrored entry pupil RE will lie, in the light path from the light source to a reflected object, in a position ahead of the entry pupil.

Furthermore, in a system with a negative back focus the chief rays of the light bundles reflected at the reflective object run divergently with respect to the optical axis of the projection objective.

A system with a positive back focus is a system in which the chief rays of the light bundles reflected at the reflective object REFLOBJ run convergently with respect to the optical axis.

The system shown in Figure 15 is similar to the system of Figure 9, but with a negative back focus. The optical element with multiple light sources is identified with the reference symbol 1100. The optical element with multiple light sources is for example a multi-source array wherein each primary light source is projected by means of an associated collector substantially into the field plane 1108. The multi- source array 1100 in the present example is located substantially in the plane in which the mirrored entry pupil RE of the projection objective lies. Components that are identical to those in Figure 9 are identified by the same reference symbols raised by 1000. Thus, the reference symbol 1120 belongs to the plane into which an image of an object in the field plane is projected, for example a reticle. The reference symbol 1112 belongs to the projection objective with negative back focus which contains in its light path from the field plane 1108 to the image plane 1116 a first mirror 1114.1 , a second mirror 1114.2, a third mirror 1114.3, a fourth mirror 1114.4, a fifth mirror 1114.5, and a sixth mirror 1114.6. The primary light sources of the multi-source array are identified as 1101.1 , 1101.2, and 1101.3.

However, it is also conceivable to place the optical element with multiple light sources 1100 at another location, in which case the images of the primary light

sources are no longer projected essentially into the field plane as illustrated in Figure 15, but into an intermediate image plane, which can lie for example upstream of the field plane 1108. The field plane 1108 will in this case show so- called defocused images of the primary light sources, which are also called illumination spots. As in the embodiment according to Figure 9, there is no further optical element arranged between the optical element 1100 with multiple light sources and the field plane 1108, i.e. the light of the primary light sources of the optical elements 1100 with a multiple of light sources after reflection at the collectors associated with each of the primary light sources of the optical element 1100, is guided directly into the field plane.

Figure 16 shows a further embodiment of a projection exposure apparatus with negative back focus and a multi-source array. The system is of the type which in addition contains a mirror 1110 in the light path from the multi-source array 1100 to the field plane 1108. This mirror 1110 is a grazing-incidence mirror, which serves to redirect the light path between the multi-source array 1100 and the field plane 1108. This is advantageous in regard to space requirements. The redirecting mirror 1110 can be configured as a planar mirror, in which case the mirror has no influence on the projected image. As an alternative, the mirror can also be designed with optical power. As a possible example, the field mirror could have a toroidal surface and serve to shape the field in the field plane. An arrangement of this kind would have the advantage that the distribution of the individual light sources on the multi-source array can conform to a regular pattern, for example in rows and columns. The ring-shaped illumination in the field plane is shaped in this case substantially by the mirror 1110 which has optical power.

Figure 17 shows a projection exposure apparatus with a similar design concept as shown in Figure 16. Figure 17 further shows the entry pupil RE which is mirrored at the field plane 1108, and it shows the plane 1200 in which the mirrored entry pupil is located. As can be seen in Figure 17, the optical element 1100 with multiple light sources is arranged in the light path downstream of the plane 1200 in which the mirrored entry pupil is located. The optical element 1100 with multiple

light sources is situated neither in a field plane nor in a pupil plane. Accordingly, the field plane is illuminated with illumination spots, meaning that the images of the light source are defocused. There is again a redirecting mirror 1110 arranged in the light path from the multi-source array to the field plane, which can again be configured as a purely planar mirror or as a mirror with optical power, as described in the context of Figure 16.

The following is a description of a concrete example of an embodiment of an optical element with multiple light sources according to the invention.

With the requirements for a telecentricity of 1 mrad and for an entry-pupil back focus (which is defined as the length of the energy-weighted central ray between the field plane and the exit pupil) of about one meter, one finds an ideal distance of less than 2 mm between the light source images in the field plane 108. With a distance of less than 2 mm, the telecentricity error in the exit pupil with a point in the field of the field plane 108 between two selected target points is less than 1 mrad.

If the light source images of the primary light sources are made larger by design, the field points between the target points will always receive light from the pupil directed to at least two target points. The telecentricity requirement in this case is also met with fewer target points, for example with 30 target points. Figures 18A and 18B show, respectively, an optical element according to the invention with round light sources and an enlarged detail portion of the optical element.

As an example, Figure 18A shows the optical element with multiple primary light sources with approximately round primary light sources 130. The reference symbol 132 indicates a detail portion of the optical element with multiple primary light sources. As Figure 18B shows, the entire optical element consists of a large number of primary light sources with at least 30 light sources being assigned to one field point. If one assumes a number of more than 50 points for the pupil, the optical element will need about 2000 to 5000 primary light sources.