WO2005015313A1

WO2005015313A1 - Illumination mask for range-resolved detection of scattered light

Info

Publication number: WO2005015313A1
Application number: PCT/EP2003/010604
Authority: WO
Inventors: Michael Arnz; Oswald Gromer; Gerd Klose; Joachim Stühler
Original assignee: Carl Zeiss SMT GmbH
Current assignee: Carl Zeiss SMT GmbH
Priority date: 2003-08-04
Filing date: 2003-09-24
Publication date: 2005-02-17
Anticipated expiration: 2006-02-04
Also published as: AU2003267402A1; US20080285018A1; US7408631B2; US7755748B2; US20050264819A1

Abstract

The invention relates to an illumination mask (10) for a device for range-resolved determination of scattered light, having one or more scattered-light measuring structures (11) which respectively include an inner dark-field zone (22) which defines a minimum scattering range, to a corresponding device, to an associated method for determining scattered light, and to a microlithographic projection exposure apparatus having such a device. According to the invention, in the case of the illumination mask the scattered-light measuring structure has a scattered-light marker zone (20) in the form of a bright-field zone which on the one hand borders the inner dark-field zone and on the other hand borders an outer dark-field zone (21) which defines a maximum scattering range.

Description

ILLUMINATION MASK FOR RANGE-RESOLVED DETECTION OF SCATTERED LIGHT

The invention relates to an illumination mask for a device for range- resolved determination of scattered light, having one or more scattered-light measuring structures which respectively include an inner dark-field zone which defines a minimum scattering range, to a cor- responding device which comprises the illumination mask for providing measuring radiation on an entrance side of an object to be tested, called test component hereinafter, and a detection part for range-resolved detection of scattered light on an exit side of the test component, to an associated method for determination of scattered light, and to a microlithographic projection exposure apparatus having such a device.

Such devices and methods are used, for example, for the purpose of drawing from the scattered light of optical components detected in a range-resolved fashion conclusions about their optical properties and optical quality, for example with regard to surface roughnesses, contaminations and material inhomogeneities. Moreover, a scattered-light portion thus determined can correctively be considered if a spatially resolved knowledge, which is as accurate as possible, of the quantity of radiation actually supplied by an optical system is desired in corresponding optical applications, such as lithographic exposure apparatuses, for example. In the text which follows, the term light is used for the sake of simplicity to denote electromagnetic radiation of arbitrary wavelength, in particular radiation in the UV or EUV range.

For the purpose of range-resolved determination of scattered light, in a conventional method a non-transparent object which acts as dark-field zone is introduced into the beam path of a directed illumi- nating radiation having a larger beam cross section compared to the object, such that a shadow image of the object is produced. If the system to be examined, presently also termed a test component, is an imaging optical system, the non-transparent object is positioned in the object plane of the system. The object is imaged by the imaging system onto an image of the object in the image plane which may be enlarged or reduced by the reproduction scale and which is denoted as an "aerial image" below, as is also denoted the pure shadow image in the case of test components which are not focus- ing imaging systems.

Scattering of the illuminating light by the test component has the effect that scattered light passes laterally into the dark zone of the surface of the aerial image. The intensity of this scattered light normally decreases with increasing distance from the edge to the middle of the aerial image. Depending on the lateral distance from the edge of the aerial image, denoted below as scattering distance or scattering range, an intensity distribution of the scattered light thus arises which is determined as a function of distance, that is to say range- resolved.

In a conventional mode of procedure, a sensor surface is introduced as a component of a detection part into the plane in which the aerial image is produced or detected. For the sake of simplicity, the term image plane is also used for this plane whenever the test component is not a focusing imaging system. It is important here that the sensor surface is smaller than the aerial image so that a minimum, lateral scattering distance remains between the edge of the sensor surface and the edge of the aerial image. In consecutive measuring operations, objects of different size are then positioned such that the size of the aerial image, and thus the minimum scattering distance, changes. It is known for this purpose to provide objects of variable size at a distance from one another on an otherwise transparent illumination mask. Chromium squares of variable edge length, for ex- ample, are used as objects. The sensor surface correspondingly has a quadratic geometry. In order to vary the minimum scattering distance, the illumination mask is displaced such that the objects applied to it are brought one after the other into the beam path and im- aged onto the sensor surface centred on the image side in the plane of the aerial image.

In one of the conventional measuring methods, the entire intensity of the scattered light impinging on the sensor surface over a fixed time interval is determined integrally for each object size, for example with the aid of an electro-optical measuring element. Each object size is assigned a minimum scattering distance in the image plane, the maximum scattering distance being infinity in theory for all objects. The scattered light distribution can be reconstructed from the intensity distribution as a function of the minimum scattering distance.

Instead of directly measuring the intensity of the scattered light on the sensor surface, in the case of an alternative conventional meas- uring method a photoresist layer is provided in the image plane as sensor surface, and in different measuring operations a respectively different region of the photoresist layer is irradiated with illuminating radiation of increasing intensity. Then, that limiting value for the intensity is determined in the case of which the aerial image first van- ishes entirely as a structure in the photoresist layer. This limiting value for the intensity of the illuminating light is used instead of the light intensity integrated over the sensor surface in order to determine the scattered light distribution. These and further details relating to conventional methods of determining scattered light are to be found in the relevant literature, see, for example, the magazine article by J.P. Kirk, "Scattered Light in Photolithographic Lenses", SPIE, Volume 2197 (1994), pages 566-572, and the magazine article by Eugene L. Church, "Fractal Surface Finish", Applied Optics, Volume 27, No. 8 (1988), pages 1518-1526. In the case of the conventional methods outlined above, effects with a long scattering range are normally superimposed on effects of short to medium scattering ranges. The non-transparent objects on the illumination mask certainly define a minimum scattering range, but the maximum scattering range is limited only by the size of the object field, that is to say the illumination mask. Moreover, since the squares have a relatively large spacing from one another, in order not to disturb one another during the measurements, and are typi- cally arranged in a nonsymmetrical distribution, the long-range scattering light intensity is a function of the field point, and so systematic measuring errors can occur.

The devices and methods currently of interest for determining scat- tered light are chiefly used in the field of medium scattering ranges, what corresponds, e.g., to typical object-side scattering ranges in a region from approximately 4 μm to approximately 1000 μm and/or image-side scattering ranges from approximately 1 μm to approximately 250 μm.

It is the technical problem underlying the invention to provide an illumination mask for a device for range-resolved determination of scattered light, and to provide such a device and an associated method for determining scattered light with the aid of such a device, which permit scattered light to be determined with a high range resolution in conjunction with a good signal-to-noise ratio. It is also an aim of the invention to provide a microlithographic projection exposure apparatus equipped with such a device.

The invention solves this problem by providing an illumination mask having the features of Claim 1 , a device having the features of Claim 7, a method having the features of Claim 10 or 11 , and a microlithographic projection exposure apparatus according to Claim 12. The illumination mask according to the invention has one or more scattered-light measuring structures in which a scattered-light marker zone in the form of a bright-field zone borders an inner and an outer dark-field zone. By contrast with the above-described conventional illumination mask, the result of this is to fix not only a minimum, but also a maximum scattering range. This reduces the long-range scattered-light component, and this can contribute to a significant improvement in the signal-to-noise ratio and to the avoid- ance of systematic measuring errors. The geometry of the scattered-light marker zone can be adapted in this case to the geometry of the scattered-light effects to be measured or of the test component.

In accordance with Claim 2, the illumination mask has a plurality of scattered-light marker zones with inner and/or outer dark-field zones of different sizes. It is thereby possible to vary the minimum and/or the maximum scattering range.

In a refinement of the invention according to Claim 3, the scattered- light marker zone is designed as a ring, something which is advantageous, in particular, when the scattered-light effects to be measured have a rotationally symmetrical structure. This is the case, for example, whenever the test component likewise has a substantially rotationally symmetrical structure, as is the case with many optical components.

In a development of the invention according to Claim 4, the illumination mask has a plurality of scattered-light marker zones designed as rings and having different ring radii. As a result, the inner and the outer dark-field zones, and thus the minimum and maximum scattering range, vary from ring to ring, and this can be used advantageously in determining the scattered light in a range-resolved fashion. Since the scattered light generally decreases exponentially with increasing scattering range, it can be favourable to enlarge the ring width suitably with increasing ring radii, in order thereby to compensate for the decrease in the measured scattered-light intensity at least partially.

In order to achieve an improvement in the signal-to-noise ratio, the scattered-light marker zones designed as rings can be selected with their radii overlapping. Alternatively, the ring radii are selected not to be overlapping, and this increases the range separation when de- termining the scattered light.

In accordance with Claim 5, the illumination mask is advantageously designed as a dark-field mask with the scattered-light marker zone or zones as bright-field zones in the mask dark field. It proves to be favourable in this case that in essence only the part of the illumination mask left free from the scattered-light marker zones is of transparent design such that, in particular, no long-range scattered light is caused by large transparent surface regions.

In a development of the invention in accordance with Claim 6, the scattered-light measuring structures and one or more dark-field area elements are arranged in the form of a matrix in a regular distribution on the illumination mask such that all the nearest neighbours of the dark-field area elements are scattered-light measuring struc- tures. The dark-field area elements can be used, for example, for calibration purposes. Since all the nearest neighbours of the dark- field area elements are scattered-light measuring structures, the influence exerted by the surrounding scattered- light measuring structures on the aerial image of a dark-field area element virtually does not differ from dark-field area element to dark-field area element.

The device according to the invention for range-resolved measurement of scattered light at a test component has as illumination mask one according to one of Claims 1 to 6. By using the illumination mask according to the invention, it is possible to reduce the component of long-range scattered light, in particular, in the case of such a device, and this can have an advantageous effect on the quality of the measurement.

In a development of the inventive device according to Claim 8, the test component is an optical imaging system in which the illumination mask is arranged in or near the object plane. The detection part has a field stop to be arranged in or near the image plane, a micro- objective and a sensor surface for scattered-light measurement. The field stop opening is imaged onto the sensor surface by the micro- objective. The field stop screens off any disturbing, external scattered light which does not originate from the scattered-light measuring structure which is imaged onto the image plane by the optical imaging system. The sensor surface can be positioned outside the image plane by using the micro-objective. The scattered-light information obtained by the sensor surface can be used by the associated evaluation part to determine in a range-resolved fashion the scattered light caused by the test component.

In a development of the inventive device according to Claim 9, the sensor surface is implemented in a first variant as an electro-optical sensor. In a second variant, the sensor surface is designed as a radiation-sensitive layer which upon irradiation is changed from a first state to a second state. In the latter case, the detection part has suitable means for measuring the state of the radiation-sensitive layer. In this second variant, the intensity or dose of the illuminating light can be increased for the purpose of range-dependent determination of scattered light until the aerial image vanishes as a struc- ture in the radiation-sensitive layer. The corresponding intensity or dose of the illumination and/or the structural dimensions of the regions exposed by scattered light with different intensities of illumination are then a measure of the scattered light caused by the test component at each respective scattering distance. In the case of the inventive method according to Claim 10, the electro-optical sensor surface is divided into subregions, for example into annular regions. When using a CCD array, this can be achieved by virtue of the fact that a multiplicity of CCD pixels are respectively combined to form corresponding subregions, and the intensity of the radiation falling onto the pixels of the respective subregion in a fixed time period is determined. The range resolution can be improved by evaluating subregions of the sensor surface, since several subre- gions of different scattering distance can be evaluated separately on the image side for each individual scattered-light measuring structure provided on the object side.

In the case of the inventive method according to Claim 11 , the same scattered-light measuring structure is imaged onto different regions of the photosensitive layer in various measuring operations with a successive increase in intensity of the measuring radiation, and the range-resolved determination of scattered light is performed using state data for these regions of the photosensitive layer and/or using data for the intensity or dose of the measuring radiation. Depending on what is needed, it is possible thereby to make use only of the radiation dose that is required to cause the image of the scattered- light measuring structure on the photosensitive layer to vanish, or only of the layer state data for these regions, or of both parameters in combination for the range-resolved determination of scattered light.

The invention comprises as an important field of application a microlithographic projection exposure apparatus having a device for range-resolved determination of scattered light at a projection objective of the projection exposure apparatus according to one of Claims 7 to 9. In this case, the projection objective need not be removed from the projection exposure apparatus for the purpose of range- resolved determination of scattered light. Advantageous exemplary embodiments of the invention are illustrated in the drawings and will be described below. In the drawings:

Figure 1 shows a schematic side view of a device for range- resolved measurement of scattered light at a projection objective of a microlithographic projection exposure apparatus,

Figure 2 shows a top view of an illumination mask having matri- cially arranged scattered-light measuring structures for a device for range-resolved determination of scattered light in accordance with Figure 1 ,

Figure 3 shows a top view of one of the scattered-light measuring structures of the illumination mask of Figure 2 having a scattered- light marker zone designed as a ring,

Figure 4 shows schematic top views of a sensor surface of a CCD array for a detection part of the device of Figure 1 for the pur- pose of illustrating a subdivision into subregions for measurement purposes, and

Figure 5 shows diagrams for the illustration of a simulation result for a range-resolved determination of scattered light, as can be exe- cuted using the device of Figure 1.

Figure 1 shows a schematic side view of a device for range-resolved measurement of scattered light in operating position at a projection objective 1 , for example of a microlithographic projection exposure apparatus for semiconductor wafer patterning. The device includes an illumination mask 10 positioned in the object plane of the objective 1 , a detection part which has a circular field stop 3 positioned in the image plane of the objective 1 , a micro-objective 4 and a CCD array 6 as sensor surface, and an evaluation part 7. The micro- objective 4 images the field stop opening onto the CCD array 6 via a deflecting mirror 5. The illuminating light is provided by a conventional illuminating system (not shown).

The illumination mask 10 to be positioned in the object plane of the objective 1 , for example by means of a conventional reticle holder, has one or more scattered-light measuring structures which in each case include a scattered-light marker zone in the form of an annular bright-field zone between an inner and an outer dark-field zone. In the measurement mode, a respective one of the scattered-light measuring structures is positioned in the object plane such that the imaging of the scattered-light measuring structure with the aid of the objective 1 in the image plane thereof produces an associated aerial image centred in the interior of the circular field stop 3.

Here, the object-side focal plane of the micro-objective 4 corresponds to the field stop plane. In order to avoid beam vignetting at a high numerical aperture of the objective 1 , that is to say for numerical apertures greater than approximately 0.7, the field stop 3 is very thin, that is to say it has a thickness of less than 0.5 μm, preferably of only approximately 0.2 μm. On the one hand, the image of the stop edge is completely visible on the CCD array 6, and so detection is possible up to a distance of approximately 0.5 μm from the stop edge for image-side numerical apertures of the objective 1 as far as at least 0.95 in a fashion virtually free from vignetting, and on the other hand it fills up the sensor surface 6 as well as possible so that as many CCD pixels as possible are situated in the image. Spatial averaging of the scattered-light signal over many CCD pixels helps to improve the signal-to-noise ratio. The numerical aperture of the micro-objective is selected to be greater than the image-side numerical aperture of the objective 1. If no pupil apodization or other inhomogeneous weighting of beam direction occurs over the pupil of the micro-objective 4, the image picked up by the CCD array 6 of the field stop interior thus constitutes an ideal enlarged image of the aerial image. The field stop 3 can be used to screen off the aerial image from disturbing external scattered light. The micro-objective 4 enlarges the aerial image such that it can be adequately resolved by the CCD array 6.

In a concrete dimensioning example, a pixel spacing of 20 nm with reference to the field stop plane results, that is to say a CCD array 6 with 1024 x 1024 pixels can detect the whole of a circular field stop 3 with a diameter of 20 μm, virtually 75% of the CCD pixels being situated in the image of the stop interior.

The evaluation part 7 serves the purpose of range-resolved determination of the scattered light caused by the test component with the aid of the scattered-light data picked up by the detection part.

Alternatively, instead of the CCD array 6, it is possible to use any other desired conventional electro-optical sensor system. As a further alternative, it is possible to use as sensor surface a radiation- sensitive layer, in particular a photoresist layer, which upon irradia- tion can be changed from a first state to a second state. In the case of such an embodiment, the micro-objective 4 and the field stop 3 can be eliminated, and the radiation-sensitive layer is placed directly in the image plane during examination at a microlithographic projection objective, for example in the form of a wafer provided with the photoresist layer.

Figure 2 shows an advantageous implementation of the illumination mask 10 for the device for range-resolved determination of scattered light in accordance with Figure 1. In this example, the mask 10 has a plurality of scattered-light measuring structures 11 and dark-field area elements 12 which are arranged in the form of a matrix in a regular distribution, each dark-field area element 12 being surrounded only by scattered-light measuring structures 11 as nearest neighbours. Specifically, the matrix arrangement comprises 7 x 7 fields which in each case include either a scattered-light measuring structure 11 or a dark-field area element 12, the edge fields and every second row and every second column of the matrix arrangement being occupied continuously by scattered-light measuring structures 11. The centres of neighbouring matrix fields have a sufficient spacing a. The illumination mask 2 is designed as a dark-field object, that is to say the surface regions between the scattered-light measuring structures and the dark-field area elements consist of a material which is not transparent to the illuminating light.

As may be seen more clearly from Figure 3, each scattered-light measuring structure 11 has an annular, transparent scattered-light marker zone 20. This encloses an inner dark-field zone 22 and is bounded on the outside by an outer dark-field zone 21. Both dark- field zones are illustrated by hatching in Figure 3. Provided in the outer and inner dark-field zones 21 and 22, respectively, for the purpose of positioning the scattered-light measuring structure 11 are alignment marks 23 which comprise a cruciform bright-field zone which is generally negligibly small by comparison with the scattered- light marker zone 20. Four alignment marks 23 are provided abaxi- ally in the example shown at a 90° spacing and at the same radial distance from the centre. Alternatively, a central alignment mark can be provided in addition. Omission of the central alignment mark avoids any interference based on light passing through said mark.

The scattered-light marker zones 20 of the scattered-light measuring structures 11 on the illumination mask 10 of Figure 2 have different ring radii, that is to say the inner radius and the outer radius, and thus the radii of the inner dark-field zone 22 and of the outer dark- field zone 21 , differ from one scattered-light measuring structure 11 to the other. The ring radii of the various scattered-light marker zones 20 are selected to overlap, for example, that is to say the inner radius of one scattered-light marker zone 20 is larger than the inner radius and smaller than the outer radius of another scattered- light marker zone, and this contributes to improving the signal-to- noise ratio. Alternatively, the ring radii can also be selected not to overlap, for example with outer and inner radii which directly adjoin one another such that the range separation is increased during the determination of scattered light. The selection of scattered-light marker zones 20 which overlap or do not overlap in size, is not, of course, limited to the use of rings, but can be used for any desired scattered-light marker geometries.

Figure 4 shows three schematic top views of the CCD array 6 from Figure 1 , more specifically on the pixel field 30 thereof, for the purpose of illustrating a selective subdivision of the CCD pixel field 30 into subregions during measurements of scattered light. The circular image of the field stop opening defines an inner, circular pixel field surface region 32 in which the aerial image imaged by the micro- objective is situated. The part 31 , remaining outside thereof, of the square pixel field 30 of the CCD array 6 has no function for the measurement of scattered light. The inner, circular region 32 is subdivided selectively into subregions for the purpose of measurement and evaluation. In a first measuring step, only one central circular region 33 of the pixel field 30 is used for the measurement, as illustrated in the left-hand partial illustration of Figure 4. In a second measuring step, only one centralized pixel field annular region 34 is actively switched, as illustrated in the middle partial illustration of Figure 4. In a third measuring step, only one second pixel field annular region 35 is activated, the radius thereof being greater than that of the first annular region 34, as is shown in the right-hand partial illustration of Figure 4. In the same way, further measuring steps continue with a progressively larger radius of the selectively acti- vated pixel field annular region until the entire active region 32 is covered. In alternative implementations, the consecutive annular regions can be selected with their outer and inner radii overlapping, or not overlapping. In a measurement cycle with the aid of the device of Figure 1 , all scattered-light measuring structures 11 present on the illumination mask 10 are brought in sequence into the beam path and imaged by the projection objective 1 onto the interior of the field stop 3, and from there onto the CCD array 6 via the micro-objective 4. The entire circular inner surface region 32, which is not vignetted by the field stop, is divided into the successively activated subregions in accordance with Figure 4 by driving the CCD array 6 electronically in appropriate fashion.

Each scattered-light measuring structure 11 defines by its scattered- light marker zone 20 a minimum and maximum scattering range which differ from scattered-light measuring structure 11 to scattered- light measuring structure 11 such that it is possible to achieve a good range resolution. A contribution is also made to this effect by the decomposition of the CCD array 6 into the selectively evaluable subregions of the surface region 32 relevant to the measurement. If a number n of scattered-light measuring structures 11 are present on the illumination mask, and if the surface region 32 of the CCD array 6 relevant to the measurement is decomposed into a number m of subregions with a different associated scattering distance in each case, a total number n m of measured values is obtained which can be used to determine the range-resolved scattered light. In this process, the light intensity respectively scattered into a subregion is determined in integral fashion by averaging over the intensity of the scattered light of all the pixels in the subregion.

The dark-field area elements 12 are used to calibrate the measured values, such elements being introduced successively into the beam path, just like the scattered-light measuring structures 1 1 , and so the contribution of the scattered light produced by the dark-field area elements can be measured for the individual subregions of the CCD array and for each pixel thereof. This contribution is subtracted from the contribution which is produced by the respective scattered-light marker zone 20 for the same subregion or the same pixel as, in the case of illumination of the test component with the respective scattered-light marker zone 20, it is not caused by the scattered light which is due to the test component and is actually to be measured. The differential contribution of scattered light is subjected to time normalization by dividing by a prescribable integration exposure time, and divided by a time-normalized bright-field differential contribution in order to generate calibrated scattered-light measured values. The said differential contribution is determined as the difference between a bright-field measurement for the respective CCD subregion and a dark-field measurement with the illumination system switched off, this difference being subjected, in turn, to time normalization by dividing by the selected time of exposure or measurement.

As an alternative to the above-described use of an electro-optical sensor surface, such as the CCD array 6, the invention also comprises the use of photosensitive layers as sensor surface, the layer state of which changes detectably upon exposure to the scattered light to be detected. A photoresist layer applied to a wafer or another carrier material can, for example, serve as photosensitive layer. The spatial resolution of typical photoresist layers is so high that the aerial image can be transferred directly to the photoresist layer even without enlargement, that is to say the photoresist layer is preferably introduced into the plane of the aerial image produced. In the case of use in the device from Figure 1 , this means that the photoresist layer is positioned in the image plane of the objective 1 being examined. The field stop 3, the micro-objective 4 and the deflecting mirror 5 can be omitted.

For the purpose of detecting scattered light by means of a photoresist layer or another suitable photosensitive layer, the respective scattered-light measuring structure 11 of the illumination mask 10 is imaged successively onto different regions of the photoresist layer with different illumination intensities or illumination doses. For this purpose, the photoresist layer is displaced suitably laterally from measurement to measurement in a way known per se by a distance which is substantially greater than the dimension, on the side of the aerial image, of the relevant scattered-light measuring structure. The displacement is effected, for example, by means of a conventional wafer holder in which the wafer with the photoresist layer applied thereto is fixed. For each scattered-light measuring structure, a measurement series is undertaken in which the illumination intensity or illumination dose is increased until the photoresist layer structure corresponding to the inner dark-field zone 22 has been completely exposed by the influence of scattered light, that is to say has reached the same layer state as the surrounding photoresist layer annular region, which corresponds to the associated scattered-light marker ring 20. Depending on the type of photoresist layer, this means, for example, that the photoresist layer region corresponding to the inner dark-field zone 21 is firstly retained virtually completely in the case of a very low exposure dosage, whereas the surrounding photoresist layer ring, corresponding to the scattered-light marker ring 20, is developed away, and with a higher exposure dosage the radius of the retained, inner photoresist layer circle becomes ever smaller until this photoresist layer circle has entirely vanished at a certain limiting value of the exposure dosage. The radii of the inner photoresist layer circles retained in a respective measurement series can subsequently be determined in a conventional way, for ex- ample in a scanning electron microscope.

In a simple variant of evaluation, for each scattered-light measuring structure use is made as evaluation variable for the range-resolved determination of scattered light only of the exposure dosage limiting value determined in this way at which the photoresist layer structure corresponding to the inner dark-field zone 21 has firstly entirely vanished. It is sufficient in this case to observe the exposure dosage at which the relevant photoresist layer structure has completely vanished. In alternative evaluation techniques, the radii of the retained, inner photoresist layer circles are determined quantitatively as a function of the illumination intensity or illumination dosage, and taken into account additionally or as an alternative to the limiting value of the illumination dosage during the evaluation. The quality of the evaluation and, in particular, the range separation can thereby be substantially improved. Taking into account the measured relationship of the radii of the remaining photoresist layer structures as a function of the illumination dosage corresponds to the above- described mode of procedure of dividing a pixel field up into a plural- ity of separately evaluated subregions in the case of the CCD array 6.

The determination of the scattered light distribution function from which it is then possible to deduce the optical quality or the optical properties of the test component and, in particular, the scattering behaviour thereof is performed subsequently in the usual way by solving the appropriate inverse scattering problem. The starting point for this purpose is an ansatz which describes the calibrated scattered light measured values as the sum of two components, specifically a first, residual long-range component and a second component which includes the scattered-light distribution to be determined. This second component can be described by a double integral of the scattered-light distribution over the relevant object-side and image-side surface regions, in the case of rotationally symmet- rical systems, for example, by a first radial integration over the surface of the scattered-light marker zone 20 being considered, and a second radial integration over the surface of the annular or circular subregion being considered on the CCD array, that is to say the integrand is the range-dependent scattered-light distribution function which depends on the lateral differential vector between the scattered position and detector position. The contribution of the second component is, furthermore, divided by the total area of the subregion being viewed, for the purpose of normalization. In order to solve the inverse scattering problem, an advantageous mode of procedure in the case of rotational symmetry is to use a radially symmetrical step function of variable step height as an ansatz for the scattered-light distribution function. The unknown step heights, and thus the solution vector of the inverse scattering problem, can then be determined using customary methods for solving inverse scattering problems by means of a least square fit from the calibrated measured values provided with suitable weightings.

Figure 5 illustrates in the form of a diagram the result of a practical simulation which proceeds from the device in accordance with Figures 1 to 4 and the algorithm, discussed above, for solving the inverse scattering problem, illuminating radiation with a wavelength of 193 nm having been assumed. The left-hand partial illustration shows an assumed distribution of the relative scattered-light component as a function of the scattering distance. The middle partial illustration of Figure 5 shows the associated scattering data, calculated by integration, as a characteristic diagram for the respective detector-side ring region j and illumination-side ring region i. Assum- ing a realistic signal-to-noise ratio, these simulated measured values were used to back-calculate the scattered-light distribution by solving the inverse scattering problem using the ansatzes explained above. The associated result, shown in the right-hand partial illustration of Figure 5, makes it clear that the original scattered-light distri- bution is reproduced very well, and the ansatzes used are therefore realistic and justified.

The device according to the invention can be used not only for range-resolved determination of scattered light at projection objec- tives of microlithographic projection exposure apparatuses, but also at any other desired optical systems. Again, it is possible as an alternative to the closed annular shape shown to use scattered-light marker zones of another, in particular also not rotationally symmetrical shape, adapted to the geometry of the optical systems to be ex- amined or of the symmetry of their scattering behaviour, for example the shapes of an annular segment.

Claims

Patent claims

1. Illumination mask for a device for range-resolved determination of scattered light, having one or more scattered-light measuring structures (11) which respectively include an inner dark-field zone (22) which defines a minimum scattering range, wherein the scattered-light measuring structure or at least one of the plurality of scattered-light measuring structures (11) has a scattered-light marker zone (20) in the form of a bright-field zone which on the one hand borders the inner dark-field zone (22) and on the other hand borders an outer dark-field zone (21 ) which defines a maximum scattering range.

2. Illumination mask according to Claim 1 , wherein the illumination mask has a plurality of scattered-light marker zones with inner and/or outer dark-field zones of different sizes.

3. Illumination mask according to Claim 1 or 2, wherein the respective scattered-light marker zone is designed as a ring.

4. Illumination mask according to Claim 3, wherein it has a plurality of scattered-light marker zones designed as rings and having different ring radii.

5. Illumination mask according to one of the preceding claims, wherein the illumination mask is designed as a dark-field mask with the scattered-light marker zone or zones as bright- field zones in the mask dark field.

. Illumination mask according to one of the preceding claims, wherein the scattered-light measuring structures and one or more dark-field area elements (12) are arranged in the form of a matrix in a regular distribution on the illumination mask, all the nearest neighbours of each dark-field area element of the matrix arrangement being scattered-light measuring structures.

7. Device for range-resolved determination of scattered light at a test component, having an illumination mask (20) for providing measuring radiation on an entrance side of the test component, and a detection part (8) for range-resolved detection of scattered light on an exit side of the test component, wherein the illumination mask (8) is one according to one of claims 1 to 6.

8. Device according to Claim 7, wherein it is set up for measurement at an optical imaging system as test component, in which the illumination mask is to be arranged in or near the object plane, the detection part has a field stop (3) to be arranged in or near the image plane, a micro-objective (4) and a sensor surface (6) for scattered-light measurement, and an evaluation part (7) for range-resolved determination of the scattered light caused by the test component is provided by using the scattered-light information obtained by means of the sensor surface.

9. Device according to Claim 7 or 8, wherein the sensor surface is implemented by an electro-optical sensor surface (30) or by a radiation-sensitive layer which upon irradiation is changed detectably from a first layer state to a second layer state.

10. Method for range- resolved determination of scattered light with the aid of a device according to Claim 9, wherein the electro-optical sensor surface (30) is divided into separately driven subregions (33, 34, 35) which correspond to different scattering ranges, and the measured values are picked up individually for each subregion.

11. Method for range-resolved determination of scattered light with the aid of a device according to Claim 9, wherein the same scattered-light measuring structure is imaged onto different regions of the photosensitive layer in various measuring operations with a successive increase in intensity of the measuring radiation, and the range-resolved determination of scattered light is performed using detected state data for these regions of the photosensitive layer and/or using data for the intensity or dose of the measuring radiation.

12. Microlithographic projection exposure apparatus having a device for range-resolved determination of scattered light at a projection objective (1) of the projection exposure apparatus, wherein the device is one according to one of Claims 7 to 9.