DE102018219127A1 - Method and device for correcting aberrations of imaging optics - Google Patents

Abstract

The invention relates to a method (10) for correcting aberrations of an imaging optical system (40) comprising a plurality of optical elements (42, 44, 46, 48), in which correction surfaces (62, 64, 66) in the optical path of the imaging optics (40) Areas of the imaging optics (40) are selected, which change the wavefront of the imaging optics (40) passing through wave. Furthermore, a field-point as well as direction-resolved wavefront error distribution of the imaging optics (40) is measured (12). Furthermore, correction values are determined by tomographic backprojection of the wavefront error distribution onto the correction surfaces (62, 64, 66) and the respective shape of the correction surfaces (62, 64, 66) is modified (30) on the basis of the correction values. Furthermore, the invention relates to a corresponding device for the correction of aberrations.

Description

Background of the invention

The invention relates to a method for correcting aberrations of an imaging optical system comprising a plurality of optical elements, in which a measurement of a field point and direction-resolved wavefront aberration distribution of the imaging optics is performed. Furthermore, the invention relates to a device for correcting aberrations of an imaging optical system with a plurality of optical elements.

Various methods and devices for measuring a field-point and direction-resolved wavefront or a wavefront error distribution are known from the prior art. For example, such a measurement is made using phase-shifting interferometry techniques, such as shearing interferometry. A point diffraction interferometry or point diffraction interferometry known to those skilled in the art is also suitable for the directionally resolved measurement of a wavefront at different field points of a field plane of the imaging optics. In microlithography, for example, corresponding measuring devices are used in imaging optics for imaging mask structures on a substrate for determining wavefront errors in the image plane.

For the correction of aberrations with the aid of a measured wavefront error distribution, conventional methods first of all perform a decomposition of measured wavefronts into a predefined selection of so-called Zernike polynomials and associated Zernike coefficients. Zernike polynomials are basic functions of a linear space and are particularly suitable for describing aberrations in imaging optics. For example, Zernike polynomials are known to those skilled in the art from Chapter 13.2.3 of the textbook "Optical Shop Testing, 2nd Edition (1992) by Daniel Malacara, Ed. John Wiley & Sons, Inc."

A correction in an imaging optics can be done, for example, by means of a change in bearing by means of manipulators or a corresponding processing of one or more optical surfaces of optical elements of the imaging optics. An effect of manipulation paths x in manipulators or a corresponding shaping processing is usually described in the space of wavefront changes with a sensitivity matrix M. In order to determine suitable adjustment paths or shape changes for a correction, in known methods with a measured wavefront error distribution a, the following is a minimization of an objective function: $$\text{min}\frac{1}{2}{\Vert Mx-a\Vert }^2=\text{min}\left(\frac{1}{2}{x}^T{M}^{T}Mx-{\left(M^{T}a\right)}^{T}x\right)$$

A disadvantage of such an optimization with a decomposition of a wavefront in Zernike polynomials is that these polynomials are only poorly suited for high-frequency interference and in particular periodic high-frequency interference. Such errors can occur, for example, with optical elements having a substrate material in which the hardness and thus also a material removal behavior vary periodically. Another cause of these errors can be artifacts of an interferometric pass measurement technique. Determined Zernike coefficients in these cases are very sensitive to noise, so that the solution of the inverse problem is poorly conditioned and by noise amplification a very small correction effect can occur.

Underlying task

It is an object of the invention to provide a method and a device, with which the aforementioned problems are solved, and in particular a better correction of aberrations is achieved in an imaging optics.

Inventive solution

The above object can be achieved according to the invention, for example, with a method for correcting aberrations of an imaging optical system comprising a plurality of optical elements, wherein the correction surfaces in the beam path of the imaging optics are selected from the optical surfaces of the imaging optics, which change the wavefront of a wave passing through the imaging optics.

Furthermore, in the method according to the invention, a field-point as well as direction-resolved wavefront error distribution of the imaging optics are measured and correction values are determined by tomographic backprojection of the wavefront error distribution onto the correction surfaces. Finally, the respective shape of the correction surfaces is changed based on the correction values.

As correction surfaces, for example, optical surfaces which are easy to machine or which can be adjusted with the aid of manipulators in their optical properties are used. For measuring a wavefront error of the imaging optics, phase-shifting interferometry techniques, such as shear interferometry or point diffraction interferometry, can be used, for example. With a tomographic back projection of the measured Wavefront error distribution is called a back projection along the beam path of the imaging optics to different correction surfaces at different locations in the imaging optics.

The shape change of a correction surface is carried out according to an embodiment by means of a suitable tool for material removal, such as a polishing pad, a magnetorheological tool, an electron beam or an ion beam. Preferably, the machining with the tool is performed by means of a corresponding tool function, e.g. a rotationally symmetric Gaussian function with a characteristic width described. In an alternative embodiment, instead of a removal of material, a travel change of a manipulator and thus a change in position of the correction surface of an optical element of the imaging optics occur.

With the backprojection and the use of the information of the radiation propagation in the imaging optics according to the method according to the invention, a determination of correction values for the reduction of aberrations which is much more robust than conventional methods and better adapted to higher-frequency errors is achieved.

According to an embodiment of the invention, the correction surfaces are each divided into correction sections, the imaging beam path of the imaging optics is simulated, and individual beams of the imaging beam path are assigned to the correction sections such that each of the associated individual beams undergoes a characteristic combination of the correction surfaces. By the term "passing through" of correction surfaces or optical surfaces, in addition to passing an optical surface, for example a surface of a lens, a reflection on an optical surface, e.g. at a mirror, or a diffraction at an optical surface and thus generally understood an interaction of a single beam with an optical surface or a correction surface. The number of correction sections of a correction surface in a lateral direction may, according to one embodiment, be greater than fifty, in particular greater than one hundred. According to one embodiment variant, the total number of correction sections per correction surface in both lateral dimensions, i. in the area, a few thousand. A division of a correction surface takes place, for example, in equally sized correction sections, which cover the entire correction surface. Alternatively, a division into different sized correction sections, not covering the entire correction surface covering correction sections or both. Correction sections can also be determined on the basis of selected individual beams which pass through the respective correction section.

According to a further embodiment according to the invention, one of the individual beams is identified as a faulty single beam by assigning an error value from the wavefront error distribution to the identified single beam, and correction values are assigned to correction sections traversed by the identified single beam. Each individual beam is defined by the field point at which it has a field plane of the imaging optics, e.g. the image plane, passes through, and the angle of incidence clearly defined. With the measured field-point and direction-resolved wavefront error distribution, it is thus possible to associate a possibly present wavefront error with each individual beam.

According to one embodiment, the field-point as well as direction-resolved wavefront error distribution of the imaging optics is measured with respect to a desired wavefront. Furthermore, identifying a single beam, assigning correction values, calculating a corrected target wavefront by simulating the beam path based on the imaging optics provided with the correction values, and determining a corrected wavefront error distribution of the imaging optics with respect to the corrected target wavefront are iteratively performed until the wavefront error distribution falls below a predetermined threshold. Furthermore, a change of the respective shape of the correction surfaces takes place on the basis of the correction values which are present when the predefined threshold is exceeded by the corrected wavefront error distribution.

Thus, in an iteration, an assignment of correction values to respective correction sections of correction surfaces is first carried out with the aid of a backprojection of faulty individual beams, and then in a forward calculation a simulation of the beam path with correspondingly corrected correction surfaces. In this case, according to one embodiment, a specification of possible corrections can be carried out by means of a tool function. The corrected target wavefront determined in the simulation is used to determine a corrected wavefront error distribution for the next iteration. For example, to determine the corrected wavefront error distribution, a comparison of the corrected nominal wavefront with the measured wavefront takes place.

In a further embodiment of the invention are in Assigning a correction value to a correction section takes into account the error value associated with a single beam and the correction surfaces traversed by the single beam. Preferably, in one embodiment, the number of swept correction areas is taken into account. Further, an appropriate weighting of the swept correction areas may be performed upon assigning a correction value to a correction portion of a correction area.

und somit für eine Wellenfrontabweichung pro Korrekturfläche WFR_fl(x, y; k_x, k_y) $${\text{WFR}}_{\text{fl}}\left(x,y;k_x,k_y\right)=-g_{\text{fl}}\text{WFR}\left(x,y;k_x,k_y\right)$$

In this case, according to one embodiment, the error value associated with a single beam is distributed among the correction surfaces traversed by the individual beam. For example, a wavefront deviation WFR (x, y; k _x , k _y ) to be corrected with the field point coordinates (x, y) and the pupil coordinates (k _x , k _y ) is distributed uniformly over all the correction surfaces passed through. With the number N of correction surfaces traversed, this results in a distribution factor g _fl $$G_{fl}=\frac{1}{N}$$

and thus for one wavefront deviation per correction surface WFR _fl (x, y; k _x , k _y ) $${\text{WFR}}_{\text{fl}}\left(x.y;k_x.k_y\right)=-G_{\text{fl}}\text{WFR}\left(x.y;k_x.k_y\right)$$

erfolgen. Die Summe über alle Wellenfrontabweichung pro Korrekturfläche kompensiert somit den Gesamtfehler.In alternative embodiments may also be another suitable distribution with $${\displaystyle \underset{\text{fl}=1}{\overset{N}{\Sigma }}{G}_{\text{fl}}=1}$$

respectively. The sum over all wavefront deviation per correction surface thus compensates for the total error.

According to an embodiment of the invention, when assigning a correction value to a correction section, a plurality or all of the individual beams passing through the correction section are taken into account. For example, there is an addition of the correction wave front portions Δp _{fl of} all the individual beams passing through the correction section. The correction wavefront component Δp _{fl of} a single beam corresponds, for example, to the portion of the wavefront deviation assigned to the correction surface: $$\mathrm{\Delta }{p}_{\text{fl}}\left(j.k\right)=-G_{\text{fl}}\text{WFR}\left(x.y;k_x.k_y\right)$$

In one embodiment according to the invention, a suitable weighting of the error values of the individual beams passing through a correction section takes place for an assignment of a correction value to the correction section. In the weighting of the error values, for example, the number of individual beams passing through the correction section can be taken into account. According to one embodiment, the weighting is averaging of the individual error values over the number of individual beams.

According to a further embodiment of the invention, the imaging optics comprise a plurality of optical surfaces which change the wavefront of a wave passing through the imaging optics, and the correction surfaces comprise only a subset of the optical surfaces. This means that not all optical surfaces of the imaging optics serve as correction surfaces. In a sense, the imaging optics for correction to the correction surfaces is reduced. For example, the optical surfaces in a mirror objective comprise the mirror surfaces and in the case of a lens objective the lens surfaces. Furthermore, the optical surfaces may also include areas of correction elements introduced specifically for the wavefront correction into the beam path of the imaging optics. Of all optical surfaces, only one part is used as correction surfaces.

According to an embodiment of the invention, the individual beams are assigned to correction sections such that the combination of correction sections of a respective single beam differs from the respective combination of the remaining individual beams. Thus, each individual beam is assigned a combination of correction sections on different correction surfaces that is not assigned to any other single beams. For example, individual beams with a same combination of correction sections can be combined to form a single beam and an error value can be assigned to this single beam based on the error values of the combined individual beams.

According to a further embodiment of the invention, the error value associated with the identified individual beam is the value of the wavefront error distribution which corresponds to the field and direction coordinates of the identified individual beam in the image plane of the imaging optics. In other words, in the assignment of the error value to the identified individual beam, that value of the field point and direction resolved wavefront error distribution is determined, which is present at the position coordinate and the direction coordinate of the intersection point of the individual beam with the image plane, and assigned to the individual beam. The wavefront error distribution is determined, for example, as a deviation of a location-dependent and direction-dependent wavefront measured in the image plane from a desired wavefront.

According to an embodiment of the invention, when an adjustment value is assigned to a correction section, error values of individual beams are weighted lower or not taken into account with respect to other individual beams of lower measurement accuracy. For example, a limit for a measurement accuracy is specified and at a measurement accuracy below this limit, the corresponding error value against error values of other individual beams less weighted or not considered. In other embodiments, filter operations to reduce noise effects during an iteration for determining correction values may additionally or alternatively be provided.

According to one embodiment of the invention, the imaging optics is configured as a projection objective of a projection exposure apparatus for microlithography. In microlithography during the exposure process, the projection objective is used to image mask structures onto a substrate in the form of a wafer. In this case, high demands are placed on the imaging quality of the projection objective for the most error-free imaging of the mask structures. The projection lens should have very little wavefront aberrations.

Furthermore, in another embodiment of the invention, the imaging optics is designed for operation with EUV radiation. Ultraviolet wavelength (EUV) radiation is electromagnetic radiation having a wavelength less than 100 nm. In particular, the imaging optics are configured for radiation having a wavelength of about 13.5 nm or about 6.7 nm. The imaging optics therefore essentially comprises mirrors as optical elements. Compared to imaging optics for radiation in another, longer-wavelength spectral range, EUV imaging optics usually have significantly fewer optical elements or optical surfaces. Backprojection of individual beams onto correction surfaces for correction of aberrations can therefore be carried out quickly and with high accuracy.

The aforementioned object can furthermore be achieved, for example, with a device for correcting imaging aberrations of an imaging optical system with a plurality of optical elements. The apparatus comprises a measuring device for the field-point and direction-resolved measurement of a wavefront error distribution of the imaging optics, a determination module for determining correction values by tomographic backprojection of the wavefront error distribution on correction surfaces, wherein the correction surfaces are selected from optical surfaces of the imaging optics which change the wavefront of a wave passing through the imaging optics, and a device for changing the respective shape of the correction surfaces based on the correction values.

Analogous to the method according to the invention, the backprojection of the wavefront error distribution along the beam path to different correction surfaces at different locations in the imaging optics and thus the use of the radiation propagation information in the imaging optics achieves a better correction of aberrations than conventional devices.

The features specified with regard to the above-mentioned embodiments, exemplary embodiments or design variants, etc. of the correction method according to the invention can be correspondingly transferred to the correction device according to the invention. These and other features of the embodiments according to the invention are explained in the description of the figures and the claims. The individual features can be realized either separately or in combination as embodiments of the invention. Furthermore, they can describe advantageous embodiments which are independently protectable and their protection is possibly claimed only during or after pending the application.

list of figures

• 1 ein schematisches Flussdiagramm zur Veranschaulichung eines Ausführungsbeispiels des erfindungsgemäßen Verfahrens,
• 2 eine schematische Darstellung einer beispielhaften Abbildungsoptik bei einer Korrektur von Abbildungsfehlern mit dem Ausführungsbeispiel nach 1, sowie
• 3a bis 3e eine Veranschaulichung von Verfahrensschritten des Ausführungsbeispiels nach 1 bei einer Korrektur der Abbildungsoptik gemäß 2.

The foregoing and other advantageous features of the invention are illustrated in the following detailed description of exemplary embodiments according to the invention with reference to the accompanying diagrammatic drawings. It shows:

• 1 a schematic flow diagram illustrating an embodiment of the method according to the invention,
• 2 a schematic representation of an exemplary imaging optics in a correction of aberrations with the embodiment according to 1 , such as
• 3a to 3e an illustration of method steps of the embodiment according to 1 in a correction of the imaging optics according to 2 ,

Detailed description of inventive embodiments

In the embodiments or embodiments or design variants described below, functionally or structurally similar elements are as far as possible provided with the same or similar reference numerals. Therefore, for the understanding of the features of the individual elements of a particular embodiment, reference should be made to the description of other embodiments or the general description of the invention.

To simplify the description, some drawings show Cartesian xyz Coordinate system specified, from which the respective positional relationship of the components shown in the figures results. In 2 the y-direction is perpendicular to the plane out of this, the x-direction to the right and the z-direction down.

In 1 becomes a procedure 10 for correcting aberrations of imaging optics in a schematic flow diagram. The procedure 10 is particularly suitable for correcting aberrations in projection lenses of EUV microlithography. With a projection objective, mask structures are imaged on a substrate in the form of a wafer in microlithography. With ever decreasing structures, an image with the lowest possible aberrations is required. EUV microlithography uses electromagnetic radiation in the extreme ultraviolet (EUV) wavelength range to illuminate the substrate. The wavelength is preferably less than 100 nm and is for example about 13.5 nm or about 6.7 nm. In such imaging optics essentially mirrors are used as optical elements. Furthermore, the number of optical elements used is usually smaller than with projection lenses for longer wavelengths.

In the process 10 a measurement takes place 12 a field point and direction-dependent wavefront error distribution in a field plane, such as the image plane of the imaging optics. For this purpose, a measuring method known to those skilled in the art for the determination of wavefronts, such as a phase-shifting interferometry, for example a shear or shearing interferometry or a point diffraction interferometry, is used. The wavefront error distribution can be determined, for example, by means of a comparison of the wavefront measured at different field points of the field plane with a desired wavefront. The measured, field point and direction resolved wavefront error distribution or wavefront deviation is also referred to here as WFR (x, y; k _x , k _y ) with the field point coordinates (x, y) and the direction or pupil coordinates (k _x , k _y ).

Furthermore, in the method 10 a provision 14 performed by correction surfaces. The imaging optics comprises a multiplicity of optical surfaces which change the wavefront of a wave passing through the imaging optics. Examples of such optical surfaces are reflecting surfaces of mirrors, surfaces of lenses or prisms, or surfaces with diffractive structures. Among these optical surfaces, only a part is determined as correction surfaces. The correction surfaces thus represent a subset of the optical surfaces of the imaging optics. In a selection of correction surfaces, inter alia, a suitability of the correction surface for surface processing or for position or shape change by manipulators, or their correction potential for certain aberrations can be considered.

Each correction surface is in one process step 16 divided into correction sections. The division of a correction surface into correction sections takes place, for example, taking into account puncture points of relevant individual beams. In this case, a division into equal or different sized correction sections take place, which cover the entire correction surface or only a part thereof.

The correction sections of a correction surface fl are also referred to below as p _fl (j, k), where j and k represent the position coordinates of the correction section on the associated correction surface.

Subsequently, an assignment in each case of an error value of the measured wavefront error distribution WFR (x, y; k _x , k _y ) to a single beam is performed. The field point coordinates x, y and direction coordinates k _x , k _y in each case uniquely identify a single beam through the imaging optics. In addition, each individual beam is assigned to the correction section p _fl (j, k) passed through by the single beam with the aid of a back projection at each correction area. Overall, in a process step 18 hence the following assignment: $$\text{WFR}\left(x.y;k_x.k_y\right)\leftrightarrow \text{beam}\leftrightarrow {\text{p}}_{fl}\left(j.k\right)=f\left(\text{fl}.x.y.k_x.k_y\right)$$

The function f thus assigns each single-beam-resolved wavefront error distribution to an error distribution on suitable correction sections. In this case, a translation factor of a surface deviation Δh can additionally be assigned to the respective correction section. For example, for a mirror, ΔWFR = -2 Δh cos (a), where α denotes the angle of incidence relative to the local surface solder.

und somit für eine Wellenfrontabweichung pro Korrekturfläche WFR_fl(x, y; k_x, k_y) $${\text{WFR}}_{\text{fl}}\left(x,y;k_x,k_y\right)=-g_{\text{fl}}\text{WFR}\left(x,y;k_x,k_y\right)$$

After a process step 20 in which the correction values are initialized at each correction section, for example by setting the correction values equal to zero, in one method step 22 an addition of a weighted error value or wavefront contribution for all measured field-dependent wavefront positions and performed at each correction section. In this exemplary embodiment, the weighting used is a uniform distribution of the error value assigned to a single beam to all the correction surfaces. With the number N of correction surfaces passed, the weighting factor g _fl results $$G_{fl}=\frac{1}{N}$$

and thus for one wavefront deviation per correction surface WFR _fl (x, y; k _x , k _y ) $${\text{WFR}}_{\text{fl}}\left(x.y;k_x.k_y\right)=-G_{\text{fl}}\text{WFR}\left(x.y;k_x.k_y\right)$$

erfolgen. Die Summe über alle den Korrekturflächen zugeordneten Wellenfrontabweichung kompensiert somit den Gesamtfehler. Die gewichtete Wellenfrontabweichung WFR_fl (x, y; k_x, k_y) eines Einzelstrahls wird dem Korrekturabschnitt der Korrekturfläche als Korrekturanteil bzw. Korrekturwert Δp_fl hinzuaddiert, welche von dem Einzelstrahl durchstoßen wird: $$\mathrm{\Delta }{p}_{\text{fl}}\left(j,k\right)=-g_{\text{fl}}\text{WFR}\left(x,y;k_x,k_y\right)$$

In alternative embodiments may also be another suitable distribution with $${\displaystyle \underset{\text{fl}=1}{\overset{N}{\Sigma }}{G}_{\text{fl}}=1}$$

respectively. The sum over all the wavefront deviations associated with the correction surfaces thus compensates for the total error. The weighted wavefront deviation WFR _fl (x, y; k _x , k _y ) of a single beam is added to the correction portion of the correction surface as the correction amount Δp _fl pierced by the single beam: $$\mathrm{\Delta }{p}_{\text{fl}}\left(j.k\right)=-G_{\text{fl}}\text{WFR}\left(x.y;k_x.k_y\right)$$

For a correction section, correction components of different wavefronts or individual beams can be added up and the sum used as a correction value. Furthermore, in this embodiment, in the summation of correction portions for a correction section, the number of passing single beams is taken into consideration. Such weighting will be discussed below with reference to 2 and 3 explained in more detail.

und der Korrekturwert Δp_fl (j, k) durch diesen Normierungswert G_fl geteilt. Feldpunkte mit unterschiedlichen Bildfehlergewichten können beispielsweise bei Abbildungsoptiken in der Mikrolithographie durch unterschiedliche Helligkeiten auftreten, etwa bei einer Beleuchtungsrampe entlang einer Scanlinie.In the event that in a measurement of the wavefront error distribution field points have different image error weights, the field weights are multiplied according to an embodiment with the respective wavefront deviations. For each correction section, the sum of the field weights g _{fl of} the individual beams passing through is created for normalization $${\text{G}}_{fl}\left(j.k\right)={\displaystyle {\Sigma }_{StraHlenauf{p}_{fl}\left(j.k\right)}{G}_{fl}\left(j.k\right)}$$

and the correction value Δp _fl (j, k) is divided by this normalization value G _fl . Field points with different image error weights can occur, for example, in imaging optics in microlithography by different brightnesses, such as a lighting ramp along a scan line.

With the determined correction values and a correction function of a tool used is in a next step 24 calculates an expected surface change Δh _fl ^Korr (x, y) for each correction surface. In this case, in the case of a material-removing tool, for example a polishing tool, a magnetorheological tool or an electron, ion or plasma jet, a rotationally symmetric Gaussian function having a characteristic width β can be used as correction function or tool function w (Δx, Δy): $$\begin{array}{ccc}w\left(\mathrm{\Delta }r\right)=k{e}^{-{\left(\frac{\mathrm{\Delta }r}{\beta }\right)}^2}& \text{With}& \mathrm{\Delta }r=\sqrt{\mathrm{\Delta }{x}^2+\mathrm{\Delta }{y}^2}\end{array}$$

The calculation of an expected surface change may further include occupying the correction surface with the current correction values and a convolution of the resulting spatially distributed correction values with the tool function. The calculated surface change Δh _fl ^Korr (x, y) is optionally added to an already calculated surface change according to a previous iteration.

Subsequently, in a process step 26 performed in a forward calculation, a calculation of corrected wavefronts, as expected from the measured wavefronts and the correction surfaces with updated surface changes. It follows in one process step 28 a check whether the corrected wavefronts meet a given quality criterion. For example, it can be checked whether selected Zernike values or RMS values are below predefined tolerance values. In other words, it is checked whether a residual error of the corrected wavefronts is acceptable.

If this applies, a corresponding processing of the correction surfaces takes place according to the calculated surface changes. Furthermore, the method is aborted if a maximum number of iterations has been reached or calculated surface changes are below a predetermined convergence threshold. If no abort criterion is met, another iteration takes place, see arrow 32 , This will be in one step 34 used the simulated corrected wavefront values as new input values for an updated wavefront aberration.

2 shows a schematic view of an exemplary imaging optics 40 with a correction of aberrations with the described method 10 , The imaging optics 40 exemplarily includes four optical elements 42 . 44 . 46 . 48 with which a mapping of structures in an object plane 50 on an image plane 52 the imaging optics 40 he follows. In 2 are three field points FP1 . FP2 . FP3 the object level 50 shown, which in the picture plane 52 be imaged.

For every field point FP1 . FP2 . FP3 are in 2 the beam paths of three exemplarily selected individual beams are shown. The from the field point FP1 in the object plane 50 outgoing single rays 54 are shown as solid lines from the field point FP2 outgoing single rays 56 as dashed lines, and from the field point FP3 outgoing single rays 58 as dot-dashed lines. Below the picture plane 52 are in 2 for each field point FP1 . FP2 . FP3 the different directional or pupil coordinates of the individual beams 54 . 56 . 58 as spatial coordinates in a pupil plane 60 shown.

According to the method described above 10 For correction of aberrations, three optical surfaces of the imaging optics are used as correction surfaces 62 . 64 , 66 has been determined. Thus, for example, optical surfaces of the fourth optical element serve 48 not as correction surfaces. The correction surfaces 62 . 64 . 66 thus provide only a subset of all optical surfaces of the imaging optics 40 represents.

Each of the correction surfaces 62 . 64 . 66 is in three correction sections K1 . K2 . K3 divided with different size. It is clear in 2 to recognize that the first correction section K1 the first correction surface 62 only from a single beam 54 of the first field point FP1 is going through. Through the second correction section K2 the first correction surface, however, run the other two individual beams of the first field point FP1 , all three rays 56 of the second field point FP2 and two single beams 58 of the third field point FP3 , The third correction section K3 the first correction surface 62 is only a single beam 58 of the third field point FP3 run through. Accordingly, in 2 recognizable, which individual beams respectively the correction sections of the second and third correction surface 64 , 66 go through.

In 3a to 3e become individual process steps of the embodiment after 1 in a correction of the imaging optics according to 2 illustrated. For this, first, as in 3a presented a unit disorder 68 in the first optical element in the region of the first correction section K1 at the first correction surface 62 accepted. This area is only used by a single beam 54 of the first field point FP1 run through. An in 3b Measurement of the wavefront error distribution shown in pupil coordinates therefore results in a single beam only for this one 54 at the first field point FP1 a wavefront error 70 ,

3c illustrates an assignment of correction values to correction sections K1 . K2 . K3 , Since the measured wavefront error distribution does not reveal which optical element 42 . 44 . 46 . 48 the error 70 causes the unit error 70 on all three correction surfaces 62 . 64 . 66 distributed equally. As already described, a uniform distribution of the individual beam thus takes place as a weighting 54 assigned error value 70 on all correction surfaces 62 . 64 . 66 , Each first correction section K1 The correction surface first becomes the correction value 1 / 3 assigned.

In addition, the number of individual beams is taken into account by means of averaging 54 . 56 . 58 which is a correction section K1 . K2 . K3 run through. The faulty single beam 54 goes through only the first correction section K1 the first correction surface 62. This correction section K1 thus becomes a correction value 1 / 3 assigned. The first correction section K1 the second correction surface 64 is next to the single beam 54 also from a single beam 58 of the third field point FP3 run through. For this single jet 58 no error was measured. The averaging (1/3 + 0) / 2, ie the sum of the error values of the individual beams divided by the number of individual beams in a correction section, yields for the first correction section K1 the second correction surface 64 the correction value 1/6.

Accordingly, the first correction section becomes K1 the third correction surface 66 from the faulty single beam 54 of the first field point FP1 and each of a faultless single beam 56 . 58 of the second and third field points FP2 . FP3 run through. An averaging results here thus ( 1 / 3 + 0 + 0) / 3 = 1/9 as a correction value for the first correction section K1 the third correction surface 66 , In the case of all other correction sections, this is during the initialization in the method step 20 assigned correction value zero not changed. These correction values occur in the method step 26 a determination of corrected wavefronts.

3d shows the change in the wavefronts caused by the correction values, which results in the determination of corrected wavefronts for each individual beam 54 . 56 . 58 occur. The in 3d left single beam 54 of the first field point FP1 goes through all three first correction sections provided with a correction value K1 , The left single beam 56 of the second field point FP2 goes through the second correction sections K2 the first and second correction surfaces 62 . 64 without correction value and then the first correction section K1 the third correction surface 66 with the correction value 1/9. The left single beam 58 of the third field point FP3 in particular, goes through the first correction sections K1 the second and third correction surfaces 64 , 66.

3e shows the result of a subtraction of the newly reconstructed corrected wavefront according to FIG 3d from the measured wavefront 3b , This result is used as input for a re iteration. In the process, an equal distribution of the error values on all three correction surfaces is again achieved 62 . 64 . 66 and then, as an average in each correction section, the sum of the correction values is divided by the number of individual beams passing through. For a re-calculation of corrected wavefronts, the correction values of all previous iterations are first added at each correction section and these values are used in a wavefront simulation.

The iterations are carried out until the process step 28 check for an acceptable residual error in the corrected wavefronts or a maximum number of iterations. With an acceptable residual value, the correction values of each correction section are finally used for processing the correction surfaces.

The above description of exemplary embodiments, embodiments or embodiments is to be understood as an example. The disclosure thus made makes it possible for the skilled person, on the one hand, to understand the present invention and the associated advantages, and on the other hand, in the understanding of the person skilled in the art, also encompasses obvious modifications and modifications of the structures and methods described. It is therefore intended that all such alterations and modifications as fall within the scope of the invention as defined by the appended claims, as well as equivalents, be covered by the scope of the claims.

LIST OF REFERENCE NUMBERS

10

Method of correcting aberrations

12

Measuring wavefront error distribution

14

Determine correction surfaces

16

Dividing correction sections

18

Assignment error value - single beam - correction section

20

Initialization correction values

22

Add correction values

24

Calculate surface change

26

Determine corrected wavefronts

28

Check the residual error

30

Editing the correction surfaces

32

iteration

40

imaging optics

42, 44

first, second optical element

46, 48

third, fourth optical element

50

object level

52

image plane

54

Single beams FP1

56

Single beams FP2

58

Single beams FP3

60

pupil plane

62, 64, 66

first, second, third correction surface

68

disorder

70

measured wavefront error

FP1-FP3

field points

K1, K2, K3

correcting sections

Claims (15)

A method (10) for correcting aberrations of an imaging optic (40) comprising a plurality of optical elements (42, 44, 46, 48), in which: - Correction surfaces (62, 64, 66) in the beam path of the imaging optics (40) are selected from optical surfaces of the imaging optics (40) which change the wavefront of a wave passing through the imaging optics (40), a field-point as well as direction-resolved wavefront error distribution of the imaging optics (40) is measured (12), Correction values are determined by tomographic backprojection of the wavefront error distribution onto the correction surfaces (62, 64, 66), as well as - The respective shape of the correction surfaces (62, 64, 66) is changed based on the correction values (30). Method according to Claim 1 in which: - the correction surfaces (62, 64, 66) are each divided into correction sections (K1, K2, K3), - the imaging beam path of the imaging optics (40) is simulated, and - individual beams (54, 56, 58) of the imaging beam path are assigned to the correction sections (K1, K2, K3) such that each of the associated individual beams (54, 56, 58) undergoes a characteristic combination of the correction surfaces (62, 64, 66). Method according to Claim 2 in which one of the individual beams (54, 56, 58) is identified as a faulty single beam (54) by assigning an error value from the wavefront error distribution to the identified single beam (54), and correction values to correction sections (54) passed through the identified single beam (54) K1) (18). Method according to Claim 3 in which the field point and direction resolved wavefront error distribution of the imaging optics (40) is measured (12) with respect to a desired wavefront, identifying a single beam (54), assigning (18) correction values, calculating a corrected target wavefront by simulating the beam path based on the imaging optics (40) provided with the correction values, and determining (26) a corrected wavefront aberration distribution of the imaging optics (40) with respect to the corrected nominal wavefront iteratively executed until the wavefront error distribution falls below a predetermined threshold, and the respective shape of the correction surfaces (62, 64, 66) is changed (30) on the basis of the correction values which are present when the corrected wavefront error distribution falls below the predetermined threshold. Method according to Claim 3 or 4 in that, when assigning (18) a correction value to a correction section (K1, K2, K3), the error value associated with a single beam (54, 56, 58) and the correction surfaces (62, 64, 66). Method according to Claim 5 in which the error value assigned to a single beam (54, 56, 58) is split between the correction surfaces (62, 64, 66) passed through by the single beam (54, 56, 58). Method according to one of Claims 3 to 6 , wherein upon assignment of a correction value to a correction section (K1, K2, K3) several or all of the correction section (K1, K2, K3) passing through individual beams (54, 56, 58) are taken into account. Method according to Claim 7 wherein a suitable weighting of the error values of the individual beams (54, 56, 58) passing through a correction section (K1, K2, K3) for an assignment of a correction value to the correction section (K1, K2, K3) takes place. The method of any one of the preceding claims, wherein the imaging optics (40) comprises a plurality of optical surfaces that change the wavefront of a wave passing through the imaging optic (40), and the correction surfaces (62, 64, 66) comprise only a subset of the optical surfaces , Method according to one of Claims 2 to 9 in which the individual beams (54, 56, 58) are assigned to correction sections (K1, K2, K3) such that the combination of correction sections (K1, K2, K3) of a respective single beam (54, 56, 58) differs from the respective one Combination of the remaining individual beams (54, 56, 58) differs. Method according to one of Claims 3 to 10 wherein the error value associated with the identified single beam (54) is the value of the wavefront error distribution corresponding to the field and direction coordinates of the identified single beam (54) in the image plane of the imaging optics (40). Method according to one of Claims 3 to 11 In the case of an assignment of a correction value to a correction section (K1, K2, K3), error values of individual beams are weighted lower or not taken into account with less accuracy than other individual beams (54, 56, 58). Method according to one of the preceding claims, in which the imaging optics (40) is configured as a projection objective of a projection exposure apparatus for microlithography. Method according to one of the preceding claims, in which the imaging optics (40) are designed for operation with EUV radiation. An apparatus for correcting aberrations of an imaging optic (40) having a plurality of optical elements (42, 44, 46, 48), comprising: a measuring device for field-point and direction-resolved measurement of a wavefront error distribution of the imaging optics, a determination module for determining correction values by tomographic backprojection of the wavefront error distribution on correction surfaces (62, 64, 66), the correction surfaces being selected from optical surfaces of the imaging optics (40) which change the wavefront of a wave passing through the imaging optics (40), and - Means for changing the respective shape of the correction surfaces based on the correction values.

Images