WO2012103933A1

WO2012103933A1 - Method and apparatus for correcting errors in an euv lithography system

Info

Publication number: WO2012103933A1
Application number: PCT/EP2011/051374
Authority: WO
Inventors: Markus Mengel; Franz-Josef Stickel; Ulrich Wegmann; Bernd DÖRBAND; Sergey Oshemkov
Original assignee: Carl Zeiss Smt Gmbh; Pixer Technology Ltd.
Priority date: 2011-02-01
Filing date: 2011-02-01
Publication date: 2012-08-09

Abstract

The invention relates to a method for correcting a local surface defect (430, 440, 530, 540) of a reflective optical element (170, 171, 172, 173, 74, 175, 200, 300, 400, 500, 700, 800) of an extreme ultraviolet lithography system (100) comprises the steps of analyzing the local surface defect (430, 440, 530, 540), correcting the local surface defect (430, 440, 530, 540) by focusing femto- or picosecond light pulses of a laser system (600) onto positions of the local surface defect (430, 440, 530, 540), and verifying that a wave front sensor (190) associated with the extreme ultraviolet lithography system (100) can no longer detect the local surface defect (430, 440, 530, 540).

Description

Method and apparatus for correcting

in an EUV lithography system

Field of the invention

The present invention relates to the field of correcting errors in am EUV lithography system.

2. Background of the invention

As a result of the steadily increasing integration density of integrated circuits (ICs), photolithographic masks have to transfer smaller and smaller features to a photoresist arranged on a wafer. To meet this demand, the actinic wavelength of the imaging tool has been reduced in steps from the near ultraviolet across the mean ultraviolet into the far ultraviolet region of the electromagnetic spectrum. Furthermore, immersion lithography has been introduced to enlarge the numerical aperture of the projection systems. In order to use significantly smaller actinic wavelengths in lithographic illumination systems, extreme ultraviolet (EUV) lithography systems for wavelengths at approximately 13.5 nm are presently in development. Presently there are no transmissive optical elements available for the

EUV wavelength range, EUV lithography systems will most probably use reflective optical elements to form and direct the optical beam from the EUV source to the photolithographic mask and from the mask to the wafer. In a lithography system the optical elements of the projection sys- tern directing the EUV light from the mask onto the wafer are of particular interest as the EUV beam striking on the components of the projection system already carries the spatial information of the photomask pattern. The errors of the optical elements forming the projection systems have to be very small in order not to impair the pattern information of the mask at the imaging plane of the projection system of the photolithographic system.

Imperfections of the surface topography of optical elements lead to various imaging errors and to a decreasing transmission of the optical system. Typically, these deficiencies are characterized by three parameters. They are separated by their scale across the surface of the optical element:

(a) Figure errors have a spatial frequency down to approximately one tenth of the clear aperture of the optical element. They mainly influence aberrations and thereby the fidelity of the image (e.g. circular holes become elliptical).

(b) Mid spatial frequency (MSF) errors extend from the figure error scale to approximately 1 μηι. These errors result in intrafield scattering, thereby decreasing the contrast of the image.

(c) High spatial frequency (HSR) errors have a spatial wavelength from the end of the MSF range down to approximately the actinic wavelength of the lithographic illumination system. These surface imperfections of the optical element lead to greater angular deviations of the reflected photons from the local reflection direction and thus disturb the proper constructive overlay of the sub-wave fronts reflected from the multilayer coating on the optical element and result in a drop of the reflectivity of the optical element.

The surface topography of an optical element is measured across the various regions of the above mentioned spatial frequency ranges by using different metrology tools adapted to the different spatial frequencies The measured surface topology is then compared with a predetermined surface form of the optical element and the local deviations of the sur- face are averaged across the surface of the optical element and expressed as a route mean square (RMS) value for the respective spatial frequency range. In order to ensure the capability of the projection system, numerical values for figure, MSF and HSF errors are specified for the overall projection system. From this specification, the specification of the individual optical elements forming the projection system is derived. The projection system comprises only optical elements which fulfil each of the figure, MSF and HSF specifications.

However, due to the statistical definition of the individual specification parameters, an optical element may still have individual local surface defects, as this or these defects add contributions to the figure, MSF and HSF errors which are below the specified threshold. As the area of the local defect is very small compared to the imaging area of the overall optical element, the contribution of the local surface defect(s) to the imaging capability of the projection system is negligible. Modern lithographic systems have a wave front sensor which can be placed in the imaging plane of the projection system instead of the wafer. This sensor analyzes the wave front of the optical beam normally illuminating the photoresist on the wafer. The obtained data are used to control the imaging power of the projection system. Further, this data en- ables the optimization of the figure correction of the projection system.

Local surface defects of individual optical elements of the projection system can be recognized in the data of the wave front sensor. The user of the lithographic system may be worried about this data. Moreover, the detected defect(s) may hamper the control and the adjustment of the lithography system in particular of the projection system. In certain critical cases the local surface defect(s) can also impair the image quality of the projection system. It is therefore one object of the present invention to provide a method and an apparatus for minimizing the effects of local surface defects of an optical element of a EUV lithography system at its imaging plane in order to avoid at least in part the above mentioned problems.

3. Summary of the invention

According to a first aspect of the invention, a method according to patent claim l is provided. In an embodiment, a method for correcting a local surface defect of a reflective optical element of an extreme ultraviolet lithography system comprises the steps of (a) analyzing the local surface defect, (b) correcting the local surface defect by focusing femto- or picosecond light pulses of a laser system onto positions of the local surface defect, and (c) verifying that a wave front sensor associated with the ex- treme ultraviolet lithography system can no longer detect the local surface defect.

The inventive method reduces the wave front distortions of an individual local surface defect at the imaging plane of a projection system of a li- thography system so that the local surface defect can no longer be recognized by the wave front sensor associated with the EUV lithography system. For this purpose, femto- or picosecond light pulses of a laser system correct local deviations of the surface height of the optical element from a predetermined surface height to such an extent that the defect volume of the corrected local surface defect is below the resolution limit of the wave front sensor of the EUV lithography system.

As the corrected local surface defect is no longer detectable, the user of the lithographic system is not worried about an alleged problem of the lithography system. Furthermore, the control and the adjustment of the system based on the data determined from the wave front sensor measurements is not impaired by the impact of one or several local surface defects. In addition, use cases can no longer occur in which local surface defect(s) impair the imaging performance of the projection system. Local surface defects of an optical element are typically detected at the end of the fabrication process. Consequently, the surface has its final form as well as its final roughness or smoothness. The inventive method avoids the removal of identified local surface defects by an abrasion process. An abrasion process is time-consuming. Furthermore, this process destroys or at least damages the surface roughness or smoothness of the corrected surface of the optical element. Therefore, the surface roughness has to be re-established by subsequent polishing of the corrected surface. If the local surface defect has the form of a bump, these processes can locally be performed at the defect position. However, if the local defect has the form of a recess, material has to be removed across the overall surface in order to remove the local surface defect. Thus, a significant part of the manufacturing process of the optical ele- ment has to be repeated.

Another aspect of the inventive method further comprises repeating steps (a) and (b) until analysis of the local surface defect satisfies a predetermined condition.

Analysis of the local surface defect may comprise measuring of the contour or of the volume of a local surface defect. Furthermore, satisfying a predetermined condition may for example mean that the volume of the remaining local surface defect is smaller than a predetermined threshold value.

After the local surface defect has been corrected by the application of the laser system onto positions of the local surface defect, the remaining defect can again be measured in order to detect whether its volume has been diminished below a predetermined threshold. When the remaining defect dimensions meet the threshold condition, this is a strong indication that the wave front sensor of the EUV lithography system can no longer detect the remaining local surface defect. On the other hand, if the remaining defect volume does not fulfil the threshold criterion, the cor- rection process is repeated based on the measured data of the remaining local surface defect until the defect volume is below the threshold condition. In another aspect, the step of analyzing the local surface defect comprises using of an interferometer and/or a scanning probe microscope and/or a scanning particle microscope.

The applied tool is adapted to the spatial frequency range to be meas- ured. For low spatial frequencies, an interferometer can be used. A micro-interferometer is often applied for higher spatial frequencies. For high spatial frequencies a scanning probe microscope can be used, as for example an atomic force microscope (AFM). Alternatively or additionally, a scanning particle microscope (e.g. a scanning electron microscope or a scanning ion beam apparatus) can scan the surface of the optical element in order to analyze the surface topology.

In a further aspect, the step of analyzing the local surface defect comprises analyzing the local surface defect across a spatial frequency range from a lateral dimension of the optical element to an illumination wavelength of the extreme ultraviolet lithography system.

It is one of the major challenges of EUV lithography that the surfaces of the optical elements forming a projection system have to fulfil very tight specifications across the spatial frequency range from the centimetre down to the nanometer range. Only by complying the strict surface specifications, projection systems can be manufactured having a diffraction limited performance. According to another aspect, the step of correcting the local surface defect does not modify a surface roughness of the reflective optical element. As already mentioned above, prior art removal of local surface defects destroys the surface roughness of the fabricated optical element which has to be re-established by a polishing process. As the inventive correction method does not change the surface roughness, a subsequent pol- ishing of the corrected surface is not necessary.

In another aspect, the step of correcting a height of the local defect exceeding a predetermined height comprises generating of a plurality of color centers by the focussed light pulses and/or correcting the height of the local defect falling below the predetermined height comprises locally depositing an energy exceeding a threshold of a local void and/or crack formation.

The inventive method allows correcting both types of local surface defects, defects having the form of a bump and defects having the form of a recess. Moreover, the absorption of color centers generated in case of a bump defect can be used to monitor their density which is proportional to the height variation or reduction at the position of the local surface defect.

In another aspect, an energy density of the light pulses is selected to locally generate either color centers or to locally form voids and/or cracks.

According to a further aspect, the step of correcting the local surface de- feet occurs prior to a deposition of a multilayer coating. In a further beneficial aspect, the step of correcting the local surface defect comprises introducing the light pulses into the optical element through the surface onto which a multilayer coating is to be deposited. As the local surface defect is corrected at the end of the fabrication process but prior to depositing the multilayer coating, the light pulses can be introduced through the surface onto which the multilayer coating is deposited. Furthermore, the coating can not be damaged by the application of the femto- or picosecond light pulses. In another aspect, the step of correcting the local surface defect comprises light pulses having a duration of 10 fs to loooo fs, preferably 20 fs to 1000 fs. In a further aspect, correcting the local surface defect com- prises light pulses having a focal spot diameter of 0.3 μηι to 20 μηι, preferably 0.5 μηι to 5 μηι. According to still another aspect, correcting the local surface defect comprises light pulses having a focal point dimension in beam direction of 0.5 μηι to 400 μηι, preferably 1 μηι to 200 μηι. In a further aspect, correcting the local surface defect comprises using light pulses having a repetition rate in the range of 1 Hz to 10 MHz, preferably 5 kHz to 5 MHz.

The laser beam parameters of the laser system producing the light pulses depend on the type and the form of the local surface defect. Further, the laser beam parameters must be adjusted to the material of the optical element. Moreover, these parameters vary with the wavelength of the applied laser system.

In still another aspect, the step correcting the local surface defect com- prises using light pulses having a density in the optical element in the range of 10³ mm ² to 10⁸ mm ², preferably 5-10³ mm ² to 5-10⁶ mm ². In a further beneficial aspect, correcting the local surface defect comprises focusing light pulses to a surface comprising the local surface defect of the optical element in a layer extending from 50 μηι above the surface to 1500 μηι below the surface comprising the local surface defect. In another aspect, correcting the local surface defect comprises focusing light pulses to a surface opposite to the local surface defect of the optical element in a layer extending within the optical element from 10 μηι to 1500 μηι below the surface containing the local surface defect. According to a further aspect, the step of correcting the local surface defect comprises focusing light pulses in multiple layers of different depth below the surface containing the local optical defect. The type of local surface defect and the material of the optical element determine the local writing of femto- or picosecond laser pulses in the optical element at the position of the surface defect. The writing of ultrashort laser pulses in the optical element is in the following called writ- ing of pixels. The effect of the pixels with respect to a local increase or a local decrease of the surface height due to a local compaction or expansion of the material of the optical element decrease with increasing distance of the pixels from the surface having the local surface defect. In a further aspect, the step of correcting the local surface defect comprises a lateral resolution of a position of the light pulses of less than 10 μηι.

As a consequence of the high lateral resolution of the position of the light pulses the writing of correcting pixels can follow the contour of the local surface defect with high precision. Thus, the contour of complex defects can also reproducibly be corrected.

In still another aspect, the optical element comprises a material having a low expansion coefficient material in particular fused silica.

In addition to the extremely low surface deficiencies of optical elements, the optical elements may also not change their surface form during the operation of the lithography system as for example due to a temperature change due to absorbed EUV photons.

In another aspect, the local surface defect contributes less than 10% to a predetermined figure error and/or to a predetermined mean spatial frequency error and/or a predetermined high spatial frequency error. In yet another aspect, a lateral dimension of the local surface defect is in the range of 0.01 mm to 2.0 mm and a height deviation is in the range of 1 nm to 20 nm. In a further beneficial aspect, the step of correcting the local surface defect comprises reducing its error volume below a resolution limit for wave front errors of the wave front sensor. The resolving power of a wave front sensor of a lithography system is adjusted so that it can resolve the imaging relevant errors of the lithography system. The imaging relevant errors are figure and MSF errors. Their spatial frequency ranges depend on the illumination wavelength. Therefore, the resolving power of the wave front sensor is designed to detect these kinds of errors.

As already mentioned, the EUVbeam in the projection system of the lithography system carries the pattern information of the photolithographic mask from the mask to the photoresist on the wafer. Therefore, specific demands are made on the optical elements forming the projection system.

Finally, according to another aspect, an apparatus for correcting a local surface defect of a reflective optical element of an extreme ultraviolet lithography system comprises (a) at least one metrology tool operable to analyze the local surface defect, (b) at least one laser system operable to generate focussed femto- or picosecond light pulses correcting the local surface defect, and (c) at least one wave front sensor associated with the extreme ultraviolet lithography system and operable to verify that the local surface defect can no longer be detected.

Further aspects of the invention are defined in further dependent claims. 4. Description of the drawings

In order to better understand the present invention and to appreciate its practical applications, the following Figures are provided and referenced hereafter. It should be noted that the Figures are given as examples only and in no way limit the scope of the invention. Fig. l schematically represents a block diagram of some of the major components of an EUV lithography system; Fig. 2 schematically shows a section of an optical element of the projection system of the EUV lithography system of Fig. l;

Fig. 3 schematically illustrates the section of the optical element of

Fig. 2 without the multilayer coating;

Fig. 4 schematically represents a section of the optical element of Fig.

3 having two bumps as local surface defects;

Fig. 5 schematically shows a section of the optical element of Fig. 3 having a bump and a recess as local surface defects;

Fig. 6 schematically shows a block diagram of an apparatus for the correction of local optical defects of optical elements; Fig. 7 schematically illustrates the correction of the local surface defects of Fig. 4; and

Fig. 8 schematically represents the correction of the local surface defects of Fig. 5.

5. Detailed description of preferred embodiments

The US Provisional US 61 324 467 describes in detail how the writing of pixels with a femto- or picosecond laser beam can change a surface height of an optical element. It is hereby incorporated herein in their entirety by reference. In particular, this document explains that the writing of an arrangement of pixels in an optical element can increase as well as decrease its surface height. In the following, the present invention will be more fully described hereinafter in more details with reference to accompanying Figures, in which exemplary embodiments of the invention are illustrated. However, the present invention may be embodied in different forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and will convey the scope of the invention to persons skilled in the art.

Fig. l shows a functional sketch of a possible embodiment of a future EUV lithography system loo. The source no generates EUV photons for example at a wavelength of 13.5 nm. EUV radiation can for example be obtained from synchrotrons, laser induced plasmas or gas discharge plasma sources. The source 110 has typically a filter which restricts the wavelength of the emitted EUV beam to a bandwidth of ±2 % of the center wavelength. A collector 120 collects the photons of the source 110 and focuses them in an intermediate focus 130 at the entrance of the illuminator 140. In the example of Fig. 1, the illuminator 140 comprises four imaging mirrors which direct the EUV photons onto the reflective photomask at the reticle or mask stage 150. From the mask arranged at the stage 150, the EUV beam is reflected in the projection system 160 of the EUV lithography system 100. In the presented example the projection system 160 comprises six optical elements 170, 171, 172, 173, 174, 175 which guide the EUV beam with a 1:4 reduction onto the wafer arranged on the wafer stage 180. All optical elements 170-175 of the projection system 160 are reflective imaging mirrors.

The wafer stage 180 can be removed and the wave front sensor 190 can be placed at the position of the imaging plane of the projection system 160. The wave front sensor 190 can include any detector which is capable to detect EUV photons with a predetermined spatial resolution. For example, a silicon photodiode having a respective wavelength filter can be scanned across the imaging plane of the projection system 160. Alternatively, a CCD (charge-coupled device) sensor can also be used for the detection of the distribution EUV beam across the imaging plane. The resolving power of the wave front sensor 190 is designed so that it can resolve the imaging relevant deficiencies of the optical elements 170- 175 of the projection system 160. The imaging relevant surface errors of the optical elements 170-175 comprise the spatial frequency range of the figure errors and of the MSF errors. For the EUV wavelength of 13.5 nm the MSF range extends to dimensions of approximately 1 μιη. Consequently, the wave front sensor 190 needs a resolution to measure the imaging relevant aspects of the optical elements which are caused by surface defects extending to lateral dimensions of approximately 1 μιη.

The wave front sensor 190 detects the sum of the imaging errors of the optical elements 170-175 of the projection system 160 according to the error propagation. Further, the wave front sensor 190 recognizes the in- dividual local surface defects of each of the optical elements 170-175.

Fig. 2 shows a section of one of the optical elements 170-175 of the projection system 160 of the EUV lithography system 100. The optical element 200 comprises a material having a low expansion coefficient, such as fused silica. Other dielectrics, glass materials or semiconducting materials may also be applied as material for the fabrication of optical elements 170-175. Presently, materials are preferred having very low expansion coefficients as for example ZERODUR®, ULE® or CLEARCE- RAM®.

The optical element 200 of Fig. 2 has on its front surface 210 a multilayer coating in order to optimize the reflectivity of the optical element 170-175, 200 for the EUV beam. The multi-layer coating comprises for example of alternating molybdenum (Mo) 230 and silicon (Si) layers 240 (referred to in the following as MoSi layers). In the multi-layer coating system, the Mo layers 230 act as scattering layers, whereas the silicon layers 240 function as separation layers. Fig. 3 represents the optical element 200 of Fig. 2 prior to the deposition of the MoSi multi-layer coating. In this example, the section of the front surface 310 and of the rear surface 320 of the optical element 300 is plain. In contrast to this, Fig. 4 illustrates a section of an optical element 400 having two bumps 430 and 440 on the front surface 410. For comparison purpose, the dashed line 450 represents a predetermined plain surface.

The sections of the optical elements 200, 300 and 400 are represented as plain sections, although the optical elements 170-175 of the projection 160 are curved imaging optical elements. In a first approximation, a small section of the optical elements 170-175 can be regarded as plain. Moreover, the sections of the optical elements 200, 300, 400, 500, 700, 800 are intended to illustrate the principle of the inventive method, but not to show to scale drawings.

Fig. 5 schematically depicts a section of an optical element 500 having on its front surface 510 a recess 530 on the left side and a bump 540 at the right side. Similar to Fig. 4, the dashed line 550 indicates a prede- termined plain front surface 510 of the optical element 500.

The local surface defects 430, 440, 530, 540 illustrated in Figs. 4 and 5 are analyzed by measuring the surface of the fabricated optical elements 400 and 500 prior to the deposition of the multi-layer coating by using an appropriate metrology tool. For the low and medium spatial frequency ranges an interferometer or a micro-interferometer, respectively, may be applied. The high spatial frequency range of the surface of the optical element 400 and 500 may be detected with an atomic force microscope (AFM). In addition or alternatively, an electron scanning mi- croscope or an ion scanning device may be applied to detect the surface contour.

Fig. 6 depicts a schematic block diagram of an apparatus 600 which can be used to correct the local surface defect 430, 440, 530 and 540 of the optical elements 400 and 500 of Figs. 4 and 5, respectively. The apparatus 600 comprises a chuck 620 which may be movable in three dimensions. The movement of the chuck 620 in two dimensions in the plane of the chuck 620 is indicated in Fig. 6 by crossed arrows. The optical element 400 or 500 may be fixed to the chuck 620 by using various techniques as for example by mechanical clamping. The optical element 400 or 500 is mounted on the chuck 620 so that its front surface 410 or 510 is directed towards the objective 640. Alternatively, the optical element 400 or 500 may be mounted on the chuck 620 so that its rear surface 420 or 520 is directed towards the objective 640. In this arrangement the pixels are written through the rear surface 420 or 420 of the optical element 400 or 500.

The apparatus 600 includes a pulse laser source 630 which produces a beam or a light beam 635 of pulses or light pulses. The laser source 630 generates light pulses of variable duration. The adjustable range of several import parameters of the laser source 630 is summarized in the following table. Table 1 represents an overview of laser beam parameters of a Ti:Sapphire laser system which can be used in an embodiment of the inventive method.

Table 1: Numerical values of selected laser beam parameters for a

Ti:Sapphire laser system

Overview

Parameter Numerical value Unit

Pulse energy 0.05 - 5 μ

Pulse width 0.05 - 100 ps

Repetition rate 0.001 - 10 000 kHz

Pulse density 1 000 - 100 000 000 mm ²

NA (numerical aperture) 0.1 - 0.9

Wavelength 800 nm In an alternative embodiment, the light pulses may be generated by an Ytterbium-doped fiber laser operating at a wavelength of 1030 nm.

However, the various aspects of the inventive method described in the following are not limited to these laser types, principally all laser types may be used having a photon energy which is smaller than the band gap of the optical element 400 or 500 and which are able to generate pulses with durations in the femto- and picosecond range.

The steering mirror 690 directs the pulsed laser beam 635 into the focus- ing objective 640. The objective 640 focuses the pulsed laser beam 635 onto the optical element 400 or 500. The NA (numerical aperture) of the applied objectives 640 depends on the predetermined spot size of the focal point and the position of the focal point within the optical element 400 or 500. As indicated in table 1, the NA of the objective 640 may be up to 0.9 which results in a focal point spot diameter of essentially 1 μηι at a wavelength of 800 nm.

The apparatus 600 also includes a controller 680 and a computer 660 which manage the translations of the two-axis positioning stage of the chuck 620 in the plane perpendicular to the laser beam 635 (x and y directions). The controller 680 and the computer 660 also control the translation of the objective 640 perpendicular to the plane of the chuck 620 (z direction) via the one-axis positioning stage 650 to which the objective 640 is fixed. It should be noted that in other embodiments of the apparatus 600 the chuck 620 may be equipped with a three-axis positioning system in order to move the optical element 400 or 500 to the target location and the objective 640 may be fixed, or the chuck 620 may be fixed and the objective 640 may be moveable in three dimensions. Although not economical, it is also conceivable to equip both the objec- tive 640 and the chuck 620 with three-axis positioning systems. It should be noted that manual positioning stages can also be used for the movement of the optical element 400 or 500 to the target location of the pulsed laser beam 635 in x, y and z directions and/or the objective 640 may have manual positioning stages for a movement in three dimensions.

The computer 660 may be a microprocessor, a general purpose proces- sor, a special purpose processor, a CPU (central processing unit), a GPU (graphic processing unit), or the like. It may be arranged in the controller 680, or may be a separate unit such as a PC (personal computer), a workstation, a mainframe, etc. The computer 660 may further comprise I/O (input/output) units like a keyboard, a touchpad, a mouse, a video/ graphic display, a printer, etc. In addition, the computer 660 may also comprise a volatile and/or a non-volatile memory. The computer 660 may be realized in hardware, software, firmware, or any combination thereof. Moreover, the computer 660 may control the tool used to analyze the surface of the optical elements 400 and 500 (not indicated in Fig. 6).

Further, the apparatus 600 may also provide a viewing system including a CCD (charge-coupled device) camera 665 which receives light from an illumination source arranged to the chuck 620 via the dichroic mirror 645. The viewing system facilitates navigation of the optical element 400 or 500 to the target position. Further, the viewing system may also be used to observe the formation of a modified or corrected area on the optical element 400 or 500 by the pulsed laser beam 635 of the light source 630.

The computer system 660 can display an image of the local surface defect 430, 440, 530, 540 of the optical element 400 or 500, respectively, calculated from the data of a surface scanning tool. Further, the computer system 660 may contain algorithms, realized in hardware and/or software, which allow extracting control signals from the experimental surface contour data. The control signals control the writing of an arrangement of pixels in the optical element 400 or 500 by the laser source 630 in order to correct the local surface defects 430, 440, 530, 540 of the optical elements 400 and 500 (cf. Figs. 7 and 8 below). Fig. 7 schematically illustrates the correction of the local surface errors 430 and 440 of Fig. 4. The optical element 700 corresponds to the optical element 400 prior to the application of the inventive method. In the optical element 700 color centers are generated by scanning focused femto- or picosecond light pulses of the pulsed laser source 630 in planes of various distance from the front surface 710 across the regions having the bumps 430 and 440, so that the distribution of the generated color centers correlates to the local deviation of the front surface 710 from the predetermined front surface 450. The laser pulses can enter the optical element 400 either through the front surface 7100 or through the rear surface 720.

The color center distribution 730 of Fig. 7 is formed using pulses which have a duration of 200 fs (femtosecond) and a repetition rate of 100 kHz. The pulse energy is 1.5 μJ at a focal point diameter of 2.5 μηι and a pitch (distance between two pixels in one layer) of 0.2 μηι. The minimum distance of the focal point of the femto- or picosecond light pulses from the front surface 710 of the optical element 700 is 10 μηι and the distance between the different planes of generated color centers is 100 μηι. As indicated in Fig. 7, the generated color center distribution 730 essentially removes the bumps 430 and 440 of the associated local surface defects of the optical element 400 by a local compaction of the material of the optical element 400 which results in a local reduction of the surface height.

Color centers are effectively generated using femtosecond laser pulses. For example using laser pulses with a wavelength of 800 nm

(Ti:Sapphire laser), a pulse width of 200 fs, and using an objective with an NA of 0.3, color centers are effectively generated with a pulse energy of 0.3 μJ to 1.5 μ<1. The concentration of color centers increases with increasing pulse energy. The pulse repetition is 100 kHz and the pulse density represents 4·ιο⁶ cm ². Fig. 8 schematically shows how the recess 530 of the front surface 510 of the optical element 500 can essentially be removed. The application of localized femtosecond light pulses exceeding a certain energy density threshold leads to a local optical breakdown of the fused silica of the op- tical element 500 of Fig. 5. This breakdown results to a local expansion of the damaged zone and the formation of expansion stress in the material of the optical element 500 resulting in a local formation of voids and/or cracks. Thus, the generation of local voids and/or crack zones 830 which correlates to the height deviation of the recess 530 from the predetermined surface 550 locally increases the front surface 810 of the optical element 800 of Fig. 8. In the breakdown mode, the pulse duration of the femtosecond light pulses is 1000 fs and repetition rate is 100 kHz. The focal point diameter is 2.5 μηι at a pitch of 0.7 μηι. The distance of the plane of femtosecond light pulses from the front surface 810 is 100 μηι and the distance between the layers of the void and/ or crack formation zones generated by the femtosecond light pulses is 200 μηι.

The bump 540 of Fig. 5 is corrected analog to the procedure described in the context of the discussion of Fig. 7. As indicated in Fig. 8, the front surface 810 of the optical element 800 corresponds essentially to the predetermined front surface 550 of the optical element 500 in Fig. 5 after the writing of the respective pixel arrangements.

In order to obtain a local optical breakdown leading a local formation of voids and/ or cracks and using again an objective with a NA of 0.3, longer pulses in the range of 1 ps are preferably applied. For example, using again the Ti:Sapphire laser system (800 nm), the application of pulse widths in the range of 1.0 ps to 1.5 ps having a pulse energy in the interval between 0.8 μJ and 1.5 μJ effectively generates local voids and/or cracks.

As indicated above, by varying of the laser pulse energy, pulse width, wavelength, NA of the focusing objective, repetition rate of the laser pulses and the density of the pulses (overlapping degree), it can be de- cided whether color centers 730 are generated or a local optical breakdown forms voids and/or cracks 830 in the optical element 800.

It is a great advantage of the inventive method that both types of local surface defects - bumps and recesses - can be corrected with the laser system 600 in a single correction process by just adjusting the operation parameters of the laser system 600, in particular the energy density and/or the pulse width. As a consequence, the combination of both, the formation of local voids and/or cracks 830 and the local generation of color centers 730 can correct local surface defects 430, 440, 530, 540 of optical elements 170-175. The writing of the pixel arrangements at the positions of the local surface defects 430, 440, 530, 540 do not modify the surface roughness of the optical elements 400 and 500. Thus, the optical elements 170-175 have essentially a surface contour following the predetermined contour.

Therefore, a multilayer coating can be deposited on the corrected optical elements 700 and 800 without any further process steps. The wave front sensor 190 of an EUV lithography system 100 can no longer detect a signal from a projection system 160 comprising optical elements 700 and 800 having corrected local surface defects 430, 440, 530 540 as optical elements 170-175. Hence, the corrected local surface defects 430, 440, 530, 540 of the optical elements 170-175 can not disturb or impair the data obtained from the wave front sensor 190 for the adjustment of the EUV lithography system 100, and it can also no longer impair the imaging performance in certain critical cases.

Claims

A method for correcting a local surface defect (430, 440, 530, 540) of a reflective optical element (170, 171, 172, 173, 174, 175, 200, 300, 400, 500, 700, 800) of an extreme ultraviolet lithography system (100), comprising the steps of: a. analyzing the local surface defect (430, 440, 530, 540); b. correcting the local surface defect (430, 440, 530, 540) by focusing femto- or picosecond light pulses of a laser system (600) onto positions of the local surface defect (430, 440, 530, 540); and c. verifying that a wave front sensor (190) associated with the extreme ultraviolet lithography system (100) can no longer detect the local surface defect (430, 440, 530, 540).

2. The method of claim 1, further comprising repeating steps a. and b. until analysis of the local surface defect (430, 440, 530, 540) satisfies a predetermined condition. 3· The method of claim 1 or 2, wherein the step of analyzing the local surface defect (430, 440, 530, 540) comprises using of an interferometer and/or a scanning probe microscope and/or a scanning particle microscope. 4- The method according to any of the preceding claims, wherein the step of analyzing the local surface defect (430, 440, 530, 540) comprises analyzing the local surface defect (430, 440, 530, 540) across a spatial frequency range from a lateral dimension of the op- tical element to an illumination wavelength of the extreme ultraviolet lithography system (100).

The method according to any of the preceding claims, wherein the step of correcting the local surface defect (430, 440, 530, 540) does not modify a surface roughness of the reflective optical element (170, 171, 172, 173, 174, 175, 200, 300, 400, 500, 700, 800).

The method according to any of the preceding claims, wherein the step of correcting a height of the local defect (430, 440, 530, 540) exceeding a predetermined height (450, 550) comprises generating of a plurality of color centers (730) by the focussed light pulses and/or correcting the height of the local surface defect (430, 440, 530, 540) falling below the predetermined height (450, 550) comprises locally depositing an energy exceeding a threshold of an optical breakdown (830).

The method according to claim 6, wherein in step b. of claim 1 an energy density of the light pulses is selected to locally generate either color centers (730) or to locally form voids and/or cracks (830).

The method according to any of the preceding claims, wherein the step of correcting the local surface defect (430, 440, 530, 540) occurs prior to a deposition of a multilayer coating of the reflective optical element (170, 171, 172, 173, 174, 175, 200, 300, 400, 500, 700, 800).

The method according to any of the preceding claims, wherein the step of correcting the local surface defect (430, 440, 530, 540) comprises introducing the light pulses into the optical element (170, 171, 172, 173, 174, 175, 200, 300, 400, 500, 700, 800) through the surface (210, 310, 410, 510, 710, 810) onto which a multilayer coating is to be deposited. The method according to any of the preceding claims, wherein the step of correcting the local surface defect (430, 440, 530, 540) comprises light pulses having a duration of 10 fs to 10000 fs, preferably 20 fs to 1000 fs.

The method according to any of the preceding claims, wherein the step of correcting the local surface defect (430, 440, 530, 540) comprises light pulses having a focal spot diameter of 0.3 μηι to 20 μηι, preferably 0.5 μηι to 5 μηι.

The method according to any of the preceding claims, wherein the step of correcting the local surface defect (430, 440, 530, 540) comprises light pulses having a focal point dimension in beam direction of 0.5 μηι to 400 μηι, preferably 1 μηι to 200 μηι.

The method according to any of the preceding claims, wherein the step of correcting the local surface defect (430, 440, 530, 540) comprises using light pulses having a repetition rate in the range of 1 Hz to 10 MHz, preferably 5 kHz to 5 MHz.

The method according to any of the preceding claims, wherein the step of correcting the local surface defect (430, 440, 530, 540) comprises using light pulses having a density in the optical element (170, 171, 172, 173, 174, 175, 200, 300, 400, 500, 700, 800) in the range of 10³ to 10⁸ mm ², preferably 5-io³ mm ² to 5-10⁶ mm ².

The method according to any of the preceding claims, wherein the step of correcting the local surface defect (430, 440, 530, 540) comprises focusing light pulses to a surface (210, 310, 410, 510, 710, 810) comprising the local surface defect (430, 440, 530, 540) of the optical element (170, 171, 172, 173, 174, 175, 200, 300, 400, 500, 700, 800) in a layer extending from 50 μηι above the surface (210, 310, 410, 510, 710, 810) to 1500 μηι below the surface (210, 310, 410, 510, 710, 810).

The method according to any of the preceding claims, wherein the step of correcting the local surface defect (430, 440, 530, 540) comprises focusing light pulses to a surface (220, 320, 420, 520, 720, 820) opposite to the local surface defect (430, 440, 530, 540) of the optical element (170, 171, 172, 173, 174, 175, 200, 300, 400, 500, 700, 800) in a layer extending within the optical element (170, 171, 172, 173, 174, 175, 200, 300, 400, 500, 700, 800) from 10 μηι below to 1500 μηι below the surface (210, 310, 410, 510, 710, 810) comprising the local surface defect (430, 440, 530, 540).

The method according to any of the preceding claims, wherein the step of correcting the local surface defect (430, 440, 530, 540) comprises focusing light pulses in multiple layers of different depth below the surface (210, 310, 410, 510, 710, 810) containing the local surface defect (430, 440, 530, 540).

The method according to any of the preceding claims, wherein the step of correcting the local surface defect (430, 440, 530, 540) comprises a lateral resolution of a position of the light pulses of less than 10 μηι.

The method according to any of the preceding claims, wherein the optical element comprises a material having a low expansion coefficient, in particular fused silica.

20. The method according to any of the preceding claims, wherein the local surface defect (430, 440, 530, 540) contributes less than 10% to a predetermined figure error and/or to a predetermined mean spatial frequency error and/or to a predetermined high spatial frequency error.

21. The method according to any of the preceding claims, wherein a lateral dimension of the local surface defect (430, 440, 530, 540) is in the range of 0.01 mm to 2.0 mm and a height deviation is in the range of 1 nm to 20 nm.

22. The method according to any of the preceding claims, wherein the step of correcting the local surface defect (430, 440, 530, 540) comprises reducing its error volume below a resolution limit for wave front errors of the wave front sensor (190).

23. An system for correcting a local surface defect (430, 440, 530, 540) of a reflective optical element (170, 171, 172, 173, 174, 175, 200, 300, 400, 500, 700, 800) of an extreme ultraviolet lithography system (100), comprising: a. at least one metrology tool operable to analyze the local surface defect (430, 440, 530, 540); b. at least one laser system (600) operable to generate focussed femto- and/or picosecond light pulses correcting the local surface defect (430, 440, 530, 540); and c. at least one wave front sensor (190) associated with the extreme ultraviolet lithography system (100) and operable to verify that the local surface defect (430, 440, 530, 540) can no longer be detected.

24. The system of claim 23, further comprises repeated application of the at least one metrology tool and of the at least one laser system (600) until analysis of the local surface defect (430, 440, 530, 540) satisfies a predetermined condition.