WO2021165078A1

WO2021165078A1 - Method for operating an optical assembly for euv lithography, and optical assembly for euv lithography

Info

Publication number: WO2021165078A1
Application number: PCT/EP2021/052932
Authority: WO
Inventors: Udo Dinger; Stephanus Fengler; Daniel Zemann; Dirk Heinrich Ehm; Moritz Becker; Tina GRABER
Original assignee: Carl Zeiss Smt Gmbh
Priority date: 2020-02-20
Filing date: 2021-02-08
Publication date: 2021-08-26
Also published as: DE102020202179A1

Abstract

The invention relates to a method for operating an optical assembly (1) for EUV lithography, having the steps of: generating EUV radiation (5) at an operating wavelength (λ_B) in the EUV wavelength range using an EUV radiation source (2) of the optical assembly (1); generating a plasma (43) in a vacuum housing (27) of the optical assembly (1), said housing being equipped with at least one optical element (12), by means of an interaction between the EUV radiation (5) and a residual gas (40) contained in the vacuum housing (27); irradiating excitation radiation (31) into the vacuum housing (27) at at least one excitation wavelength (λ_A), which differs from the operating wavelength (AB), by means of an excitation device (30); and exciting at least one photo-induced reaction in the residual gas (40) in the vacuum housing (27) in a controlled manner by actuating the excitation device (30). The invention also relates to an optical assembly (1) for EUV lithography.

Description

METHOD OF OPERATING AN OPTICAL ARRANGEMENT FOR THE EUV LITHOGRAPHY AND OPTICAL ARRANGEMENT FOR THE EUV

LITHOGRAPH

Reference to related applications

This application claims the priority of the German patent application DE 102020202179.1 of January 20, 2020, the entire disclosure content of which is incorporated into the content of this application by reference.

Background of the invention

The invention relates to a method for operating an optical arrangement for EUV lithography, comprising: generating EUV radiation at an operating wavelength in the EUV wavelength range by means of an EUV radiation source of the optical arrangement, and generating a plasma in a vacuum housing of the optical Arrangement in which at least one optical element is arranged due to an interaction of the EUV radiation with a residual gas contained in the vacuum housing during operation of the optical arrangement. The invention also relates to an optical arrangement for EUV lithography, comprising: an EUV radiation source for generating EUV radiation at an operating wavelength in the EUV wavelength range, and at least one vacuum housing in which at least one (reflective) optical element is arranged is, wherein the vacuum housing contains a residual gas, and wherein the optical arrangement is designed to generate a plasma in the vacuum housing during operation by an interaction of the EUV radiation with the residual gas. The invention also relates to a method for determining a setpoint value for a target plasma parameter, in particular for a target plasma parameter which determines the service life of at least one optical element.

The optical arrangement for EUV lithography can be an EUV lithography system (EUV scanner) for exposing a wafer or another optical arrangement that uses EUV radiation, for example an EUV metrology system, for example for inspection of masks, wafers or the like used in EUV lithography to create an EUV test setup, etc. Such an optical arrangement is made with EUV radiation at an operating wavelength in the EUV wavelength range between approx nm operated. The operating wavelength can be, for example, 13.5 nm or 6.8 nm. Due to the low transmission of all gases at wavelengths in the range of 13.5 nm - or at 6.8 nm - it is necessary to operate the optical elements of such an optical arrangement in a vacuum environment, which is in (at least) one vacuum housing is formed. Due to the low transmission of practically all solid-state materials at these wavelengths, reflective optical elements (e.g. mirrors) are usually used in such an optical arrangement.

To avoid contamination or for cleaning purposes, such an optical arrangement is not operated in a high vacuum, but at an operating pressure of approx. 5 Pa in the vacuum housing, typically with molecular hydrogen as residual gas or as flushing gas. In the residual gas, which is contained in the vacuum housing, gases other than hydrogen, e.g. Ar, He, as well as slight oxygen and / or nitrogen admixtures, as well as H2O and carbon, in particular in the form of CO2 and in the form of hydrocarbons (C _n H _m ) be included. With 92 eV photon energy, EUV radiation is ionizing for all of the gases contained in the residual gas - the ionization energies in argon and hydrogen are, for example, in the range of 15 eV. The photo-absorption cross-section is almost completely determined by the various ionization channels (single, double, dissociative photoionization). The photoelectrons have sufficient energy to induce further secondary ionizations. If the EUV radiation hits a surface, for example the surface of a reflective optical element, it also releases (on average comparatively low-energy, ie in the range of a few eV) electrons from the surface and generates a photocurrent. At longer wavelengths - in particular in the range from 13.5 nm to 300 nm - further photo-induced reactions such as photodissociation or vibratory excitation with subsequent electron attachment can occur.

In hydrogen, different ionic hydrogen species (H2 ⁺ , H3 ⁺ , H ⁺ , H, ...) as well as neutral hydrogen species, e.g. in the form of hydrogen radicals H or of neutral hydrogen species H *, are combined in one excited electron state formed. Particularly in the case of oxygen admixtures in the residual gas, secondary reactions also form negatively charged oxygen ions (O, O2, C, O4) in _{addition to the positive oxygen ions (0 2} ⁺ , 0 ⁺ , 0 ^{2+ ..)} . In nitrogen-containing residual gas or plasmas, for example, IS FT, NFU ⁺ and all species that lead to ammonia NFI3 play a significant role. In residual gases in the form of hydrogen / nitrogen / oxygen mixtures, more than 200 species, including, for example, FhO ⁺ or FINO3, can arise.

Due to the plasma dynamics, the plasma formed by the interaction with the EUV radiation is at a more or less constant positive potential with respect to the earthed chamber walls of the vacuum housing and of optical or non-optical ones arranged in the vacuum housing Components from which it is separated by a steep drop in potential, known as plasma sheath or plasma equalization layer. In this area, under the relevant operating conditions, field strengths in the range of a few kV / m prevail.

This electric field, plasma density gradients and thermal movement create a constant flow of positive and negative charge carriers to the boundaries of the plasma. Ions are accelerated in the sheath, while electrons and negative ions are decelerated. Neutral species migrate to the limits by diffusion. Recombination, reflection or more complicated surface reactions then occur on the walls.

The plasma potential ensures that in the equilibrium state the electronic and ionic currents compensate each other towards the limits, otherwise a constantly growing space charge would arise. During the pulsed excitation, ie during a respective EUV radiation pulse which, in the case of an EUV radiation source in the form of an LPP ("laser produced plasma") radiation source, has a pulse duration of generally between approx. 10 ns and approx. 100 ns, which with an EUV radiation source in the form of a GDPP ("gas discharge produced plasma") radiation source has a pulse duration of usually between approx. 50 ns and approx. 500 ns and that with an EUV light source that has a free electron laser , has a pulse duration of usually less than one ps, this compensation can, however, be violated at least for a short time in an optical arrangement for EUV lithography, so that a transient plasma state is present.

The ionic plasma species and the neutral plasma species (for example in the form of radicals) can on the one hand have advantageous effects, for example in the case of hydrogen a cleaning effect, but the ion or radical flows can also cause disadvantageous effects. For example Ion fluxes incident on the mirror can lead to adverse effects such as sputtering of the surface, contamination of the surface by ion deposition, ion implantation or triggering of surface chemical reactions and, in the worst case, to the destruction of layers, all of which affect the quality and service life of the mirror or others reduce the components arranged in the vacuum housing. Furthermore, electrical currents can endanger the functionality of electrically sensitive components.

To optimize the plasma in the vacuum housing, it is advantageous to specifically influence the ion flows on the components arranged in the vacuum housing or, if necessary, other plasma parameters.

WO 2017/202545 A1 describes a projection exposure system which has at least one partial volume which is closed off from the environment and which contains a flushing gas, for example hydrogen, from which a plasma can be generated. The projection exposure system has conditioning elements for conditioning the plasma in the partial volume, which are suitable for at least partial neutralization of the plasma. The conditioning elements can be suitable for locally introducing a gas to reduce the plasma concentration, for example nitrogen, which leads to an increased formation of NH compounds, whereby the concentration of the hydrogen ions can be reduced. The conditioning elements can also be suitable for locally influencing partial pressures of the flushing gas or for influencing the electrical potential of elements arranged in the partial volume. By influencing the electrical potential, positively charged ions, for example, can be kept away from reaching the surfaces. However, a positive bias accelerates the electrons present in the plasma and can enable them to trigger additional ionizations. DE 102009029 121 A1 describes a generator for atomic hydrogen which has a vessel with a gas inlet for introducing molecular hydrogen and a gas outlet. A UV radiation source is arranged outside the vessel and the vessel is designed in such a way that the UV radiation can penetrate into the interior of the vessel. Homolytic decomposition of the molecular hydrogen into atomic hydrogen takes place in the vessel, which exits the vessel via the gas outlet and can be used, for example, to clean the surfaces of an EUV lithography device.

In WO 2009/053476 A1 a device and a method for cleaning one or more optical or non-optical elements are described, which are in particular housed hermetically in a housing of an EUV lithography system. The at least one optical or non-optical element is / are irradiated with radiation from a broadband light source. Due to the broadband spectrum, a large number of possible contaminants are excited. The excitation makes it possible to identify the contaminants so that a targeted cleaning of the contaminants can take place. For this purpose, instead of a broadband light source, a light source can be used that emits light at a specific wavelength that is adapted to the targeted removal of the identified contaminants.

From DE102017213406A1 a reflective optical element is known which has an edge. At least along one edge section, the edge has a bevel and / or a rounding.

The bevel and / or rounding is intended to avoid a localized increase in the electrical field, which may otherwise develop when the (sharp) edge is exposed to a plasma. DE102017213406A1 also describes a method for adapting a geometry of at least one surface area of a component of an optical arrangement. Object of the invention

The object of the invention is to provide a method for operating an optical arrangement for EUV lithography and an optical arrangement for EUV lithography, which specifically influence the plasma in the vacuum housing, in particular with regard to the service life of optical elements arranged there , enable.

Subject of the invention

This object is achieved according to a first aspect by a method of the type mentioned at the beginning for operating an optical arrangement, further comprising: irradiating excitation radiation into the vacuum housing at at least one excitation wavelength that differs from the operating wavelength by means of a Excitation device of the optical arrangement, as well as targeted excitation of at least one photo-induced reaction in the residual gas in the vacuum housing by controlling the excitation device, in particular with the aid of a control device of the optical arrangement.

According to a second aspect, this object is achieved by an optical arrangement of the type mentioned at the outset, further comprising: an excitation device which is designed to radiate excitation radiation into the vacuum housing at at least one excitation wavelength that differs from the operating Wavelength differs, as well as a control device which is designed or programmed to control the excitation device, in particular to specifically excite at least one photo-induced reaction in the residual gas in the vacuum housing. By stimulating the photo-induced reaction, the plasma in the vacuum housing can be influenced. The inventors have recognized that the negative effects of the plasma described above cannot generally be completely prevented, especially since some of these are even wanted. Instead of the partial neutralization of the plasma, as described in WO 2017/202545 A1, the present application proposes to specifically influence the plasma in the vacuum housing, ie to specifically control or regulate the plasma chemistry. By controlling the plasma chemistry, an optimal balance can be found between desired effects of the plasma, for example carbon cleaning, and undesired effects of the plasma, for example sputtering, oxidation or redeposition of species contained in the plasma on the delimiting surfaces.

According to the above aspects of the invention, it is proposed not to use electromagnetic fields, for example in the form of electrical fields or potentials or homogeneous or inhomogeneous magnetic fields, for targeted influencing of the plasma in the vacuum housing, but rather the spectral dependence of the cross-section of the photo-induced reactions in the plasma to be used in a targeted manner in order to influence the plasma, in particular in order to adjust the ratio between different species present in the plasma.

This requires knowledge of the dependence of the photochemical cross-sections of the respective reaction on the wavelength or on the photon energy. For the cross sections of a plurality of photoinduced reactions relevant to the present application, see the article “Photdissociation and photoionization of atoms and molecules of astrophysical interest”, A. N. Fleays et al. , A&A 602, A105 (2017).

The photo-induced reaction can be excited on primary components of the residual gas, ie on components that are present even without the plasma contained in the vacuum housing. The photo-induced reaction can, however, also be excited on species or on constituents of the residual gas which are only formed in the plasma by secondary reactions. For example, a photo-induced reaction on NH3 or on H ₃ 0 ^{+ can be} excited.

The embodiments or further developments according to the second aspect of the invention described below in connection with the optical arrangement also apply analogously to the method according to the first aspect of the invention.

In one embodiment, the excitation device is designed to radiate the excitation radiation into the vacuum housing in a wavelength range which is between 30 nm and 300 nm, preferably between 100 nm and 200 nm. The photochemical cross-sections of most of the chemical reactions relevant for the present application increase with lower photon energy and thus with increasing wavelength. It is therefore advantageous if the excitation device is designed to radiate excitation radiation into the vacuum housing at longer wavelengths than in the EUV wavelength range, ie above 30 nm. In particular in the VUV or DUV wavelength range up to a maximum of approx. 300 nm, the excitation radiation can trigger chemical reactions that are relevant for influencing the plasma. Certain gases that can form constituents of the residual gas, eg hydrogen and nitrogen, are largely transparent for wavelengths greater than approx. 100 nm, ie no photo-induced chemical reactions of these gases are to be expected at more than approx. 100 nm. Other gases that can form constituents of the residual gas, for example oxygen, water or ammonia, also have a very rich photochemistry above 100 nm to approx. 200 nm and possibly above. By stimulating photochemical reactions of these gases at wavelengths of more than 100 nm, the composition of the plasma can be changed without this directly affects the photo-induced reactions of gases such as hydrogen that are transparent at these wavelengths.

For example, by irradiating the excitation radiation, a photo-induced reaction of oxygen can be specifically excited. Oxygen does not show any ionization at wavelengths greater than 103 nm, but has some absorption lines at wavelengths greater than 100 nm and at which a photo-induced dissociation of the oxygen molecule O2 takes place through a photon Y, i.e. g + O2 - 2 0.

The formation of oxygen radicals (0) by photo-induced dissociation is beneficial in order to counteract the reducing effect of hydrogen contained in the plasma, without the negative effects occurring that are caused by positively charged oxygen ions due to the plasma potential accelerated towards the bounding surfaces. The so-called Schumann-Runge bands at approx. 193 nm and the broad Schumann-Runge continuum in the wavelength range between 130 nm and 170 nm are of particular interest for the photo-induced dissociation of oxygen. It has been shown that by irradiating excitation radiation in the VUV wavelength range, in particular within the wavelength range between 130 nm and 170 nm, the concentration of the oxygen radicals can be clearly controlled without the hydrogen plasma (at least directly) influence.

For the absorption of radiation by oxygen and the associated reactions, see the article "Fligh-temperature absorption studies of the Schumann-Runge band of oxygen at ArF laser wavelengths", RC See et al. , J. Opt. Soc. At the. B, vol. 17, no. 3, 2000, to the article “Absorption and photodissociation in the Schumann-Runge bands of molecular oxygen in the terrestrial atmosphere ”, G. Kockarts, Planet. Space Sci., Vol. 24, Issue 6, pp. 589, 1976, as well as to the article “Polynomial coefficients for calculating O2 Schumann-Runge cross sections at 0.5cm ¹ resolution”, K. Minschwaner et al. , Journal of Geophysical Research, Vol.97, Issue D9, pp10103, 1992, see also the article “Dissociative photoionization cross sections of O2 from threshold to 120 A”, JAR Samson et al., J. Chem. Phys. 76, 393 (1982).

For photo-induced chemical reactions of hydrogen, see the article “Dissociative photoionization of H2 from 18 to 124 eV”, Y. M. Chung et al., J. Chem. Phys. 99, 885 (1993) as well as the article “Photoionization Cross Sections of He and H2”, M. Yan et al., The Astrophysical Journal, 496, pp. 1044-1050, 1998. All of the above articles are incorporated by reference in their entirety into the content of this application.

In a further embodiment, the excitation device is designed to generate the excitation radiation at the at least one excitation wavelength by spectral filtering of (broadband) radiation emitted by the EUV radiation source at wavelengths outside the EUV wavelength range. In particular in the event that the EUV radiation source is a plasma radiation source in which a laser-induced or gas discharge-induced plasma is generated to generate the EUV radiation, so-called “out-of-band” is typically also used in the plasma "Radiation generated at wavelengths outside the EUV wavelength range. To generate the excitation radiation, an excitation wavelength or a (narrower ) The excitation wavelength range which forms the excitation radiation can be filtered. In the event that the excitation radiation has not only one excitation wavelength but a continuous excitation wavelength range, it has proven to be beneficial if this excitation wavelength range is not too large, i.e. it should have a spectral width of approx nm, preferably approx. 20 nm, particularly preferably approx. 10 nm, in particular approx. 1 nm.

In the simplest case, a spectral filter already present in the optical arrangement can be used for spectral filtering, for example a collector mirror, which is used to select the EUV radiation at the operating wavelength from the broadband spectrum of the EUV radiation source. If necessary, the radiation emitted by the EUV radiation source outside the EUV wavelength range - usually outside the EUV beam path - can be filtered with the aid of an additional spectral filter in order to form the excitation radiation. In this case, the excitation device can radiate the excitation radiation into the vacuum housing with the aid of suitable beam-guiding or beam-deflecting optical elements.

In an alternative embodiment, the excitation device has an excitation radiation source for generating the excitation radiation. The excitation radiation source is an additional (secondary) radiation source, i.e. not the EUV radiation source. There are several options for designing the secondary radiation source:

In a further development, the excitation device is designed to form the excitation radiation at the at least one excitation wavelength by spectral filtering of radiation that is emitted by the excitation radiation source. In this case, the excitation radiation source is a broadband radiation source, for example a plasma radiation source, for example in the form of a gas discharge lamp, in particular a spectral lamp, or a laser-plasma radiation source in the form of the EUV radiation source. The radiation emitted by such a broadband radiation source can be filtered through a suitable spectral filter, e.g. is designed as a diffraction grating or as optics operated with grazing incidence (eg “grazing incidence” mirror), can be filtered in order to generate the desired excitation wavelength (s). Even with radiation sources that do not emit radiation in a broadband wavelength spectrum but with several discrete wavelengths, e.g. gas discharge lamps or spectral lamps that only contain a single chemical element in the gas phase, it can be useful to filter out unwanted wavelengths. As described above, the excitation wavelength range of the excitation radiation remaining after the spectral filtering should not be too large, ie it should have a spectral bandwidth of approx. 40 nm, preferably approx. 20 nm, particularly preferably approx. 10 nm , in particular of approx. 1 nm.

In an alternative development, the excitation radiation source is designed to generate the excitation radiation at a single excitation wavelength. In this case, the excitation radiation source is usually a laser source. The laser source can optionally be tailored to emit the desired excitation wavelength, for example it can be a soft x-ray laser, see, for example, the article "Demonstration of a High Average Power Tabletop Soft X-Ray Laser" by BR Benware et al. , Phys. Rev. Lett. 81: 5804 (1998). A high harmony generation (HHG) laser source can also serve as an excitation radiation source. To promote the photo-induced oxygen dissociation reaction described above, the radiation source can in particular be an ArF laser (excitation wavelength 193 nm) or an F2 laser (excitation wavelength 157 nm). The former is with one comparatively large power of more than 90 W available, the latter with a power of about 20 W, see the article “High-power excimer lasers for 157-nm lithography”, S. Spratte et al., Proc. SPIE 5040: 1344-1351, 2003.

These powers are sufficient to enable the desired influencing of the plasma by the excitation radiation.

In a further embodiment, the excitation radiation source is designed to generate the excitation radiation at at least one excitation wavelength at which an effective cross section of the photoinduced reaction with the residual gas excited by the excitation radiation is at least 70%, preferably at least 80%, in particular at least 90% % of a maximum cross-section of the photo-induced reaction with the residual gas is.

As described above, it is beneficial for targeted influencing of the plasma to selectively stimulate one or possibly several desired photoinduced reactions in the residual gas, ideally without stimulating other photoinduced reactions in an undesirable manner. In the event that the excitation radiation has a (narrow) excitation wavelength range of, for example, not more than 40 nm, possibly not more than 20 nm, particularly preferably not more than 10 nm, in particular not more than 1 nm , this excitation wavelength range should also meet the above condition with regard to the cross section. The photo-induced reaction can in particular be a photodissociation or a photoionization reaction of a species contained in the residual gas, for example H2, O2, N2,... Ideally, the photo-induced reaction has its maximum effective cross-section at the excitation wavelength or in the center of the excitation wavelength range. In a further embodiment, the excitation device has a coupling element for coupling the excitation radiation into a beam path of the optical arrangement. The coupling element can be a mirror, for example. The coupling into the beam path of the optical arrangement is favorable, since this passes through the vacuum housing with the at least one optical element, so that the excitation radiation is radiated into the vacuum housing. A prerequisite for this is of course that the coupling element is arranged in the beam path in front of the vacuum housing.

In principle, the excitation radiation can be coupled into the beam path of the optical arrangement at any desired location. The coupling element can be arranged in the beam path of the optical arrangement. In this case, it is advantageous if the coupling-in element is arranged at a location in the beam path which has only a slight influence on the image. This is the case, for example, behind a beam trap for the laser beam that is used to generate the plasma, or on or behind an obscuration screen for shading a partial area of the beam path. There is no need to provide a coupling device if the excitation radiation is radiated directly into the vacuum housing, e.g. via a window, or if the radiation generated by the EUV radiation source is suitably spectrally filtered by an optical element already in the beam path is to form the excitation radiation.

In a further embodiment, the optical arrangement comprises a supply device for supplying at least one gas into the vacuum housing. As described above, it is advantageous if hydrogen is supplied to the vacuum housing as a flushing gas for cleaning the optical elements. The supply device can also be used to supply oxygen or other gases, the components of the in the Form vacuum housing existing residual gas. The supply device can influence the composition of the constituents of the residual gas present in the vacuum housing, whereby the plasma in the vacuum housing can be influenced. However, the composition of the residual gas can generally only be changed within small limits without influencing the image quality or the transmission of the optical arrangement. Correspondingly, the setting of the composition of the residual gas offers only limited and often unsatisfactory possibilities for controlling the flow densities of individual plasma species on the surfaces of the vacuum housing that delimit the plasma.

In a further embodiment, the optical arrangement is designed to generate the plasma through an interaction of the EUV radiation with at least one component of the residual gas which is selected from the group: hydrogen, argon, helium, oxygen and nitrogen. The constituents of the residual gas can be fed to the vacuum housing with the aid of the feed device described above, or they are already present in the vacuum housing anyway. The residual gas in the vacuum housing can, for example, have a residual gas pressure in a range between 0.1 Pa and 20 Pa, in particular between 1 Pa and 10 Pa.

In a further embodiment, the control device is designed to control the excitation device, to change the power of the excitation radiation, the excitation wavelength and / or an intensity distribution of the excitation radiation. By controlling the power of the excitation radiation, the excited photo-induced reaction can be promoted or inhibited in a targeted manner. By changing the excitation wavelength, different photo-induced reactions can be specifically excited or promoted. The change in the excitation wavelength can be implemented, for example, with the aid of a tunable spectral filter, for example by means of a Diffraction grating that can be tilted relative to the incident broadband radiation. A change in the intensity distribution of the excitation radiation can also be beneficial in order to influence the plasma in the vacuum housing in a targeted manner in a location-dependent manner. It goes without saying that the radiation of the excitation radiation can also be changed over time. This is particularly advantageous if the equilibrium state of the plasma has not yet been established, ie in a transient state of the plasma, such as is present, for example, immediately after the optical arrangement has been put into operation.

The change in the properties of the excitation radiation is used to specifically influence the plasma through the excitation of specific photo-induced reactions, for example in relation to the relative flows of different, relevant plasma species on the surfaces exposed to the plasma, in particular on the at least one optical element to optimize suitable quality functions. The properties of the excitation radiation which are suitable for a given lighting setting of the optical arrangement and which are required to optimize the quality function can be stored in a memory device of the control device, for example in a table. Suitable parameters or properties of the excitation radiation can thus be selected for a respective lighting setting that generates an associated spatial intensity distribution of the EUV radiation in the vacuum housing.

With an optimized quality function, the maximum service life of the optical element or elements or other components in the vacuum housing can be achieved. The quality function is optimal for certain target values of target plasma parameters relevant for the plasma, for example for the ratio of the concentrations or the flow rates of different plasma species. The target value or values of at least one target plasma parameter can be obtained from simulations - possibly in combination with the measurement of test plasmas - as described in more detail below.

In a further embodiment, the optical arrangement has at least one sensor for measuring at least one plasma parameter of the plasma formed in the vacuum housing and the control device is designed to adapt the radiation of the excitation radiation into the vacuum housing as a function of the at least one measured plasma parameter. In this embodiment, for example for different values of the measured plasma parameter, associated parameters for the properties (e.g. power, excitation wavelength, intensity distribution, ...) of the excitation radiation can be stored in the control device, which parameters are set with the aid of the control device. The respective parameters of the excitation radiation, which can be stored in a table or a database, are selected in such a way that the plasma in the vacuum housing is optimized in particular with regard to the flow rates on the relevant surfaces.

In a further embodiment, the control device is designed to regulate at least one target plasma parameter of the plasma in the vacuum housing to a predetermined target value when the optical arrangement is in operation, in particular when the plasma is in a transient state. In this case, the target plasma parameter is regulated to the target value on the basis of the at least one measured plasma parameter with the aid of a suitable control circuit that implements a feed-forward control or a feedback control. The measured plasma parameter and the target plasma parameter can be one and the same plasma parameter or different plasma parameters. For example, the target plasma parameter can be composed of or depend on two or more measured plasma parameters. The target plasma parameter can in particular be one for the at least one optical element Act life-determining plasma parameters. If necessary, the control loop can regulate more than one target plasma parameter to a setpoint value.

The control loop can be designed to regulate the target parameter (s) on all relevant time scales. The regulation can also only take place in the steady state of the plasma. The control loop can be designed for rapid regulation in order to enable regulation even in a transient state of the plasma. In order to implement rapid regulation, the excitation device preferably has an excitation radiation source in the form of a laser source, for example in the form of an HHG laser source. At least one parameter or at least one property of the excitation radiation is adapted as a manipulated variable for the regulation.

In a further embodiment, the target plasma parameter forms a ratio of the concentrations of at least two plasma species, preferably a ratio between a concentration of at least one oxygen species, in particular a radical oxygen species 0, and a concentration of at least one hydrogen species, in particular one ionic hydrogen species, e.g. Fh ⁺ . The concentration of the respective plasma species is understood to mean their partial pressure in the vacuum housing or a variable proportional to the partial pressure. Since the ratio of the concentrations of the plasma species forms the target parameter, it is not absolutely necessary to know the absolute concentrations of the respective plasma species in the vacuum housing. In addition to the ratio between a radical oxygen species and an ionic hydrogen species, the ratio between other plasma species can also serve as a target plasma parameter, for example the ratio between nitrogen-containing and carbon-containing plasma species formed by secondary reactions (e.g. N2FT: H , C0 ₂ ⁺ : H ₂ ⁺ , NFU ⁺ : Fh ⁺ , or any combination of the species listed in the next sentence). At the goal Plasma parameters can in particular also be a target concentration of a specific plasma species, possibly also formed by secondary reactions, for example the concentration of (H ⁺ , H2 ⁺ , H3 ⁺ , H, H, N2 ⁺ , N ⁺ , N ²⁺ , 0 ₂ ⁺ , 0 ⁺ , 0 ²⁺ , 0, 0-, 0 ₂ -, 03, 0 ₄ -, 03, H ₂ 0, H ₂ 0 ⁺ , H30 ⁺ , 0H, 0H ⁺ , 0H-, NH, NH ₂ , NH ₃ , NH ⁺ , NH ₂ ⁺ , NH ₃ ⁺ , NH ₄ ⁺ , N ₂ H ⁺ , NO, NO ⁺ , HNO, HN0 ⁺ , C0 ₂ ⁺ , C0, CO ⁺ , C _n H _m , C _n H _m +, C _n H _m Oi, C _n H _m Oi ⁺ , etc., where n, m, l are small (<5) integers that parameterize small organic molecules stoichiometrically correctly) .

As described above, a hydrogen plasma is generated under the influence of EUV radiation in the residual gas containing hydrogen in the vacuum housing, ie ionic hydrogen species (H2 ⁺ , H3 ⁺ , H ⁺ , H, ... ), which have a cleaning effect for carbon which is deposited on the surfaces of optical elements arranged in the vacuum housing. However, the hydrogen plasma also has negative effects, since it can lead to undesirable reduction reactions on other components arranged in the vacuum housing - or on the substrate of the optical elements themselves. Therefore, a small amount of oxygen is added to the residual gas in order to compensate for the undesired reduction reactions by counter-rotating oxidation reactions as far as possible.

Due to the induced by the EUV radiation ionization of molecular oxygen, two positively charged ionic species plasma 0 ₂ ^+, 0 ^+, and a radical plasma species are 0 formed in approximately equal parts. As described above, the positively charged ionic plasma species can, however, lead to adverse effects on components that are also arranged in the vacuum housing. For example, the positively charged ionic oxygen species 0 ₂ ⁺ , 0 ⁺ are accelerated by the positive plasma potential towards the surfaces exposed to the plasma and can lead to implantation or sputtering effects there. The concentration of the oxygen radicals useful in terms of the desired oxidation effect, from which the negatively charged oxygen ion 0 is formed by electron capture, can be increased by the excitation radiation, which promotes the corresponding photo-induced reaction (see above), without to increase the concentration of the positively charged ionic oxygen species 0 ₂ ⁺ , 0 ⁺ . In this way, the partial pressure of the molecular oxygen, which is supplied to the vacuum housing via the supply device, can be reduced as far as possible. The concentration of the oxygen radicals 0, which should be added to the residual gas in order to counteract the reducing effect of the hydrogen plasma, depends on the concentration of the ionic hydrogen species. Therefore, the ratio of the concentration of oxygen radicals to the concentration of ionic hydrogen species, for example of H3 ⁺ , represents a target parameter which, when set or controlled to a predetermined, suitable target value, the service life of the in the vacuum housing arranged components, in particular the reflective optical elements, maximized.

In a further development, the measured plasma parameter is selected from the group comprising: a concentration of at least one ionic, neutral or radical plasma species or a ratio between the concentrations of at least two ionic, neutral or radical plasma species. These and other plasma parameters can be measured with the help of suitable sensors. In the context of this application, the concentration of the plasma species is also understood to mean a flow rate or the number of the respective particles of the plasma species per unit of time, which is measured on a respective surface of a sensor that detects this plasma species. In addition or as an alternative to the plasma parameters mentioned above, other plasma parameters can also be measured with suitable sensors, for example a plasma potential, an electric field strength, a plasma density, an electron density, etc. In a further development, the sensor is selected from the group comprising: ion sensors or radical sensors, in particular emission spectroscopic sensors, absorption spectroscopic sensors, laser fluorescence spectroscopic sensors, mass spectroscopic sensors, retarding field energy (RFEA) analyzers, and Langmuir probes. Ion sensors and radical sensors are used to measure the concentrations or the (possibly relative) flow rates of ionic or radical plasma species. A Langmuir probe enables plasma parameters such as electron density, electron temperature, plasma potential, etc. to be measured. Examples of sensors that can measure the concentration or flow rate of hydrogen radicals during operation of an EUV lithography system are given in DE 102017205870A1, which is incorporated by reference in its entirety into the content of this application.

The optimization of the plasma or the determination of the target values of the target plasma parameters requires as precise a knowledge as possible of the chemical reactions taking place in the vacuum housing in order to be able to specifically set the current densities of the respective species of the plasma on the delimiting surfaces and in this way the To maximize the service life of the optical elements arranged in the vacuum housing, in particular the reflective optical elements.

In principle, one could assume that a suitable setpoint value for a target plasma parameter can be determined experimentally during operation of an optical arrangement for EUV lithography by means of suitable plasma diagnostics under operating conditions.

In practice, however, this approach often fails due to the following points: 1.) During the system design phase of the optical arrangement, the operating conditions, in particular the desired geometry and the radiation sources, cannot yet be implemented.

2.) There are no sensors that can determine the flows chemically, spatially, energetically and, in particular, temporally with sufficient accuracy.

3.) Access to real systems is often not possible to a sufficient extent for operational reasons, especially for long-term studies.

4.) For long-term predictions, “accelerated” tests using other geometries and / or otherwise (e.g. via RF-ICP (radio frequency inductively coupled plasma), RF-CCP (radio frequency capacitively coupled plasma), microwave-coupled or off-line EUV-generated plasmas are used, which may develop a significantly different plasma chemistry.

5.) Low-temperature plasmas develop through a highly non-linear dynamic. Any “simple” extrapolation of experimental data to other space / time scales, geometries or plasma stoichiometries and excitation mechanisms are therefore extremely error-prone or simply impossible.

6.) Optimizing the gas mixtures can be extremely complex experimentally if, for example, explosive (e.g. H2 / O2), corrosive (e.g. HNO3), toxic (e.g.

HCN) mixtures or species arise.

The invention also relates to a method for determining a target value for a target plasma parameter of a plasma that is in operation of an optical Arrangement is formed by an interaction of EUV radiation with a residual gas in a vacuum housing.

The method according to the invention for determining the target value for the target parameter comprises the following steps:

Determining at least one simulated plasma parameter by numerical simulation of the plasma in the vacuum housing,

Determining at least one measured plasma parameter by measuring a plasma in a test setup of the optical arrangement, and also

Determining the setpoint value for the target plasma parameter based on the at least one simulated plasma parameter and based on the at least one measured plasma parameter.

In the method according to the invention, the disadvantages described above are overcome in that a comparison is made between the numerical simulations of the plasma on the one hand and the measurements on real plasmas that may still be in the test phase on the other hand. On the basis of the comparison, values for the target plasma parameters can be determined iteratively, if necessary, and a suitable target value can be established for the respective target plasma parameter.

The target plasma parameter can in particular be a plasma parameter that determines the service life for at least one optical element which is arranged in the vacuum housing, for example a flow rate of at least one plasma species on the surface of the (reflective) optical element. The setpoint value determined with the aid of the method described above can be displayed in an optical Arrangement are stored, which is designed as described above and in which a control loop is implemented, which makes it possible to regulate the target parameter to the desired target value.

The target plasma parameter can be selected, for example, from the group comprising: a concentration of at least one ionic, neutral or radical plasma species or a ratio of the concentrations of at least two ionic, neutral or radical plasma species. In particular, the target plasma parameter can be the above-described ratio between the radical oxygen species 0 and an ionic hydrogen species, for example H3 ⁺ .

Since the geometry of the vacuum housing and the components arranged therein are not yet fixed when the method is carried out, the geometric parameters can also be used as an alternative or in addition to determining the target value for the target plasma parameters by comparing the simulated plasma parameters and the measured plasma parameters Degrees of freedom of the vacuum housing or the components arranged therein are optimized in order to ensure the maximum service life of the optical components in particular.

When determining the at least one simulated plasma parameter, the plasma chemistry, in particular with regard to the flux densities on the surfaces of interest, can be modeled by computer simulations that are as accurate as possible for all relevant plasma configurations. The simulation is preferably carried out under operating conditions over as long periods of time as possible, in the pulsed case, generally the so-called quasiperiodic state, or for cw-excited plasmas for the "steady state". The conditions of the test set-ups for the (specially accelerated) life test or for the determination of the at least one measured plasma parameter are preferably modeled in as much detail as possible, also here up to the (quasi-periodic) "steady state".

In both cases mentioned above, i.e. when determining the simulated and the measured plasma parameter, the transient behavior is preferably determined until equilibrium is reached. The differences or deviations between the simulated plasma and the plasma in the test setup are documented. It is also possible that certain plasma parameters or certain plasma species are taken into account in the simulations that are not taken into account in the test setup, or vice versa.

The cleaning, sputtering, etching rates, etc. of individual flow-induced reactions obtained from the simulations and possibly the experimental tests are preferably converted to the corresponding flows under operating conditions in order to enable an optimal service life prediction under operating conditions. The (geometrical and plasma-chemical) parameters of the test setup can, if necessary, be modified in such a way that the resulting plasma chemistry reproduces the relationships under operating conditions as precisely as possible with regard to a suitably defined quality function.

The parameters to be adapted are not only partial pressures and radiation powers, but also e.g. electrical biasing of the components, geometrical adjustments by means of screens, optimized edges e.g. on the optical elements, electrical antennas etc. Background gas. If necessary, magnetic fields can also be used. The test setup of the optical arrangement can have a simplified geometry, for example in that the vacuum housing is designed as a substantially cuboidal space or the like. The parameters of the real system, ie the optical arrangement, are optimized, if possible, to the extent that the longest possible service life results. The calculated flux densities on the surfaces are preferably used as input for further simulations, e.g. for the induced surface chemistry or diffusion in the near-surface areas of the exposed surfaces in order to characterize the damage mechanisms. It has proven to be beneficial if, for the simulations, the generally very complex plasma chemistry is reduced in advance to a significant subset of relevant species and reactions through detailed modeling. In general, several hundred plasma species and many thousands of reaction channels can be expected, which in a reduced model can be reduced to a double-digit number of plasma species and several hundred reaction channels. In addition, electrical interference signals can also be detected on the relevant surfaces exposed to the plasma.

Various so-called multiphysics methods can be used for the simulation, e.g. phase space-based 5D / 6D kinetic models ("particle in cell Monte-Carlo", PIC-MC, e.g. for modeling a plasma in narrow slots), simplified fluid-based 2D / 3D Models, eg magnetohydrodynamic models or rate equation-based global 0D models (especially for recording extremely complex chemical systems).

Here it is beneficial or necessary to self-consistently calculate the particle dynamics in external electromagnetic fields generated by their own charge and current distribution, taking into account the reaction kinetics (“reactive, charged multifluid flow”).

Further features and advantages of the invention emerge from the following description of exemplary embodiments of the invention, with reference to the figures of the drawing, which show details essential to the invention, and from the claims. The individual features can be individually or individually to be realized in a variant of the invention to several in any combination.

drawing

Exemplary embodiments are shown in the schematic drawing and are explained in the following description. It shows

1 shows a schematic representation of an optical arrangement in the form of an EUV lithography system with an excitation device for irradiating excitation radiation into a vacuum housing,

2a, b schematic representations of two excitation

Radiation sources for generating excitation radiation,

3a, b are schematic representations of the cross sections of several photoinduced reactions of molecular hydrogen,

Fig. 4a, b schematic representations of the effect of a hydrogen

Plasma and an oxygen plasma on an optical element or on a non-optical component, as well

5 shows a schematic representation of a flow chart of a

Method for determining nominal values for a target plasma parameter in the form of a reaction rate or a concentration of a plasma species on a surface of a reflective optical element.

In the following description of the drawings, identical reference symbols are used for identical or functionally identical components. 1 shows schematically the structure of an optical arrangement for EUV lithography in the form of an EUV lithography system 1, namely a so-called wafer scanner. The EUV lithography system 1 has an EUV radiation source 2 for generating EUV radiation 5, which has a high energy density in the EUV wavelength range below 30 nm, in particular between approx. 5 nm and approx. 15 nm. In the example shown, the EUV radiation source 2 is designed in the form of a plasma radiation source. The plasma radiation source 2 has a target 3 in the form of tin droplets, which is bombarded with a laser beam 4. The laser beam 4 is generated by a laser source (not shown), for example a CO2 laser. The laser beam 4 strikes the target 3 and generates a tin plasma which emits the EUV radiation 5. The EUV radiation 5 emanating from the tin plasma at the position of the target 3 is captured by a collector mirror 6, which bundles the EUV radiation 5 into an illumination beam path 7 and in this way further increases the energy density. In order to prevent the laser beam 4 from passing through the illumination beam path 7, the EUV lithography system 1 has a beam trap 8 for the laser beam 4.

The illumination beam path 7 is used to illuminate a structured object M by means of an illumination system 10 which, in the present example, has five reflective optical elements 12 to 16 (mirrors).

The structured object M can be, for example, a reflective photomask which has reflective and non-reflective or at least less strongly reflective regions for generating at least one structure on the object M. Alternatively, the structured object M can be a plurality of micromirrors which are arranged in a one-dimensional or multi-dimensional arrangement and which, if necessary, are at least one axis can be moved in order to adjust the angle of incidence of the EUV radiation on the respective mirror.

The structured object M reflects part of the illumination beam path 7 and forms a projection beam path 9, which carries the information about the structure of the structured object M and which is irradiated into a projection lens 20, which is an image of the structured object M or a respective sub-region thereof on a substrate W generated. The substrate W, for example a wafer, has a semiconductor material, for example silicon, and is arranged on a holder, which is also referred to as the wafer stage WS.

In the present example, the projection objective 20 has six reflective optical elements 21 to 26 (mirrors) in order to generate an image of the structure present on the structured object M on the wafer W. The number of mirrors in a projection objective 20 is typically between four and eight, but if necessary only two mirrors or even ten mirrors can be used.

The collector mirror 6, the reflective optical elements 12 to 16 of the lighting system 10 and the reflective optical elements 21 to 26 of the projection objective 20 are arranged in a vacuum environment. A respective optical element 6, 12 to 16, 21 to 26 is typically arranged in its own vacuum housing, which is also referred to as a “mini environment”. By way of example, FIG. 1 shows such a vacuum housing 27 in which the first reflective optical element 12 of the lighting system 10 is arranged. The illumination beam path 7 enters the vacuum housing 27 at an intermediate focus F and exits the vacuum housing 27 at a further opening (not shown) in the direction of the second optical element 13 of the illumination system 10. It goes without saying that several of the optical Elements 12 to 16, 21 to 26 can be arranged in a respective vacuum housing. The entire EUV lithography system 1 shown in FIG. 1 is additionally surrounded by a housing, not shown, in the interior of which the vacuum housing is arranged. The vacuum housing 27 acts in the manner of a shield and a contamination reduction unit (not shown) is assigned to it, which reduces the partial pressure of contaminating substances at least in the immediate vicinity of the optical surface of the optical element 12 compared to the partial pressure of the contaminating substances in the one surrounding the vacuum housing 27 Interior space reduced. In this regard, reference is also made to WO 2008/034582 A1, which is made part of the content of this application in its entirety by reference.

The EUV lithography system 1 shown in FIG. 1 is designed for an operating wavelength AB of the EUV radiation 5 of 13.5 nm. However, it is also possible for the EUV lithography system 1 to be configured for a different operating wavelength AB of the EUV wavelength range, such as 6.8 nm, for example. Radiation that is emitted by the EUV radiation source 2 at wavelengths other than in the EUV wavelength range is filtered out at the collector mirror 6 and, if necessary, at other spectral filter devices that are not shown in the figure, since this can have an unfavorable effect on the image of the structured object M. .

The EUV lithography system 1 shown in FIG. 1 has an excitation device 30 which is designed to irradiate excitation radiation 31 into the illumination beam path 7 and thus into the vacuum housing 27 at at least one excitation wavelength AA, which is differs from the operating wavelength AB.

In order to generate the excitation radiation 31, the excitation device 30 shown in FIG. 1 is designed, radiation 5a, which is emitted by the EUV- Radiation source 2 is emitted and which does not impinge on the collector mirror 6, via a reflective optical element 32 in the form of a diffraction grating, which is arranged outside the illumination beam path 7, to a reflective coupling element 33. The coupling element 33 is designed as a reflective optical element (mirror) and is used to couple the excitation radiation 31 into the illumination beam path 7 of the EUV lithography system 1. The coupling element 33 is arranged in a region of the illumination beam path 7 , which is shaded by the beam trap 8 for the laser beam 4, so that the coupling of the excitation radiation 31 does not or only slightly influences the imaging or the shaping of the illumination beam path 7. The coupling element 33 can also be integrated directly into the collector mirror 6 as a grid.

In the example shown, the excitation device 30 is designed to generate excitation radiation 31 at excitation wavelengths AA ZU that are outside the EUV wavelength range, specifically at excitation wavelengths AA between 30 nm and 300 nm, for example between 100 nm and 200 nm. The spectral bandwidth or the excitation wavelength range of the excitation radiation 31, which is radiated into the vacuum housing 27 by the excitation device 30, is typically no more than 40 nm, preferably no more than approx nm, particularly preferably not more than approx. 10 nm, in particular not more than approx. 1 nm. The diffraction grating 32 is designed to have the excitation wavelength AA or the excitation wavelength range surrounding it of, for example, AA − 20 nm <A <AA + 20 nm from the broadband emitted radiation 5a of the EUV radiation source 2 to filter.

The EUV lithography system 1 has a control device 34, which can be designed, for example, as a microprocessor, microcontroller or as a control computer and which is designed or programmed, the excitation device 30 via control signals (not shown) head for. In the example shown in FIG. 1, the control device 34 is designed to change the excitation wavelength AA or the excitation wavelength range by rotating the reflective diffraction grating 32, which serves as a spectral filter, and thus the angle of incidence on the diffraction grating 32 impinging radiation 5a of the EUV radiation source 2 is changed, which results in a shift in the filtered excitation wavelength AA or the excitation wavelength range. In this example, the coupling-in element 33 is also rotatably mounted in order to ensure that the excitation radiation 31 reflected at the diffraction grating 32 is coupled into the illumination beam path 7.

As an alternative to generating the excitation radiation 31 using the radiation 5a generated by the EUV radiation source 2, the excitation device 30 can have its own excitation radiation source 35a, 35b for generating the excitation radiation 31, as is exemplified in FIG. 2a , b is shown. The excitation radiation source 35a shown in Fig. 2a is designed as a broadband radiation source, ie it generates radiation 36 in a comparatively large wavelength range, which is filtered by a spectral filter 37 operated in transmission in order to absorb the excitation radiation 31 at the (at least one) To generate excitation wavelength AA and in this way to avoid excitation radiation 31 being radiated into vacuum housing 27 at undesired excitation wavelengths AA. The broadband excitation radiation source 35a in the example shown is a gas discharge lamp, but it can also be a different type of broadband radiation source, for example a plasma radiation source such as the EUV radiation source 2.

The excitation device 30 shown in Fig. 2b differs from the excitation device 30 shown in Fig. 2a in that the excitation Radiation source 35b is designed to emit radiation at a single excitation wavelength A _A. The excitation radiation source 35b shown in FIG. 2b is a laser source in the form of a soft x-ray or VUV laser which is tailor-made for the emission of the desired excitation wavelength AA.

The excitation radiation sources 35a, b of FIGS. 2a, b are designed to adjust or change the power p of the emitted radiation 36 or the power of the excitation radiation 31. The coupling element 33 is rotatably mounted and makes it possible - within certain limits - to change the intensity distribution of the excitation radiation 31 radiated into the vacuum housing 27. It goes without saying that a change in the intensity distribution of the excitation radiation 31 can also be achieved with the aid of a suitable intensity filter (gray filter) which is introduced into the beam path of the radiation 36 emitted by the excitation radiation source 35a or the excitation radiation 31, or can take place with the aid of another optical element influencing the intensity distribution.

The total output of the excitation radiation source 35a, b can also be controlled or regulated either by internal setting options or gray filters.

The activation of the excitation device 30 with the aid of the control device 34 enables at least one photo-induced reaction of the excitation radiation 31 with a residual gas 40 contained in the vacuum housing 27 (cf. FIGS. 2a, b) to be specifically excited or - if desired - to avoid such a photo-induced reaction as far as possible by not irradiating any excitation radiation 31 into the vacuum housing 27. In the example shown, the residual gas 40, which is contained in the vacuum housing 27, is a mixture of several gases, namely (molecular) hydrogen, oxygen, argon, helium and nitrogen. The total pressure of the residual gas 40 in the Vacuum housing 27 is typically between about 0.1 Pa and 20 Pa. The total pressure of the residual gas 40 is generated by means of a vacuum pump (not shown) and can be changed within certain limits.

While certain constituents of the residual gas 40 are already present in small quantities in the vacuum housing 27, other constituents that serve as flushing gases 42, in particular hydrogen and oxygen, are fed to the vacuum housing 27 via a feed device 41 in the form of a tubular gas feed. The supply device 41 can control or, if necessary, regulate the flow of the flushing gas or the flow of the individual components of the flushing gas 42 via a controllable valve or the like.

By interaction of the residual gas 40 with the EUV radiation 5, a plasma 43 is generated in the vacuum housing 27 during operation of the EUV lithography system 1, ie ionic, neutral (excited) and radical plasma species of the residual gas 40 are formed. In Fig. 2a, b this is shown by way of example for molecular hydrogen H2 as a component of the residual gas 40, in which the interaction with the EUV radiation 5 for the formation of positively charged ionic plasma species (H2 ⁺ , H3 ⁺ ) and for the formation of radical (neutral) plasma species Fl leads. It goes without saying that ionic, neutral and radical plasma species of most of the other constituents contained in the residual gas 40 are also generated. Their representation has been omitted in FIGS. 2a, b for reasons of clarity.

The hydrogen plasma 43 in the vacuum housing 27 produces a cleaning effect on the optical element 12 arranged in the vacuum housing 27 by removing carbon from the surface of the optical element 12 which is exposed to the plasma 43. The ionic, neutral and radical hydrogen plasma species can also be undesirable Reduction reactions on other components arranged in the vacuum housing 27 result, as will be described in more detail below.

The supply of the flushing gas 42 in the form of molecular hydrogen is controlled with the aid of the control device 34 in order to achieve the desired cleaning effect of the hydrogen plasma species. In order to counteract the reducing effect of the hydrogen plasma 43, the vacuum housing 27 is additionally supplied with a small proportion of molecular oxygen O2 via the supply device 41. The setting of the ratio of molecular hydrogen H2 to molecular oxygen O2, however, offers only comparatively few possibilities for influencing the plasma 44.

In order to optimize the effect of the plasma 44 on the optical element 12, ie to achieve the longest possible service life of the optical element 12, the excitation radiation 31 radiated into the vacuum housing 27 with the aid of the excitation device 30 described above is used, in order to stimulate specific photo-induced reactions in the residual gas 40 in a targeted manner.

This makes use of the fact that the effective cross section o of the residual gas 40 with a respective photon g of the excitation radiation 31 is dependent on the wavelength AA of the excitation radiation 31 or on the respective photon energy (in eV), as in FIGS. 3a, b for several photoinduced reactions with molecular hydrogen H2 is shown. The representations of the cross-section o of FIG. 3a and FIG. 3b differ only in the different scaling.

The EUV radiation 5 at the operating wavelength AB = 13.5 nm has a photon energy of 92 eV. As can be seen in particular in FIG. 3a, the photo-ionization reaction of hydrogen H2 with a photon g

at a photon energy of 92 eV a comparatively large cross-section compared to the other reactions considered, but this increases to smaller photon energies, i.e. the majority of the H2 ⁺ ions are not generated with broadband irradiation at the nominal photon energy of 92 eV, but with lower photon energies.

In order to specifically stimulate the photo-ionization reaction of molecular hydrogen h to H ₂ ⁺ , ie to intensify the formation of H2 ⁺ , excitation radiation 31 at lower photon energies, for example at less than approx. 40 eV, can enter the vacuum housing 27 are irradiated, whereby the proportion of H2 ⁺ ions in the residual gas 40 is increased. In the example shown in FIG. 3 a, for this purpose, excitation radiation 31 is radiated into vacuum housing 27 at exactly one excitation wavelength AA of 72.9 nm (corresponding to 17 eV). In this way, the ratio of the F molecules and H ⁺ ions to the Py ions in the residual gas 40 changes.

This is particularly relevant because the H2 ⁺ ions result from the very dominant reaction

H ₂ ⁺ + H ₂ -> H ₃ ⁺ + H atomic hydrogen H, ie hydrogen radicals, are formed, so that by this accompanying phenomenon the hydrogen radical flows are (partially) _{coupled to the H 2} ⁺ flow. By means of the targeted irradiation of the excitation radiation 31, the proportion of the radical hydrogen species H in the plasma 43 can therefore also be controlled. In the example shown in FIG. 3b, the excitation radiation 31 is radiated into the vacuum housing 27 at (exactly one) excitation wavelength AA of 33.5 nm (corresponding to 37 eV) in order to specifically trigger the direct photo-induced reaction

to stimulate the formation of radical hydrogen species H.

For the most efficient possible excitation of the respective photo-induced reaction, it is advantageous if the excitation wavelength AA or the excitation wavelength range is as close as possible to the wavelength at which the cross section of the respective reaction has a local or a global maximum OMAX. In particular, it is advantageous if the excitation wavelength AA or all excitation wavelengths AA of an excitation wavelength range of the excitation radiation 31 are in a wavelength interval in which the cross section o the photo-induced reaction excited by the excitation radiation 31 with the Residual gas 40 is at least 70%, possibly at least 80% or at least 90% of the maximum effective cross-section O _max of the photoinduced reaction with the residual gas 40 excited by the excitation radiation 31, as is the case in the example shown in FIG. 3a.

In addition to the photo-induced reactions described above, other reactions can of course also be used, for example the direct photo-dissociation reaction of hydrogen H2

can be influenced in a targeted manner in the manner described above, ie by the irradiation of excitation radiation 31. To optimize the effect of the plasma 40 with regard to the service life of the optical element 12, which is arranged in the vacuum housing 27, it has proven to be advantageous if the control device 34 controls the excitation radiation 31 radiated into the vacuum housing 27 as a function of at least adapts to a measured plasma parameter. For this purpose, two sensors 44, 45 are arranged in the EUV lithography system 1 shown in FIGS. 2 a, b, the first of which is designed as an ion sensor 44 and the second of which is designed as a radical sensor 45. The ion sensor 44 can be, for example, an emission spectroscopic sensor, an absorption spectroscopic sensor, a laser fluorescence spectroscopic sensor, a mass spectroscopic sensor (ie a mass spectrometer) or a retarding field energy analyzer.

The radical sensor 45 for the detection of hydrogen radicals H can for example have a material which is reversibly hydrogenatable, i.e. which reversibly forms a hydride with hydrogen, as described in DE 102017205870 A1. Some of the sensor types described above, which are used as ion sensors, can also be used for the quantitative measurement of the concentration of hydrogen radicals or of other radical species in the plasma 43. A laser fluorescence spectroscopic sensor can, for example, measure the concentration of OH radicals. Absorption spectroscopic sensors or emission spectroscopic sensors can also be used to measure the concentration of radicals.

In the examples shown in FIGS. 2a, b, the ion sensor 44 is designed to determine the (instantaneous) concentration c (H2 ⁺ ), c (H3 ⁺ ) of the ionic hydrogen species H2 ⁺ or H3 ⁺ ZU, which of the (instantaneous ) Flow rate of the ionic hydrogen species H2 ⁺ , H3 ⁺ at the location of the sensor surface of the ion sensor 44 corresponds. In the example shown, the radical sensor 45 is designed to measure the (instantaneous) concentration c (H) of the radical hydrogen species H at the location of the sensor surface of the radical sensor 45. It goes without saying that the sensors 44, 45 can also be designed to measure other plasma parameters than the concentrations of the above-mentioned hydrogen species H2 ⁺ , H3 ⁺ , H ZU, for example the concentration of radical oxygen species 0, etc. For this purpose For this purpose, further sensors can be arranged in the vacuum housing 27. Certain types of sensors 44, 45 also make it possible to measure directly the ratio of the concentrations of specific ionic and radical species, for example the concentration ratio c (H3 ⁺ ) / c (H). In the vacuum housing 27, other types of sensors can also be arranged which enable plasma diagnostics, for example Langmuir probes, etc.

^{Using the plasma parameters C (H2 +} ), C (H3 ⁺ ), C (H), ... measured with the aid of the sensors 44, 45, the control device 34 can act on the excitation device 30 to generate the plasma 43 in the vacuum housing 27 can be influenced in a targeted manner. The control device 34 can in particular have a control device, for example in the form of a linear controller, for example a PID controller, or possibly a non-linear controller, in order to control a target plasma parameter to a predetermined setpoint value. In the event that the regulation or the response times of the excitation device 30 are fast enough, the regulation of the target plasma parameter to the desired value can also take place in the transient state of the plasma 43, ie before a stationary or quasi-stationary one The equilibrium state of the plasma 43 is established. The target plasma parameter can be one of the measured plasma parameters, i.e. for example the concentration c (H2 ⁺ ), c (H3 ⁺ ), c (H), ... of ionic or radical hydrogen species H2 ⁺ , H3 ⁺ , H, the concentration ratio of ionic to radical hydrogen species, e.g. c (H3 ⁺ ) / c (H), ... or a combination of the measured plasma parameters. It goes without saying that the goal- Plasma parameters can generally be a variable which is dependent on the measured plasma parameters and which can be influenced by the excitation device 30 or the excitation radiation 31 in such a way that it can be regulated to the desired target value.

^{The concentration ratio c (O) / (C (H3 +} )) of the concentration c (O) has proven to be a particularly suitable target plasma parameter for the regulation, which has a direct effect on the service life of the optical element 12 arranged in the vacuum housing 27 ) of the radical oxygen species 0 to the concentration c (H3 ⁺ ) of the ionic hydrogen species H3 ⁺ . The reason for this fact is explained below with reference to FIGS. 4a, b.

FIG. 4a shows the reflective optical element 12 from FIG. 1 and a further component 50 which is also arranged in the vacuum housing 27 shown in FIG. 1 and which contains silicon (Si). The silicon-containing component 50 can be, for example, a sensor, an actuator, ... ULE®), or made of a glass ceramic.

The optical element 12 also has a substrate 51 made of titanium-doped quartz glass, to which a reflective multilayer coating 52 is applied, which is optimized for the reflection of EUV radiation 5 at the operating wavelength AB of 13.5 nm. The multilayer coating 52 has alternating layers 53a, 53b made of molybdenum and silicon. At the operating wavelength AB of 13.5 nm, the silicon layers 53b have a higher real part of the refractive index than the molybdenum layers 53a. Depending on the exact value of the operating wavelength AB, there are other material combinations such as molybdenum and beryllium, ruthenium and beryllium or lanthanum and B ₄ C also possible. To protect the multilayer coating 52, a protective layer 54 made of ruthenium is applied to it.

As indicated in FIG. 4a, a native SiC layer on the surface of component 50 becomes Si on contact with ionic hydrogen species H3 ⁺ , which are formed when interacting with EUV radiation 5, according to the following cumulative reaction reduced:

4 H3 ⁺ (g) + 3 S1O2 (s) +4 e - »3 Si (s) + 6 H2O (g).

As a result of the attack of radical hydrogen species H from the plasma 43, gaseous silane (SiFU) is formed from the silicon according to the following reaction equation:

At the surface 54a of the Ru protective layer, there is degradation or severe reflectivity losses due to the accumulation and possibly subsequent oxidation of silicon:

SihU (g) + Ru (s) - »Si-Ru (s) + 2 H2 (g).

Si-Ru (s) +20 (g) - »SiC> 2 (s) + Ru (s)

The Si deposition on the surface 54a of the mirror 12 is thus driven by the concentration of the ionic hydrogen species, in particular in the form of Fh ⁺ . In particular, the S1O2 deposited on the mirror leads, for example, to a reduction in reflectivity.

As described above in connection with FIGS. 2a, b, such a degradation of the surface is avoided as completely as possible 54a of the reflective optical element 12 of the vacuum housing 27 in the purge gas 42 supplied via the supply device 41, in addition to molecular hydrogen H2, oxygen O2 is also supplied. The addition of molecular oxygen O2 leads to the restoration of the native Si0 ₂ layer on component 50, as indicated in FIG. 4b and described in detail below:

Due to the interaction with the EUV radiation 5, the molecular oxygen O2 essentially forms equal parts of 0 ₂ ⁺ , 0 ⁺ and 0, i.e. positively charged ionic plasma species 0 ₂ ⁺ , 0 ⁺ and radical plasma species 0, in particular according to the following reaction equations:

Y (92eV) + 0 (g) - »0 ₂ ⁺ (g) + e-

The exposed silicon Si oxidizes to S1O2 according to the following reaction equations:

Si (s) + 20 (g) - »S1O2 (s)

Si (s) + 20 (g) - »S1O2 (s) + 2e

As a result of the attack by radical oxygen species 0 and ionic oxygen species 0 from the plasma 43, the silicon surface can be restored again, ie the native SiO ₂ layer is formed again.

The restoration rate is driven by the formation of radical oxygen species 0 43 in the plasma Ionic oxygen species 0 ₂ ^+, 0 ⁺ can on the other hand to adverse effects on components 50, lead 12 in the vacuum housing 27, for example, be this through the positive The plasma potential is accelerated towards the surfaces exposed to the plasma 40 and can lead to undesired implantation and sputtering effects there. In the case of radical oxygen species 0, this is the case to a much lesser extent.

In order to stimulate or intensify the formation of the radical oxygen species 0 in the plasma 40, it is advantageous if the excitation device 30 is designed to generate excitation radiation 31 at excitation wavelengths AA in the VUV wavelength range, more precisely at more than 100 nm to radiate into the vacuum housing 27. This takes advantage of the fact that oxygen tends to ionization reactions at wavelengths below 100 nm, but has some absorption bands at wavelengths of more than 100 nm that lead to the dissociation of the molecular oxygen O2:

Y + O2 (g) - »2 O.

In particular, the irradiation of excitation radiation 31 at excitation wavelengths AA is in the range of the so-called Schumann-Runge band around a wavelength of approx. 193 nm or in the so-called Schumann-Runge continuum between approx. 130 nm and approx. 170 nm cheap. For this purpose, the excitation device 30 shown in FIG an F2 laser for generating excitation radiation 31 at an excitation wavelength AA of 157 nm.

With both, generally finely tuned, excitation wavelengths AA, the reaction is g (193 nm) + O2 (g) - »2 O Y (157 nm) + O2 (g) - »2 O favored.

The promotion of the photo-induced dissociation reactions described above increases the proportion of radical oxygen species 0 compared to the ionic, positively charged oxygen species 2 0 ⁺ , since the latter has a very small cross-section at the above-described excitation wavelengths AA of more than 100 nm exhibit. The formation of hydrogen plasma species is likewise not influenced by the irradiation of the excitation radiation 31 at excitation wavelengths AA of more than 100 nm, since (molecular) hydrogen is practically transparent for these wavelengths.

Instead of an excitation radiation source 35b in the form of a laser source for exciting the photo-induced dissociation reaction of oxygen, a broadband excitation radiation source 35a can also be used for generating broadband excitation radiation 31, from which an interval of excitation wavelengths AA or a single Excitation wavelength AA is filtered out, which is, for example, between about 130 nm and about 170 nm.

By varying the power p or the intensity of the excitation radiation 31 irradiated at the excitation wavelength (s) AA described above, the concentration of radical oxygen species O can be controlled without (directly) the hydrogen plasma or to influence the concentration of positively charged ionic oxygen species 0 ⁺ , 0 ²⁺ . By varying the power p of the excitation radiation 31, the target plasma parameter can be the concentration ratio c (O) / C (H3 ⁺ ) of the concentration c (O) of the radical oxygen species O to the concentration c (H3 ⁺ ) the radical hydrogen species H3 ⁺ , which has a direct effect on the service life of the reflective optical element 12. In the example shown, the control device 34 is used to regulate the concentration ratio c (O) / c (H ₃ ⁺ ) to a predetermined target value (c (O) / c (H3 ⁺ )) soii. For this purpose, the power p of the excitation radiation 31 is dependent on the concentrations c (O) or c (H ₃ ⁺ ) of the radical oxygen species 0 and the radical hydrogen species H3 measured with the aid of the sensors 44, 45 ⁺ or the actual value of the concentration ratio c (O) / C (H3 ⁺ ) formed from the two concentrations is determined and regulated to the target value (c (O) / c (H3 ⁺ )) soii. The regulation can in particular be fast enough to increase the concentration ratio c (O) / C (H3 ⁺ ) to the target value (c (O) / c (H3 ⁺ )) soii even in a transient state of the plasma 43 rules.

In addition to regulating the concentration ratio c (O) / c (H3 ⁺ ), it is advantageous if the basic equilibrium of molecular hydrogen H2 to molecular oxygen O2 in the vacuum housing 27 is suitably set. The setting takes place in that the control device 34 acts on the feed device 41 in order to set the ratio of the purge gas flows of molecular hydrogen and molecular oxygen.

A typical order of magnitude for the ratio of molecular oxygen O2 supplied in the purge gas 42 to molecular hydrogen H2 for generating an approximate balance between reduction and oxidation in the plasma 40 is the interaction caused _{by the EUV radiation 4 at the operating wavelength A B} on the order of about 1 ppb O2 to H2.

Simulations have shown that when excitation radiation 31 is irradiated at an excitation wavelength AA of 157 nm, significantly higher formation rates of oxygen radicals 0 are possible than when excitation radiation 31 is irradiated at an excitation wavelength AA of 193 nm . The rate of formation of atomic oxygen 0 at 157 nm can be increased almost by a factor of 20 compared to the rate of formation in the interaction with EUV radiation 4 at the operating wavelength A _{A of 13.5 nm.} Basically, in the photo-induced reaction of molecular oxygen O2 with EUV radiation 4 at the operating wavelength AA of 13.5 nm, approx. 0.5 O atoms are formed, i.e. the dissociative ionization has a share of 50% while at 157 nm two O atoms are formed per dissociation. Even if the rate of formation of oxygen radicals O at 193 nm is lower than at 157 nm, the rate of formation of oxygen radicals O is still significantly greater than at 13.5 nm, so that even when excitation radiation 31 is irradiated during an excitation - Wavelength AA of 193 nm the concentration c (O) of radical oxygen species O in the plasma 43 can be increased.

To determine a suitable target value (c (O) / c (H3 ⁺ )) soii for the target plasma parameter in the form of the concentration ratio c (O) / c (Fl ₃ ⁺ ) or for another suitable target plasma parameter , which influences the service life of the reflective optical element 12, a method can be carried out which is described below by way of example with reference to the flowchart shown in FIG. 5.

In a first step S100 of the method, the chemical and geometric plasma degrees of freedom are defined. In a subsequent step S101, at least one simulated plasma parameter is determined by numerical simulation of the plasma 40 in the vacuum housing 27 or a suitable test plasma. In the example shown, the simulated plasma parameters are the flux densities of all relevant plasma species. In a further step S102, at least one plasma parameter is determined experimentally by measuring a plasma in a test setup of the optical arrangement in the form of the EUV lithography system 1. In the example shown, the measured plasma parameters are the reaction rates induced by the flows of all relevant plasma species (cleaning, etching, in solid diffusion, sputtering, etc.). In addition, in a further step S103, the EUV plasma is modeled in the EUVL scanner to determine the flow rates of all relevant plasma species.

In a subsequent step S104, the plasma parameters simulated or experimentally determined in steps S101 and S102 are combined in order to determine the flow-normalized rate coefficients. In a subsequent step S105, the reaction rates induced by the flows of all relevant plasma species (cleaning, etching, in solid diffusion, sputtering, etc.) are determined in the scanner, i.e. in the EUV lithography system 1. Here, the generally non-linear coupling of the plasma parameters and the flow rates of all relevant plasma species determined in step S103 in the EUVL scanner are taken into account.

In the event that the reaction rates do not match a predefined criterion for the service life of the reflective optical element 12 in the vacuum housing 27, a return is made to step S100 and the method is carried out again with a modified definition of the chemical and geometric plasma degrees of freedom until the predefined Criterion is met.

The desired value (c (O) / C (H3 ⁺ )) S _O M determined in step S105 for the target parameter relevant to service life, for example the concentration ratio of the radical oxygen species 0 to the positively charged ionic hydrogen species Fh ⁺ , e.g. can be of the order of magnitude of 10 ⁶ , is used as a specification for the target value of the regulation of the EUV lithography system 1 described above.

It goes without saying that the method shown in FIG. 5 is greatly simplified and only covers the essential steps in determining suitable target or target values for the plasma 43 in the vacuum housing 27 are shown.

It is also understood that the irradiation of excitation radiation 31 in the vacuum housing 27 can excite photoinduced reactions in addition to hydrogen and oxygen in the case of other gases contained in the plasma 43 or in the residual gas 40, for example in the case of nitrogen (N2) . In the case of nitrogen, analogously to hydrogen, the irradiation of excitation radiation 31 at wavelengths higher than the operating wavelength A _B results in the proportion of N2 ⁺ ions that are formed during the photo-ionization reaction and the proportion of N ⁺ - Ions or N-radicals, which are formed during the photo-dissociation reaction, can be influenced or adjusted.

Claims

1. A method for operating an optical arrangement (1) for EUV lithography, comprising:

Generating EUV radiation (5) at an operating wavelength (AB) in the EUV wavelength range by means of an EUV radiation source (2) of the optical arrangement (1), and

Generating a plasma (43) in a vacuum housing (27) of the optical arrangement (1), in which at least one optical element (12) is arranged, by interaction of the EUV radiation (5) with one contained in the vacuum housing (27) Residual gas (40), characterized by

Radiation of excitation radiation (31) into the vacuum housing (27) at at least one excitation wavelength (AA), which differs from the operating wavelength (AB), by means of an excitation device (30), and

Targeted excitation of at least one photo-induced reaction in the residual gas (40) in the vacuum housing (27) by controlling the excitation device (30).

2. Optical arrangement (1) for EUV lithography, comprising: an EUV radiation source (2) for generating EUV radiation (5) at an operating wavelength (AB) in the EUV wavelength range, at least one vacuum housing (27) , in which at least one optical element (12) is arranged and which contains a residual gas (40), wherein during operation of the optical arrangement (1) a plasma (43 ) is generated in the vacuum housing (27), characterized by an excitation device (30) which is used for irradiating excitation radiation (31) into the vacuum housing (27) in at least one Excitation wavelength (AA) is formed, which differs from the operating wavelength (AB), and a control device (34), which is designed to control the excitation device (30) during operation of the optical arrangement (1), in particular in order to specifically stimulate at least one photo-induced reaction in the residual gas (40) in the vacuum housing (27).

3. Optical arrangement according to claim 2, in which the excitation device (30) is designed to radiate the excitation radiation (31) into the vacuum housing (27) in a wavelength range between 30 nm and 300 nm, preferably between 100 nm and 200 nm.

4. Optical arrangement according to claim 2 or 3, wherein the excitation device (30) is designed to form the excitation radiation (31) at the at least one excitation wavelength (AA) by spectral filtering of radiation (5a), the is emitted by the EUV radiation source (2) at wavelengths outside the EUV wavelength range.

5. Optical arrangement according to claim 2 or 3, in which the excitation device (30) has at least one excitation radiation source (35a, b) for generating the excitation radiation (31).

6. Optical arrangement according to claim 5, wherein the excitation device (30) is designed to form the excitation radiation (31) at the at least one excitation wavelength (AA) by spectral filtering of radiation (36) from the excitation radiation source (35a) is emitted.

7. Optical arrangement according to claim 5, in which the excitation radiation source (35b) is designed to generate the excitation radiation (31) at a single excitation wavelength (AA).

8. Optical arrangement according to one of claims 2 to 7, wherein the excitation device (30) for generating the excitation radiation (31) is formed at at least one excitation wavelength (AA), in which an effective cross section (s) of by the excitation radiation (31) excited photoinduced reaction with the residual gas (40) is at least 70% of a maximum cross section (o _max ) of the photoinduced reaction with the residual gas (40) excited by the excitation radiation (31).

9. Optical arrangement according to one of claims 2 to 8, in which the excitation device (30) has an optical coupling element (33) for coupling the excitation radiation (31) into a beam path (7, 9) of the optical arrangement ( 1).

10. Optical arrangement according to one of claims 2 to 9, further comprising: a feed device (41) for feeding at least one flushing gas (42) into the vacuum housing (27).

11. Optical arrangement according to one of claims 2 to 10, which is designed to generate the plasma (43) by interaction of the EUV radiation (5) with at least one component of the residual gas (40) which is selected from the group: Hydrogen, argon, helium, oxygen, nitrogen and carbon dioxide.

12. Optical arrangement according to one of claims 2 to 11, in which the control device (34) is designed to control the excitation device (30), the power (p) of the excitation radiation (31), an intensity distribution of the excitation radiation (31) and / or change the excitation wavelength (AB) ZU.

13. Optical arrangement according to one of claims 2 to 12, further comprising: at least one sensor (44, 45) for measuring at least one plasma parameter (c (H2 ⁺ ), c (H3 ⁺ ), c (H), c (H3 ⁺⁾ ) / c (H)) of the plasma (43) formed in the vacuum housing (27), the control device (34) being designed to generate the excitation radiation (31) radiated into the vacuum housing (27) as a function of the at least one measured Change plasma parameters (c (H2 ⁺ ), c (H3 ⁺ ), c (H)).

14. Optical arrangement according to claim 13, in which the control device (34) is designed, during operation of the optical arrangement (1), in particular in a transient state of the plasma (43), at least one target plasma parameter (c (O) / c (H ₃ ⁺ )) to a predetermined target value ((c (O) / c (H3 ⁺ )) soii) to regulate.

15. Optical arrangement according to claim 13 or 14, in which the target plasma parameter forms a ratio of the concentrations (c (H3 ⁺ ), c (O)) of at least two plasma species (H3 ⁺ , 0), preferably a ratio between a concentration (c (0)) of at least one oxygen species, in particular a radical oxygen species (0), and a concentration (c (H3 ⁺ )) of at least one hydrogen species, in particular an ionic hydrogen species (H3 ⁺ ) .

16. Optical arrangement according to one of claims 13 to 15, in which the measured plasma parameter (c (H2 ⁺ ), c (H3 ⁺ ), c (H)) is selected from the group comprising: a concentration (c (H2 ⁺ ) , c (H3 ⁺ ), c (H)) of at least one ionic, neutral or radical plasma species (H2 ⁺ , H3 ⁺ , H) or a ratio between the concentrations of at least two ionic, neutral or radical plasma species (c (H3 ⁺ ) / c (H)).

17. Optical arrangement according to one of claims 13 to 16, in which the sensor (44, 45) is selected from the group comprising: ion sensors (44) and radical sensors (45), in particular emission spectroscopic sensors, absorption spectroscopic sensors, laser fluorescence spectroscopic sensors, mass spectroscopic sensors , Retarding Field Energy Analyzers, and Langmuir Probes.