DE102024104118A1 - Method for producing an optical imaging system for a microlithography system - Google Patents

Abstract

In a method for manufacturing an optical imaging system for a microlithography system, a basic imaging system is first constructed by installing optical modules, each having a manipulable optical element, at associated installation positions. The basic imaging system is then adjusted in an adjustment process based on at least one system measurement of the imaging quality performed with a wavefront measuring system by controlling manipulators to change the optical effect of associated optical elements in order to obtain an (adjusted) useful configuration in which the optical imaging system has a specified imaging quality. The adjustment process automatically runs through a first stage in the form of a coarse adjustment and a second stage in the form of a fine adjustment, wherein the measuring system is operated with a first measurement accuracy in the first stage and with a second measurement accuracy that is higher than the first measurement accuracy in the second stage. The coarse adjustment is terminated and the fine adjustment is automatically initiated when a predefinable first adjustment state is reached, which is defined by at least one predefinable termination criterion, wherein a first termination criterion is met when the system measurement shows that an image quality error is smaller than a predefinable first error limit and a second termination criterion is met when the image quality error is still greater than or corresponds to the first error limit when approaching the first error limit, but at the same time a predefinable improvement potential limit of the coarse adjustment is reached.

Description

FIELD OF APPLICATION AND STATE OF THE ART

The invention relates to a method for producing an optical imaging system for a microlithography system and to an optical imaging system for a microlithography system.

A preferred area of application is the manufacture or restoration of optical imaging systems designed as projection lenses for projection exposure systems. The process can be used, for example, in the manufacture or restoration of optical imaging systems designed as projection lenses for EUV projection exposure systems or EUV mask inspection systems for inspecting masks (reticles) for EUV microlithography. These operate with wavelengths in the extreme ultraviolet (EUV) range. The process can also be used in the manufacture of imaging systems for other wavelength ranges, e.g., for wavelengths in the deep ultraviolet (DUV) range.

Microlithographic projection exposure processes are predominantly used today to manufacture semiconductor devices and other finely structured components, such as masks for photolithography. These involve masks (reticles) or other pattern-generating devices that carry or form the pattern of a structure to be imaged, for example, a line pattern of a layer of a semiconductor device. The pattern is positioned in a projection exposure system between an illumination system and an optical imaging system, usually referred to as a projection lens or projection optics, in the region of the object plane of the imaging system and illuminated with illumination radiation shaped by the illumination system. The radiation modified by the pattern travels along an imaging beam path through the imaging system, which images the pattern on a reduced scale onto the substrate to be exposed. The surface of the substrate is positioned in the image plane of the imaging system, which is optically conjugate to the object plane. The substrate is usually coated with a radiation-sensitive layer (resist, photoresist).

One of the goals in the development of projection exposure systems is to lithographically create structures with increasingly smaller dimensions on the substrate, for example, to achieve higher integration densities in semiconductor components. One approach is to work with shorter wavelengths of electromagnetic radiation. Systems for the deep ultraviolet range (e.g., at wavelengths below 260 nm, e.g., at approximately 193 nm) are widely used today. They often consist only of lenses (dioptric systems) or a combination of lenses and at least one mirror (catadioptric systems). Optical systems have also been developed that use electromagnetic radiation from the extreme ultraviolet range (EUV) with operating wavelengths in the range between 5 nanometers (nm) and 30 nm, particularly at 13.5 nm. Imaging systems for EUV microlithography systems use exclusively mirrors to image structures from the object plane to the image plane, for example, from a reticle to a wafer.

Many modern EUV projection lenses have a modular design. Optical modules are installed in a common support frame (force frame) at assigned mounting positions. Each optical module has a mirror. When fully assembled, these modules are located in the designated mounting positions for the optical modules. Together with the optically effective surfaces of the mirrors, they form the imaging beam path. DUV systems can have optical modules with either a lens or a mirror.

In imaging systems of the type considered here, it is possible to at least partially compensate for imaging errors by using manipulators. The term "manipulator" refers to a device configured to actively influence an associated optical element based on control signals from an operating control system in order to change its optical effect, in particular to change it in such a way that an occurring error is at least partially compensated. A manipulator has one or more actuators integrated into an optical module, whose current control value can be changed or adjusted based on control signals from the operating control system.

A manipulator can, for example, be designed to decenter an optical element along or perpendicular to a reference axis, tilt an optical element, locally or globally heat or cool an optical element, and/or deform an optical element. If a change in the control value involves the movement of an actuator, e.g., to move or tilt an optical element according to its rigid-body degrees of freedom, a change in the control value can also be referred to as the "manipulator travel distance."

To ensure the most precise imaging possible, projection lenses with excellent image quality, in particular with the lowest possible wavefront aberrations, are required. The production of such an optical imaging system is extremely time-consuming. Production includes both the manufacture of the individual components of the optical imaging system (e.g. mirrors and/or lenses) and the assembly of the optical imaging system with its many individual components. An optical imaging system for an EUV microlithography system has several optical modules, each carrying a mirror, along an imaging beam path leading from an object plane to an image plane of the imaging system. An optical module comprises the mirror itself as well as devices connected to the mirror that enable the optical module to be mounted in the support frame of the optical imaging system.

First, a basic imaging system is constructed by installing the optical modules in their corresponding installation positions on a support frame. Even with high-precision manufacturing and assembly of all components, it is practically always the case that the basic imaging system does not yet meet the specifications required for later use. Therefore, the basic imaging system is adjusted in an adjustment process to obtain an adjusted useful configuration in which the optical imaging system exhibits the specified image quality. The imaging system is adjusted based on at least one system measurement of the image quality performed with a measuring system by controlling manipulators to change the optical effect of associated mirrors so that the imaging system is "in specification" after the adjustment is completed. In the methods considered here, the system measurement includes a measurement of the wavefront of the wave propagating along the imaging beam path. In this adjustment process, manipulators, which are already present in optical modules in the optical imaging systems considered here, are used for the purpose of adjustment.

Such an adjustment process can be very time-consuming and take several days, even if well-trained, experienced personnel carry out the adjustment.

TASK AND SOLUTION

Against this background, the present invention is based on the object of providing a method for producing an optical imaging system for a microlithography system in which the time required for the adjustment process is considerably reduced compared to conventional procedures.

To achieve this object, the invention provides a method having the features of claim 1. Advantageous further developments are specified in the dependent claims. The wording of all claims is incorporated into the description by reference.

In a method according to the claimed invention, the adjustment process automatically runs through a first stage in the form of a coarse adjustment and a second stage in the form of a fine adjustment. The term "automatic" here means that no operator intervention in the process is required for this process. During the adjustment process, the measuring system is operated or used in the first stage in a first measuring mode with a first measuring accuracy, and in the second stage in a second measuring mode with a second measuring accuracy that is higher than the first measuring accuracy. Due to the fact that only a comparatively low measuring accuracy is required in the first stage, a complete system measurement in the first measuring mode, i.e., during the first stage, requires only a relatively short measuring time. Performing a complete measurement in the second measuring mode, on the other hand, produces significantly more accurate measurement results, but requires considerably more time for a complete run than a measurement in the first measuring mode. The innovative automated adjustment process automatically ensures optimal use of the capabilities of the first measuring mode and the second measuring mode and ensures, among other things, that measurements in the second measuring mode and adjustment interventions based on them only begin when a relatively good adjustment state has already been achieved by the preceding first stage (rough adjustment).

According to the claimed invention, the coarse adjustment is terminated and the fine adjustment is automatically initiated when a predeterminable first adjustment state is reached, which is defined by at least one predeterminable termination criterion. A first termination criterion is met when a system measurement shows that an image quality error lies within a predeterminable first tolerance range or is smaller than a predeterminable first error limit. However, the fine adjustment is also automatically initiated if the image quality error is still greater than or equal to the first error limit when approaching the first error limit, but at the same time a predeterminable improvement potential limit of the coarse adjustment is reached.

The fine adjustment is completed when the image quality error is less than one is a predeterminable second error limit value, i.e. lies within a second tolerance range that is narrower than the first tolerance range.

During this adjustment process, it is possible to precisely specify the adjustment state at which an automatic transition between coarse adjustment and fine adjustment should occur. By selecting suitable values for the termination criteria, an optimal compromise can be achieved between the time required and the degree of goal achievement of the coarse adjustment and the time required and the degree of goal achievement of the fine adjustment.

The first termination criterion generally applies in cases where the coarse adjustment converges relatively quickly and leads within a reasonable time to an adjustment state in which the image quality error lies within the first tolerance range, so that the subsequent fine adjustment can take over and further improve the adjustment state in a reasonable time such that the image quality error lies within the specification, i.e. within the second tolerance range around the target image quality.

The second termination criterion covers those cases in which the coarse adjustment, after one or more iterations, has already led to an adjustment state that is close to the first error threshold, but the adjustment state is not yet good enough, so that the first termination criterion is not yet met. The second termination criterion "takes effect" when it is determined that further coarse adjustment, with further iterations, will only result in marginal improvements in image quality, which would only lead to the conditions for the first termination criterion being met after a long adjustment period. In this case, it may be more advantageous to automatically proceed to the fine adjustment in order to minimize the time required for the overall adjustment.

According to the second termination criterion, the system switches to fine adjustment even if the adjustment state quantified by the first termination criterion has not yet been fully reached, because it is more time-efficient if, in this adjustment state close to the first adjustment state, the fine adjustment leads to the desired target state more quickly than a coarse adjustment that converges too slowly would require to reach the first adjustment state.

By specifying suitable values for the first termination criterion and the second termination criterion, the total time of the adjustment process until the desired image quality is achieved can be optimized.

According to a further development, the adjustment process in the first stage comprises a sequence of several steps. The starting point is a situation in which the actuators of the manipulation system are in a specific configuration that determines the optical effect of the respectively assigned optical elements. According to step A), a system measurement is carried out in this situation in the first measurement mode in order to determine the image quality for the actuator configuration given during the system measurement. According to step B), an evaluation of the image quality determined in step A) includes a check to determine whether the image quality error lies within or outside the first tolerance range. If the image quality error is already within the first tolerance range (i.e., below the first error limit), the system automatically switches to fine adjustment.

If the image quality error lies outside the first tolerance range, a manipulator recipe is calculated in step C). This recipe contains actuator travel commands for setting a modified actuator configuration. According to an underlying system model, the modified actuator configuration leads to an improvement in image quality. The term "manipulator recipe" refers to the entire set of instructions for the manipulator travel paths.

Subsequently, in step D), a convergence test is carried out to determine whether the modified actuator configuration is likely to lead to an improvement in image quality. Step D) thus serves as a control instance with regard to the expected effectiveness of the preceding step C). If the result of the convergence test is positive, i.e. if the convergence test shows that the change in the actuator configuration suggested by the manipulator recipe should actually lead to an improvement in image quality, the modified actuator configuration is adjusted in step E) by controlling suitable actuators according to the manipulator recipe. After this adjustment step, it would then be expected that the image quality has improved or the image quality error has decreased. The preceding steps A) to E) are repeated in this way as long as the image quality error lies outside the first tolerance range and the convergence test is positive.

According to a further training, an important contribution to accelerating the adjustment process is made by performing the convergence test using a simulation calculation. The convergence test is therefore not carried out on the real imaging system by implementing the manipulator recipe and then measurement of the image quality is carried out, but virtually through simulations.

According to a further development, a further contribution to saving adjustment time is achieved by specifying an improvement threshold for the convergence test and automatically initiating the second stage (fine adjustment) if the convergence test shows that a modified actuator configuration determined in step C) would result in an improvement in image quality, but this improvement is less than the improvement threshold. This avoids time losses that could arise from running one or more iterations of the coarse adjustment without resulting in a substantial improvement in image quality that would lead to reaching the first tolerance limit or falling below the first error threshold.

Modern measurement systems for performing system measurements on imaging systems deliver a large amount of measurement data per measurement, the evaluation of which provides quantitative information about the current image quality at the time of the measurement. Converting the raw measurement data into a quantitative description of the image quality, for example to quantify the current wavefront, requires a considerable amount of time. It is theoretically possible that the measurement data itself could reveal that a measurement was disturbed or incorrect, so that no meaningful conclusions about the image quality can be drawn from the measurement data. The plausibility check can be used to determine this, so that an iteration of the rough adjustment can be aborted early if it becomes apparent from the measurement data that no meaningful results can be expected. A new measurement can then be initiated immediately.

When the conditions that mark the end of the coarse adjustment and the automatic transition to fine adjustment are met, the manipulator actuators assume certain configurations. According to a further development, a first system measurement in the second stage is performed with the same actuator configuration as a final system measurement of the first stage. In other words, the configuration of the manipulator actuators is retained during the transition from coarse adjustment to fine adjustment. Thus, the measurement results of the final system measurement of the coarse adjustment should agree, at least to the first order, with the measurement results of the first measurement of the fine adjustment with regard to the quantification of the adjustment status. This makes it possible to use the measured values determined in the second measurement mode, which requires considerable expenditure of time, to evaluate the quality of the measurement results in the preceding first measurement mode and to improve them if necessary.

Preferably, in the first measurement mode, only a single wavefront measurement is performed for each alignment state in order to determine the current image quality error. Such a system measurement requires only a relatively short measurement time. In the second measurement mode, however, several measurements are performed for each alignment state, during which the optical imaging system in the measuring machine is rotated and shifted differently in order to separate the contribution to the image errors originating from the lens component from that of the machine or sensor component. Such measurement strategies are used, for example, in WO 2019/'025218 A1 Such measurements are significantly more time-consuming.

According to a further development, the optical imaging system is designed for use in an EUV microlithography system. In this case, all optical modules each have an optical element in the form of a mirror. Such imaging systems often have fewer than ten mirrors, e.g., four, six, or eight mirrors. Preferably, at least half of all optical modules have actuators of manipulators of a wavefront manipulation system for controllably changing the optical effect of the mirrors in response to control signals from the control unit. Preferably, at least 80% are equipped with manipulable mirrors. Especially when so many manipulable optical elements are available, the entire adjustment can be performed using the method.

In DUV systems, an optical imaging system typically has ten or more optical elements, of which usually less than half are manipulable, for example, one, two, three, or four optical elements. Particularly in such cases, only a sub-process of the alignment may be performed according to the procedure to improve the alignment state. After completion of this sub-process, further measures may be necessary to correct imaging errors, e.g., material-removing processing of individual optical surfaces using an ion beam or electron beam or similar to create a corrective asphere.

BRIEF DESCRIPTION OF THE DRAWINGS

• 1 zeigt schematisch Komponenten eines in eine Messmaschine zur Wellenfrontmessung eingebauten Projektionsobjektivs einer EUV-Mikrolithographie-Projektionsbelichtungsanlage während eines Justageprozesses;
• 2 zeigt ein schematisches Flussdiagramm zur Erläuterung eines automatisch ablaufenden Justageprozesses mit Grobjustage und Feinjustage.

Further advantages and aspects of the invention emerge from the claims and from the description of embodiments of the invention, which are explained below with reference to the figures.

• 1 schematically shows components of a projection lens of an EUV microlithography projection exposure system built into a wavefront measurement measuring machine during an adjustment process;
• 2 shows a schematic flow chart to explain an automatic adjustment process with coarse adjustment and fine adjustment.

DETAILED DESCRIPTION OF THE EMBODIMENTS

The 1 schematically shows components of a projection objective PO of an EUV microlithography projection exposure system, which is used to expose a radiation-sensitive substrate arranged in the region of an image plane IS of a projection objective PO with at least one image of a pattern of a reflective mask arranged in the region of an object plane OS of the projection objective. The projection objective is configured to operate with radiation from the extreme ultraviolet range (EUV range), in particular with wavelengths between 5 nm and 15 nm. So that the projection objective can operate in this wavelength range, it is constructed with optical elements in the form of mirrors that are reflective for EUV radiation. All mirrors are coated with multilayer reflective coatings that reflect EUV radiation and can contain, for example, Mo/Si layer pairs (bilayer). The projection objective PO is an example of an optical imaging system for EUV lithography.

In the example, the projection lens PO has six mirrors M1 to M6 with concave or convex curved mirror surfaces. These can be freeform surfaces. In production operation, the projection lens projects a mask pattern arranged in the object plane of the projection lens at a reduced scale into the image plane, which then contains a substrate to be exposed, e.g., a semiconductor wafer. The image plane contains the image field optically conjugated to the object field. Other designs, e.g., with more or fewer mirrors and/or with an intermediate image, are possible.

The projection lens has a modular design. Optical modules OM1, OM2, ..., OM6 are installed at assigned mounting positions in a common support frame (force frame) FF. Each of these modules contains one of the mirrors M1 to M6. When fully assembled, these modules are located at the mounting positions designated for the optical modules. Together with the optically effective surfaces of the mirrors M1 to M6, they form the imaging beam path.

With the projection lens, it is possible to at least partially compensate for imaging errors by using manipulators. The term "manipulator" refers to a device configured to actively influence an associated mirror based on control signals from a control system ST in order to change its optical effect. A manipulator comprises one or more actuators integrated into an optics module, whose current control value can be changed or adjusted based on control signals from the operating control system. Several of the optics modules comprise such manipulators MAN1, ..., - MAN5.

The 1 shows the projection lens PO during a phase of initial production in a state in which the projection lens is installed in a measuring machine MMA. This machine has a wavefront measuring system WMS that operates at the EUV operating wavelength and is configured to measure the wavefront of the projection radiation propagating in the projection lens from the object plane to the image plane. For wavefront measurement using the wavefront sensor, a coherence reticle KR is arranged in the object plane OS. In the exemplary embodiment, the wavefront sensor comprises a diffraction grating LAT arranged in the image plane IS and a spatially resolving intensity sensor SEN arranged below the diffraction grating.

In the example, a spatially resolved measurement is planned for several field points. The measuring machine is designed so that the projection lens can be rotated differently between successive measurements in order to separate those contributions to the image errors originating from the projection lens from those of the machine or sensor component (see WO 2019/025218 A1 ).

During initial production, a basic imaging system is first constructed by installing the optics modules in the corresponding installation positions on the support frame. The imaging quality of this system typically deviates significantly from the imaging quality required during operation. To achieve a calibrated configuration in which the projection lens exhibits the imaging quality specified for its intended use, an adjustment process is then performed. System measurements are repeatedly performed during the adjustment process. Depending on the measurement results, individual or multiple mirrors are then shifted in their rigid-body degrees of freedom by controlling the manipulators to improve the imaging quality.

The calculation of the manipulator changes to be performed to correct the image quality is carried out using a travel-generating optimization algorithm, which is also referred to as a “manipulator change model” or “manipulator recipe.” Such optimization algorithms are used, for example, in WO 2010/034674 A1 as well as DE 10 2012 205 096 B3 and DE 10 2015 206 448 B4 described.

Using the schematic flow chart in 2 An example of an adjustment process will now be explained which can run completely automatically without operator intervention and, starting from an initial state which has not yet been adjusted, leads relatively quickly and purposefully to a final state of the projection lens in which its image quality corresponds to the specifications required for productive operation.

The abbreviation ST stands for the start or beginning of the process of interest here. After the optical modules have been installed in their positions, the projection lens is still in its original state, with wavefront errors still significantly outside the specifications. This basic imaging system is now automatically adjusted in a multi-step alignment process.

The phase presented here begins with an initial system measurement, in which the wavefront measurement system is operated in a first measurement mode, MEAS1. In the example, this means that only a single spatially resolved wavefront measurement is performed on the projection lens, and the measurement results are then evaluated.

The system measurement delivers a large amount of measurement data, which is then checked in the subsequent DEST (plausibility check) step to determine whether a meaningful measurement result can be expected or whether, for example, due to a fault in the measurement data acquisition, the measurement is not expected to deliver meaningful measurement data. If the plausibility check reveals that the measurement data as such is not suitable for further processing, the corresponding information (INF) is transmitted back to the measurement system, and the measurement data is not further processed to calculate a corresponding image quality.

If, however, the measured data appears plausible, the wavefront evaluation (WFB) step follows. This step assesses whether the image quality achieved by the adjustment step and quantified by the measurement meets certain termination criteria, allowing the system to automatically advance to the second stage (fine adjustment). If the wavefront evaluation (WFB) shows that the measured image quality error is already smaller than a specified first error limit, the system automatically advances to the second stage (fine adjustment). The first error limit quantifies the boundary of a first tolerance range around the ideal target image quality at which, according to the specifications, it is already worthwhile to proceed to fine adjustment.

If, however, the wavefront evaluation shows that the image quality error is still greater than the first error threshold, the adjustment process assumes that a further coarse adjustment step will result in a significant improvement in image quality within a short period of time. To achieve this, the process proceeds to the next process step, REZ-B, in which a manipulator recipe is calculated. This recipe contains instructions for the manipulator travel paths to be adjusted, which, according to an underlying optimization model, should lead to a significant improvement in image quality.

However, this calculated manipulator recipe is not yet implemented immediately. Instead, a convergence test (KONV) follows, which determines whether the modified actuator configuration resulting from the manipulator recipe is actually likely to lead to an improvement in image quality. The convergence test is performed using a numerical simulation, which essentially calculates the image quality that should result, according to an underlying model, if the manipulator recipe were implemented on the projection lens by adjusting the relevant manipulators.

If the convergence test reveals that implementing the manipulator recipe by changing the travel ranges of the manipulator actuators should lead to a significant improvement in image quality, which would, for example, subsequently result in a switch to fine adjustment, the manipulation concept calculated in step REZ-S is stored in a database and implemented in the subsequent step JU by actuating the actuators accordingly and resulting in a change in the position of the manipulators. In the best case scenario, a further measurement MESS1 in the first measurement mode would then reveal an image quality error, which would lead to an automatic switch to the second stage of the adjustment process.

However, it cannot be ruled out that several iterations of the coarse adjustment would be necessary in this way to improve the image quality to such an extent that a switch to the fine adjustment would be necessary after the wavefront evaluation WFB. In other words: The adjustment concept also takes into account cases in which changing the manipulator positions leads to a deterioration in the image quality or only to a slight improvement, so that further iterations of the coarse adjustment would be necessary before switching to fine adjustment can be carried out. Therefore, an improvement threshold is specified for the convergence test, and the second stage (fine adjustment) is automatically initiated if the convergence test shows that the previously determined modified actuator configuration would only lead to an improvement in the image quality that is less than the improvement threshold. The consideration behind this aspect of the adjustment process is that the total adjustment time until the desired image quality is achieved can be very long if the image quality converges only slowly during the coarse adjustment, whereas switching to fine adjustment could lead to the desired result more quickly.

The fully automatic adjustment process thus uses two different termination criteria to define under which circumstances the system should automatically switch from coarse adjustment to fine adjustment. The first termination criterion takes effect during the wavefront evaluation and results in an automatic switch to fine adjustment if the image quality error determined by measurement is smaller than a predefined first error limit. If this is not the case, this means that the coarse adjustment target has not been achieved, and a next coarse adjustment loop is initiated (with calculation of a manipulator recipe, convergence check, recipe storage if necessary, and implementation of the recipe before a new measurement in the first measurement mode).

However, to also cover cases where the coarse adjustment converges too slowly, a second termination criterion is used during the convergence check. If this check reveals that the result of the simulation performed during the convergence check is only slightly better than the result of the last measurement, this is interpreted as a sign that the specified improvement potential limit of the coarse adjustment has been reached. The mechanism can then automatically switch to fine adjustment, even if the image quality at that time would not be good enough to trigger a switch to fine adjustment after the first termination criterion. Switching due to reaching the improvement potential limit can therefore also occur if the first error limit has not yet been fully reached. If, on the other hand, the convergence check reveals that the new simulation promises a significant improvement compared to the last measurement, the previously calculated manipulator recipe is implemented, and the success of this measure is verified in a further measurement in the first measuring mode.

Regardless of the method used to switch from the first stage to the second stage, the adjustment process is defined such that the first system measurement in the second stage (fine adjustment) is performed using the same actuator configuration as the final system measurement in the first stage. This allows the two measurement methods to be compared for consistency in their results, thereby increasing the reliability of the adjustment process.

It has proven useful to proceed in such a way that the first error limit and/or the improvement limit are defined by one or more quantities determined either statistically or via AI/machine learning and/or acceptance-relevant and/or population-relevant, which take individual aberration components into account and/or combine a large number of aberration components into a quality criterion representing the image quality.

The entire adjustment process can thus be divided into a measurement component, a calculation component, and a component that physically implements these recipes. In one embodiment, separate scripts are provided for the two subprocesses of the adjustment process, which communicate with each other. One of the scripts regularly searches for a new manipulator recipe, implements it, and starts a new measurement. The other script regularly searches for a new measurement, then evaluates it, and, if necessary, creates a new manipulator recipe.

Because the scripts regularly check for changes in the size of each other, the various steps of the automated adjustment process run automatically with virtually no delay, eliminating unnecessary waiting times between individual steps. This also helps reduce the time required for the entire adjustment process as much as practical and reasonable.

This example demonstrates that the entire rough adjustment process can be fully automated. The adjustment process includes an automatic evaluation of the measurement with regard to convergence and target value achievement, including newly defined quality criteria. Manipulator recipes are calculated using suitable optimization methods. There are a variety of optimization algorithms from which a specialist can choose. In other words, the adjustment process and its advantages are independent of the type of optimization algorithm underlying the recipe calculation. The physical implementation of the calculated manipulator recipes and the initiation of the next measurement are each carried out via interaction with a database.

The fine adjustment procedure can generally be the same as for coarse adjustment. Different values can be used for the termination criteria. Since the system measurements in the second measurement mode provide more detailed information about the projection lens' contribution to the overall measured values, more precise manipulator recipes can be calculated, so that the adjustment may require only a single change in the manipulator configuration to achieve the adjustment goal.

In particular, in some embodiments, it is provided that the fine adjustment runs automatically, wherein the fine adjustment is terminated when a predefinable first fine adjustment state is reached, which is defined by at least one predefinable termination criterion of the fine adjustment. A first termination criterion of the fine adjustment is met when the system measurement shows that an image quality error is smaller than a predefinable first error limit of the fine adjustment, and a second termination criterion of the fine adjustment is met when the image quality error, when approaching the first error limit of the fine adjustment, is still greater than or equal to the first error limit, but at the same time a predefinable improvement potential limit of the fine adjustment is reached.

The method can also be used in principle with some imaging systems designed for other wavelength ranges, e.g., deep ultraviolet (DUV) light and/or other radiation sources (e.g., electron beam). Especially with EUV systems with many manipulable mirrors, it is possible to automate the entire alignment process as described. If necessary, only a sub-process of the overall alignment process can be automated as described.

QUOTES CONTAINED IN THE DESCRIPTION

This list of documents submitted by the applicant was generated automatically and is included solely for the convenience of the reader. This list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Cited patent literature

• WO 2019/'025218 A1 [0028, 0037]
• WO 2010/034674 A1 [0039]
• DE 10 2012 205 096 B3 [0039]
• DE 10 2015 206 448 B4 [0039]

Claims (9)

A method for producing an optical imaging system for a microlithography system, wherein the optical imaging system has a plurality of optical modules, each carrying an optical element, along an imaging beam path leading from an object plane to an image plane of the imaging system, and a plurality of the optical modules have actuators of manipulators of a wavefront manipulation system for controllably changing the optical effect of the optical elements in response to control signals from a control unit. Whereby, first, a basic imaging system is constructed by installing the optical modules at associated installation positions. Thereafter, the basic imaging system is adjusted in an adjustment process based on at least one system measurement of an imaging quality performed with a wavefront measuring system by controlling manipulators for changing the optical effect of associated optical elements in order to obtain an (adjusted) useful configuration in which the optical imaging system has a specified imaging quality. Whereby the adjustment process automatically runs through a first stage in the form of a coarse adjustment and a second stage in the form of a fine adjustment, wherein the measuring system is in a in the first measurement mode with a first measurement accuracy and, in the second stage, in a second measurement mode with a second measurement accuracy that is higher than the first measurement accuracy, wherein the coarse adjustment is terminated and the fine adjustment is automatically initiated when a predeterminable first adjustment state is reached, which is defined by at least one predeterminable termination criterion, wherein a first termination criterion is met when the system measurement shows that an image quality error is smaller than a predeterminable first error limit, and a second termination criterion is met when the image quality error is still greater than or equal to the first error limit when approaching it, but at the same time a predeterminable improvement potential limit of the coarse adjustment is reached. Procedure according to Claim 1 , characterized in that the adjustment process in the first stage comprises the following steps: A) Carrying out a system measurement in the first measurement mode to determine the image quality for an actuator configuration of actuators of the wavefront manipulation system given during the system measurement; B) Assessing the image quality, including a check to determine whether the image quality error lies within or outside the first tolerance range; C) If the image quality error lies outside the first tolerance range: Calculating a manipulator recipe with travel commands for actuators to set a modified actuator configuration which, according to a model, leads to an improvement in the image quality; D) Carrying out a convergence test to determine whether the modified actuator configuration is likely to lead to an improvement in the image quality; E) Setting the modified actuator configuration according to the manipulator recipe by controlling the actuators if the result of the convergence test is positive; F) Repeat steps A) to E) as long as the image quality error is outside the first tolerance range and the convergence test is positive. Procedure according to Claim 2 , characterized in that the convergence test is carried out using a simulation calculation. Procedure according to Claim 2 or 3 , characterized in that an improvement limit is specified for the convergence test in step D) and that the second stage is automatically initiated if the convergence test shows that a modified actuator configuration determined in step C) brings about an improvement in the imaging quality which is less than the improvement limit. Procedure according to Claim 2 , 3 or 4 , characterized in that step A) comprises a plausibility check of measurement data determined in the first measurement mode, that a determination of the image quality from the measurement data only takes place if the plausibility check does not indicate any measurement system errors and that if there is an indication of measurement system errors, an error message is transmitted to the measurement system. Method according to one of the preceding claims, characterized in that a first system measurement in the second stage is carried out with the same actuator configuration as a last system measurement in the first stage. Method according to one of the preceding claims, characterized in that the first error limit value and/or the improvement limit value are defined by one or more quantities determined either statistically or via AI/machine learning and/or acceptance-relevant and/or population-relevant, which take individual aberration components into account and/or combine a plurality of aberration components into a quality criterion representing the image quality. Method according to one of the preceding claims, characterized in that the fine adjustment runs automatically, the fine adjustment being ended when a predefinable first fine adjustment state is reached, which is defined via at least one predefinable termination criterion of the fine adjustment, a first termination criterion of the fine adjustment being met when the system measurement shows that an image quality error is smaller than a predefinable first error limit value of the fine adjustment and a second termination criterion of the fine adjustment is met when the image quality error is still greater than or corresponds to the first error limit value of the fine adjustment when approaching it, but at the same time a predefinable improvement potential limit of the fine adjustment is reached. Method according to one of the preceding claims, characterized in that the optical imaging system is designed for use in an EUV microlithography system, wherein all optical modules each have an optical element in the form of a mirror and at least half of all optical modules, preferably at least 80% of the optical modules, have actuators of manipulators of a wavefront manipulation system for controllably changing the optical effect of the mirrors in response to control signals from the control unit.