Dynamic photoetching alignment compensation method, photomask and photoetching system

Abstract

The invention relates to a dynamic photoetching alignment compensation method, a photomask and a photoetching system.The method comprises the steps that test photomasks with different geometric features and corresponding GDS files are provided, and geometric and layout features of exposure areas on the corresponding test photomasks are extracted from the GDS files; performing exposure by using the test photomask to obtain corresponding alignment error data; measuring temperature distribution data of different areas of each test photomask after exposure; based on the geometric and layout characteristics, the alignment error data and the temperature distribution data, constructing and training to obtain an identification optimization model; and aiming at a target product photomask, inputting geometric and layout characteristics of an exposure area of the target product photomask to the identifier optimization model, generating layout parameters of the reference identifier, and feeding back the layout parameters to a photomask production system. According to the invention, the reference identification layout can be dynamically adjusted according to the characteristic geometry and layout characteristics of the actual exposure area, the exposure compensation is effectively realized, and the alignment error is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of semiconductor manufacturing technology, and in particular to a dynamic lithography alignment compensation method, a photomask and a lithography system. BACKGROUND

[0002] In the lithography process of semiconductor manufacturing, overlay accuracy is one of the key factors determining device performance and yield. The lithography alignment system uses fiducial marks as key reference points to measure the relative positional deviation between the mask (also known as the photomask) and the wafer, and can further be used to detect and compensate for the aberration and distortion of the projection optical system, thereby achieving accurate positional alignment.

[0003] Currently, the alignment mark layout method commonly used in the industry is a fixed design, i.e., the number and position of the fiducial marks are predetermined during the photomask design phase, and are usually arranged in a symmetrical manner in the non-exposure area (such as the scribe lane) outside the exposure area. On a common 6-inch photomask, the maximum exposure area range is generally 26mm x 33mm. For example, as shown in Figures 1-3 , Figures 1-3 is a schematic diagram of the placement of fiducial marks on photomasks of different layouts in the prior art. There are significant differences in the common Chip Array form, wherein reference numeral 100 represents the photomask, reference numeral 200 represents the exposure area (Shot), and reference numeral 300 represents the fixedly arranged fiducial marks. Among them, Figure 1 is a 2x4 array, and the corresponding exposure area 200 (Shot) size is about 25.2um x 32.2um; Figure 2 is a 7x1 long strip array, and the exposure area 200 size is about 24.6um x 32.9um; Figure 3 is a 1x1 independent chip, and the exposure area 200 size is about 22.5um x 25.5um.

[0004] However, with the continuous miniaturization of technology nodes and the increasing diversification of chip product types, this fixed layout method has been difficult to meet the high-precision process requirements, and its technical limitations mainly lie in the following two aspects:

[0005] On the one hand, the exposure areas of different process platforms and customer designs differ significantly in area, shape, and chip array arrangement. Fixed-position, fixed-number fiducial marks are difficult to adapt to all layout features. For example, Figure 1 for a Figure 1 such exposure area, which is uniformly distributed and regularly arrayed, the fixed marks can still achieve basic compensation for thermal expansion deformation; but for a Figure 2 long strip array or Figure 3The small size of the Shot and the fact that the fiducial mark cannot effectively cover the local deformation area result in insufficient compensation of the alignment error at some positions in the exposure area.

[0006] On the other hand, when the chip size is small or the exposure area shape is special, the fixed fiducial mark is often far away from the actual pattern area. During the exposure process, the reticle generates a thermal effect (reticle heating) due to the absorption of laser energy, causing non-uniform thermal expansion. At this time, the mark located at the edge cannot accurately reflect the actual deformation of the interior of the exposure area (especially the central area), resulting in deviation of the global alignment compensation based on the measurement data thereof, and a decrease in overlay accuracy. For example, Figure 2 the central area of the long strip structure, Figure 3 the isolated Shot center, all have insufficient deformation monitoring accuracy due to the excessive distance of the mark.

[0007] Therefore, the existing fixed fiducial mark layout method cannot meet the development needs of diversified and high-precision photolithography processes. It is necessary to propose a method for dynamically optimizing the layout of fiducial marks according to the actual exposure area characteristics, so as to improve the alignment accuracy and enhance the flexibility of mask design. SUMMARY

[0008] The purpose of the present application is to provide a dynamic photolithography alignment compensation method, a mask, and a photolithography system, which can generate customized fiducial mark layouts according to the characteristic geometry and layout characteristics of the actual exposure area, accurately compensate for local deformation (especially thermal expansion), and significantly improve the photolithography alignment accuracy and process adaptability.

[0009] To achieve the above purpose, the present application provides a dynamic photolithography alignment compensation method, comprising:

[0010] providing test masks with different geometric characteristics and corresponding GDS files, and extracting the geometry and layout characteristics of the exposure area on the corresponding test mask from the GDS file;

[0011] performing exposure using the test mask to obtain corresponding alignment error data;

[0012] measuring the temperature distribution data of different areas of each test mask after exposure;

[0013] based on the geometry and layout characteristics, the alignment error data, and the temperature distribution data, constructing and training an identification optimization model;

[0014] for a target product mask, inputting the geometry and layout characteristics of the exposure area thereof into the identification optimization model, generating the layout parameters of the fiducial mark, and feeding back to the mask production system.

[0015] Optionally, the geometry and layout characteristics include one or more of the following: an area of the exposure region, a shape of the exposure region, a distribution of key patterns, and a chip array arrangement.

[0016] Optionally, the exposure region areas of the test masks with different geometry characteristics are different in size and / or the chip array arrangements are different.

[0017] Optionally, all the test masks are exposed on the same photolithography machine, and the exposure parameters and the metrology machine are kept consistent.

[0018] Optionally, a genetic algorithm or a reinforcement learning algorithm is used to construct the identification optimization model.

[0019] Optionally, when the identification optimization model is constructed based on the reinforcement learning algorithm, a reward function and a penalty function are introduced; the reward function is set based on a reduction ratio of the alignment error, and the penalty function is used to penalize a situation where the number of reference marks exceeds a first threshold or the interval is less than a second threshold.

[0020] Optionally, the layout rule followed by the identification optimization model is that the density of reference marks in a high-temperature gradient zone is high, and the density of reference marks in a low-temperature gradient zone is low.

[0021] Optionally, the layout parameters include an optimal number of reference marks, position coordinates of each reference mark, and weight distribution of each reference mark in global alignment compensation; wherein,

[0022] The optimal number of reference marks is a variable dynamically determined according to the exposure region area and the pattern complexity of the target product mask;

[0023] The position coordinates of the reference marks are two-dimensional coordinates based on the coordinate system of the exposure region;

[0024] The weight distribution of the reference marks is a normalized value used to represent the contribution of each reference mark in global compensation.

[0025] Based on this, the present application further provides a mask, the layout of the reference marks on which is generated and configured according to the dynamic photolithography alignment compensation method as described above.

[0026] Based on this, the present application further provides a photolithography system, comprising:

[0027] A photolithography machine configured to perform an exposure operation;

[0028] A control unit configured to perform the dynamic photolithography alignment compensation method as described above.

[0029] In the dynamic photolithography alignment compensation method, mask, and photolithography system provided by the present application, at least one of the following beneficial effects is achieved:

[0030] 1) By constructing the identification optimization model, customized reference mark layout can be generated for exposure areas with different geometric and layout characteristics. This layout strategy can accurately capture and compensate for complex non-uniform thermal deformation caused by mask heating effects, effectively reducing local alignment errors and achieving higher precision global overlay alignment.

[0031] 2) The traditional fixed mark layout cannot adapt to diversified chip designs. The identification optimization model of the present application can automatically calculate the optimal number and position of reference marks according to the exposure area, shape, pattern distribution and chip array arrangement of the specific product mask. This makes the scheme perfectly adapt to various designs from regular array to long strip, large size chip, etc., solving the adaptability problem caused by the "one-size-fits-all" layout method, greatly improving the flexibility of mask design;

[0032] 3) The model introduces a penalty function in the optimization process to constrain the situation of too many marks or too small spacing. This ensures that the layout scheme meets the accuracy requirements while avoiding redundant configuration of mark resources, saving valuable scribe lane space on the mask, reducing manufacturing costs, and achieving the best balance between accuracy and cost;

[0033] 4) By combining geometric characteristics, measured alignment errors and temperature distribution data for model training, and feeding back the optimization results directly to the mask production system, a fully automated and intelligent closed loop from data collection, model optimization to physical manufacturing is formed. This reduces the dependence on human experience, improves the stability and efficiency of the process, and provides a reliable technical path for precise control of advanced processes;

[0034] 5) As a software algorithm level optimization, this compensation method does not require major modifications to existing expensive hardware equipment such as photolithography machines, making it easy to quickly deploy and promote on existing production lines, with high practical value and economic benefits. BRIEF DESCRIPTION OF DRAWINGS

[0035] Those of ordinary skill in the art will understand that the provided drawings are for a better understanding of the present application and do not constitute any limitation on the scope of the present application. Among them:

[0036] Figures 1-3 A schematic diagram of the placement of reference marks on masks for different layout designs in the prior art;

[0037] Figure 4 A flowchart of the dynamic photolithography alignment compensation method provided by an embodiment of the present application. DETAILED DESCRIPTION

[0038] In order to make the objects, technical solutions and advantages of the present application clearer, the following will clearly and completely describe the technical solutions in the present application in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. The components of the embodiments of the present application described and shown in the accompanying drawings can be arranged and designed in various different configurations.

[0039] Therefore, the following detailed description of the embodiments of the present application provided in the accompanying drawings is not intended to limit the scope of the claimed present application, but only represents selected embodiments of the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application without creative labor are within the scope of protection of the present application.

[0040] In the description of the present application, it should be understood that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, or the orientation or positional relationship commonly understood by those skilled in the art, and are only for the purpose of facilitating the description of the present application and simplifying the description, and therefore cannot be understood as indicating or implying that the device or element referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as limiting the present application.

[0041] In addition, the relational terms such as "first" and "second" and the like are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply that there is any such actual relationship or order between these entities or operations. Moreover, the terms "include", "contain" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not explicitly listed or inherent to such article or device. Without more limitations, the element defined by the phrase "including a" does not exclude the presence of additional identical elements in the article or device including the element. The specific meaning of the above terms in the present application can be understood on a case-by-case basis by those of ordinary skill in the art.

[0042] Please refer to Figure 4 The embodiment of the present application provides a dynamic lithography alignment compensation method, comprising:

[0043] S1, providing test masks with different geometric features and corresponding GDS files, and extracting geometric and layout features of exposure regions on the corresponding test masks from the GDS files;

[0044] S2, exposure is performed using the test mask to obtain corresponding alignment error data;

[0045] S3, temperature distribution data of different regions of each test mask after exposure is measured;

[0046] S4, based on the geometric and layout features, the alignment error data and the temperature distribution data, an identification optimization model is constructed and trained;

[0047] S5, for a target product mask, the geometric and layout features of the exposure area thereof are input into the identification optimization model to generate layout parameters of the reference mark and feedback to the mask production system.

[0048] Firstly, S1 is performed, test masks with different geometric features and corresponding GDS files are provided, and the geometric and layout features of the exposure area on the test mask are extracted from the GDS file.

[0049] In this embodiment, the geometric and layout features include one or more of the area, shape, key pattern distribution and chip array arrangement of the exposure area.

[0050] The area of the exposure area directly determines the range of chips that can be covered by a single exposure;

[0051] The shape of the exposure area determines the thermal expansion behavior and stress distribution of the mask during exposure. A regular rectangular area has relatively uniform and predictable thermal deformation. A long and narrow strip-shaped area with a large aspect ratio (for example, a 7x1 array) is more likely to produce asymmetric and anisotropic deformation when heated.

[0052] The key pattern distribution refers to the local density distribution of circuit patterns in the exposure area. During lithographic exposure, there are differences in light reflection, scattering and heat absorption between pattern-dense areas and pattern-sparse areas, and this unevenness will directly cause local overlay errors.

[0053] The chip array arrangement refers to the arrangement and combination of multiple chip units (Die) in a single exposure field, such as a regular 2x4 grid, a 7x1 long strip array, or a 1x1 single large chip.

[0054] By extracting and analyzing these feature data, process engineers can predict the possible deformation of different design layouts during lithography, thereby optimizing process parameters in advance, dynamically adjusting alignment strategies, and ultimately achieving precise control of overlay errors.

[0055] In this embodiment, the exposure area size and / or chip array arrangement of the test masks with different geometric features are different.

[0056] It's important to note that the test masks are not chosen arbitrarily, but rather carefully designed with different geometric features to cover as many process scenarios as possible. For example, masks with very small exposure areas but long chip arrays are intentionally prepared, as well as masks with large exposure areas but containing only a single chip. The purpose of this is to enable the subsequently trained label optimization model to acquire various possible scenarios. Therefore, when faced with a new product mask, even if its layout is unprecedented, the label optimization model can make reasonable inferences based on learned patterns and generate the optimal label layout scheme.

[0057] Next, step S2 is executed, where the test photomask is exposed to obtain the corresponding alignment error data. This step is the core data foundation for building the identifier optimization model. Its goal is to simulate real process conditions and accurately measure the alignment error generated by photomasks with different geometric features after exposure, providing high-quality and clean input data for subsequent model training.

[0058] Preferably, all test masks are exposed on the same lithography machine, and the exposure parameters and measurement equipment are kept consistent. Different lithography machines have inherent systematic errors (such as slight differences in stage accuracy and optical performance). If multiple lithography machines are used for testing, the collected error data will contain a mixture of errors inherent to the mask itself and systematic deviations from different equipment, making it impossible to attribute the results solely to changes in the mask's geometric characteristics.

[0059] Furthermore, fluctuations in exposure parameters (such as light intensity and focal length) directly affect image quality and alignment. Different measurement equipment may also lead to variations in measurement principles and accuracy. Strictly controlling these variables ensures high comparability of alignment error data obtained from different test masks, thereby accurately capturing alignment errors directly caused by variations in mask geometry.

[0060] Then, step S3 is executed to measure the temperature distribution data of different areas of each test mask after exposure. During the exposure process, the mask's temperature rises due to the continuous absorption of laser energy. This reticle heating is not uniform; the rates of heat absorption and dissipation differ between dense and sparse areas, and between the central and edge regions, resulting in complex temperature gradients. This non-uniform temperature field directly causes non-uniform thermal expansion of the mask, which is one of the main factors leading to overlay errors. Therefore, accurately measuring its temperature distribution is crucial to obtaining the thermal deformation field of the mask during the actual exposure process. This data is used to train the marker optimization model and provide data support for optimizing the layout of the baseline marker. In this embodiment, non-contact infrared thermometry (such as an infrared thermal imager or a high-precision infrared thermometer) or advanced optical thermometry can be used to measure the temperature distribution data of different areas of each test mask after exposure.

[0061] Next, S4 is executed to construct and train an identifier optimization model based on the geometric and layout features, the alignment error data, and the temperature distribution data.

[0062] In practice, model training typically follows a systematic process, which includes:

[0063] Data preparation and preprocessing: First, the extracted geometric and layout features of different test masks, the measured alignment error data, and the temperature distribution data are organized, normalized, and correlated to form a structured dataset.

[0064] Algorithm selection and initialization: Select the optimal algorithm based on specific requirements. For example, if the identifier layout scheme is encoded as a "chromosome," a genetic algorithm can be used for population initialization; if it is viewed as an interaction problem between an agent and its environment, a reinforcement learning framework can be established and an initial policy can be set.

[0065] Iterative Optimization and Evaluation: The model begins to run. In genetic algorithms, a new generation of layout schemes is generated through selection, crossover, and mutation, and their performance is evaluated using a fitness function (such as the degree of reduction in alignment error). In reinforcement learning, the agent tries different layout "actions" and updates its policy based on rewards (such as error reduction) and penalties (such as excessive labeling) from environmental feedback.

[0066] Convergence and Validation: Training continues until the model's performance on the validation set no longer shows a significant improvement or reaches the preset number of iterations. The resulting identifier optimization model is able to learn the complex mapping relationship from mask geometric features to the optimal identifier layout.

[0067] In summary, this method of constructing and training an identifier optimization model based on multi-source data essentially transforms an experience-dependent engineering problem into a data-driven intelligent prediction and decision-making process. By integrating key geometric and layout features, alignment error data, and temperature distribution data, and leveraging the powerful search and learning capabilities of optimization algorithms, it can dynamically generate a "tailor-made" high-performance reference identifier layout scheme for a new photomask, thereby directly improving the overlay accuracy of the lithography process.

[0068] Preferably, when constructing the identifier optimization model based on the reinforcement learning algorithm, a reward function and a penalty function are introduced; wherein, the reward function is set based on the reduction ratio of the alignment error, which intuitively reflects the effectiveness of the model layout scheme. The higher the ratio, the more significant the layout optimization effect, and the greater the "positive feedback" (reward value) obtained by the model.

[0069] The penalty function penalizes instances where the number of reference markers exceeds a first threshold or the spacing is less than a second threshold. Both the first and second thresholds can be set based on actual conditions. When the total number of reference markers in the proposed layout exceeds the first threshold, the penalty is triggered. This directly controls production costs, as each marker occupies photomask area and increases manufacturing complexity. Avoiding excessive markers is key to controlling costs and improving the practicality of the solution. Simultaneously, a minimum allowable distance between markers (i.e., the second threshold) needs to be set to prevent overly dense markers, which could physically cause signal interference. Sufficient spacing is the physical basis for ensuring that each marker can be clearly, independently, and accurately identified and measured by the alignment system.

[0070] In this embodiment, the reward function and the penalty function do not work independently, but together constitute a multi-objective optimization framework. By adjusting the weights of the reward and penalty, the model can be guided to find the optimal balance between accuracy and cost.

[0071] Preferably, the layout rule followed by the marker optimization model is: the density of the reference markers is high in the high-temperature gradient region and low in the low-temperature gradient region. The photomask absorbs laser energy and generates heat during exposure. Due to the non-uniform pattern distribution in the chip design, different areas on the photomask have different rates of heat absorption and dissipation, thus forming a temperature gradient. The high-temperature gradient region is often an area with dense patterns or severe heat accumulation, and its physical expansion after heating is more significant and complex. The core function of the reference markers is to measure the deformation of the photomask. To accurately compensate for this non-uniform, localized deformation, it is necessary to collect denser deformation data in areas of severe deformation, allocate more marker resources to the most needed high-temperature gradient region, and conserve resources in the low-temperature region where deformation is simple, thereby achieving optimal allocation of monitoring resources. This ensures the compensation accuracy in key areas and improves the efficiency and feasibility of the overall solution.

[0072] Finally, step S4 is executed. For the target product photomask, the geometric and layout features of its exposure area are input into the marking optimization model to generate the layout parameters of the baseline marking and feed them back to the photomask production system. The geometric and layout features of the target product photomask are input into the pre-trained marking optimization model. Since this model has learned the complex mapping relationship from photomask features to the optimal marking layout, it will eventually output a set of accurate and executable layout parameters. The generated layout parameters can be sent to the photomask production system through a standardized data interface (such as the SECS / GEM protocol or a specific format data file).

[0073] Preferably, the layout parameters include the optimal number of reference markers, the position coordinates of each reference marker, and the weight allocation of each reference marker in global alignment compensation; wherein,

[0074] The optimal number of reference markers is a variable dynamically determined based on the exposure area and graphic complexity of the photomask of the target product. The larger the area and the more complex the graphic, the more markers are required.

[0075] The position coordinates of the reference marker are two-dimensional coordinates based on the coordinate system of the exposed area. The origin and axis of the coordinate system are aligned with the boundary of the exposed area to ensure that all measurements and compensation calculations are performed in a unified coordinate system, which is the basis for achieving accurate positioning.

[0076] The weight allocation of the benchmark identifiers is used to characterize the contribution of each benchmark identifier to the global compensation, and is a normalized value of 0-1. It is usually based on the local deformation gradient or temperature gradient at the identifier's location. For example, identifiers in high-temperature gradient regions or areas of severe deformation have higher weights.

[0077] Based on the same inventive concept, embodiments of the present invention also provide a photomask, wherein the layout of reference markings on the photomask is generated and configured according to the dynamic lithography alignment compensation method described above.

[0078] Since the photomask provided by this invention belongs to the same inventive concept as the dynamic lithography alignment compensation method described above, the photomask provided by this invention has all the advantages of the dynamic lithography alignment compensation method described above. Therefore, the beneficial effects of the photomask provided by this invention will not be described in detail here.

[0079] Based on the same inventive concept, embodiments of the present invention also provide a photolithography system, comprising:

[0080] A lithography machine is configured to perform exposure operations;

[0081] The control unit is configured to perform the dynamic lithography alignment compensation method as described above.

[0082] Since the lithography system provided by this invention belongs to the same inventive concept as the dynamic lithography alignment compensation method described above, the lithography system provided by this invention has all the advantages of the dynamic lithography alignment compensation method described above. Therefore, the beneficial effects of the lithography system provided by this invention will not be described in detail here.

[0083] In summary, the dynamic lithography alignment compensation method, photomask, and lithography system provided by this invention analyze the geometric and layout characteristics of the exposure area and combine them with real-time process data (such as alignment error and temperature distribution) to dynamically optimize the layout of the reference markers, thereby achieving precise compensation for local deformation (especially thermal expansion) and significantly improving lithography alignment accuracy and process adaptability.

[0084] The above description is merely a description of preferred embodiments of the present invention and is not intended to limit the scope of the invention in any way. Any changes or modifications made by those skilled in the art based on the above disclosure are within the protection scope of the present invention. Obviously, those skilled in the art can make various modifications and variations to the present invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the present invention and its equivalents, the present invention also intends to include these modifications and variations.

Claims

1. A dynamic photolithography alignment compensation method, characterized in that, include: Provide test masks with different geometric features and corresponding GDS files, and extract the geometric and layout features of the exposure area on the corresponding test mask from the GDS file; Exposure is performed using the test mask to obtain the corresponding alignment error data; Measure the temperature distribution data of different areas of each test mask after exposure; Based on the geometric and layout features, the alignment error data, and the temperature distribution data, an identifier optimization model is constructed and trained. For the target product photomask, input the geometric and layout features of its exposure area into the mark optimization model, generate the layout parameters of the baseline mark, and feed them back to the photomask production system.

2. The dynamic photolithography alignment compensation method according to claim 1, characterized in that, The geometric and layout features include one or more of the following: the area, shape, key pattern distribution, and chip array arrangement of the exposure area.

3. The dynamic photolithography alignment compensation method according to claim 1, characterized in that, The exposure area size and / or chip array arrangement of test masks with different geometric features are different.

4. The dynamic photolithography alignment compensation method according to claim 1, characterized in that, All test masks were exposed on the same lithography machine, and the exposure parameters and measurement equipment were kept consistent.

5. The dynamic photolithography alignment compensation method according to claim 1, characterized in that, The identifier optimization model is constructed using a genetic algorithm or a reinforcement learning algorithm.

6. The dynamic photolithography alignment compensation method according to claim 5, characterized in that, When constructing the identifier optimization model based on the reinforcement learning algorithm, a reward function and a penalty function are introduced; wherein, the reward function is set based on the reduction ratio of the alignment error, and the penalty function is used to penalize the case where the number of the reference identifiers exceeds a first threshold or the spacing is less than a second threshold.

7. The dynamic photolithography alignment compensation method according to claim 1, characterized in that, The layout rule followed by the identification optimization model is: the density of the reference identification is high in the high temperature gradient region and low in the low temperature gradient region.

8. The dynamic photolithography alignment compensation method according to claim 1, characterized in that, The layout parameters include the optimal number of reference markers, the position coordinates of each reference marker, and the weight allocation of each reference marker in global alignment compensation; wherein, The optimal number of the reference markers is a variable dynamically determined based on the exposure area and graphic complexity of the photomask of the target product. The position coordinates of the reference marker are two-dimensional coordinates based on the coordinate system of the exposure area; The weight allocation of the benchmark identifier is a normalized value used to characterize the contribution of each benchmark identifier in the global compensation.

9. A photomask, characterized in that, The layout of the reference markings thereon is generated and configured according to the dynamic lithography alignment compensation method according to any one of claims 1-8.

10. A photolithography system, characterized in that, include: A lithography machine is configured to perform exposure operations; The control unit is configured to perform the dynamic lithography alignment compensation method according to any one of claims 1-8.

Images