Method, device and equipment for analyzing lithography process window based on stepwise regression

Abstract

The present application relates to a method, apparatus, and device for analyzing a lithography process window based on stepwise regression. The method comprises: obtaining photoresist pattern image data and determining the constraints required for the lithography process window; determining a lithography process model that satisfies the relationship between the photoresist pattern features, exposure energy, and focal length, and a stepwise regression direction for the lithography process model; performing stepwise regression on the lithography process model based on the photoresist pattern image data in the direction indicated by the stepwise regression direction to obtain multiple candidate lithography process models and model selection index values ​​for each candidate lithography process model; determining a target lithography process model from the multiple candidate lithography process models based on the model selection index values; and determining the lithography process window based on the constraints and the target lithography process model. This method can improve the accuracy of the analysis results of focal length energy matrix data.

Description

Technical Field

[0001] The present application relates to the field of photolithography technology, and in particular to a photolithography process window analysis method, device and equipment based on stepwise regression. Background Art

[0002] Lithography window process analysis is a critical step in the semiconductor manufacturing process, ensuring that the mask pattern is correctly replicated on the silicon wafer. Lithography window process analysis is crucial to ensuring the quality and performance of integrated circuits, while lithography technology itself is constantly evolving to meet the demands of smaller line widths and higher integration densities.

[0003] Focus-Exposure Matrix (FEM) analysis is a method used to determine the photolithography process window. FEM involves measuring the photoresist pattern after exposure and development using different exposure energies in one direction and focal lengths in another direction for different exposure areas on the same wafer. This allows researchers to determine the impact of different exposure energies and focal lengths on the critical dimensions (CD) of the photoresist pattern.

[0004] In traditional technology, the analysis method of the lithography process window uses regression analysis to fit the FEM data to a certain polynomial function model to obtain the functional relationship between the critical dimension of the photoresist pattern and the exposure energy and focal length. For FEM data, the analysis results of this processing method have low accuracy. Summary of the Invention

[0005] Based on this, it is necessary to provide a lithography process window analysis method, device, computer equipment, computer-readable storage medium and computer program product based on stepwise regression that can improve the accuracy of the analysis results of focal energy matrix data in order to address the above technical problems.

[0006] In a first aspect, the present application provides a lithography process window analysis method based on stepwise regression, comprising:

[0007] Obtaining photoresist pattern image data and determining the constraints required for the photolithography process window;

[0008] Determining a photolithography process model that satisfies the relationship between photoresist pattern characteristics and exposure energy and focal length, and a stepwise regression direction of the photolithography process model;

[0009] Performing stepwise regression on the photolithography process model according to the photoresist pattern image data in the direction indicated by the stepwise regression direction to obtain a plurality of candidate photolithography process models and a model selection index value of each candidate photolithography process model;

[0010] determining a target photolithography process model from a plurality of candidate photolithography process models according to the model selection index value;

[0011] A photolithography process window is determined according to the constraint conditions and the target photolithography process model.

[0012] In one embodiment, performing stepwise regression on the photolithography process model according to the photoresist pattern image data in the direction indicated by the stepwise regression direction to obtain a plurality of candidate photolithography process models and a model selection index value of each candidate photolithography process model includes:

[0013] Determining the candidate photolithography process models obtained each time in the stepwise regression process according to the photolithography process model in the direction indicated by the stepwise regression direction until a preset regression condition is met, thereby obtaining a plurality of candidate photolithography process models;

[0014] For each of the candidate photolithography process models, fitting the candidate photolithography process model according to the photoresist pattern image data to obtain a selection index value of each model item in each of the candidate photolithography process models;

[0015] According to the selection index value, a model selection index value of each candidate photolithography process model is determined.

[0016] In one embodiment, determining the candidate photolithography process models obtained each time in the stepwise regression process according to the photolithography process model in the direction indicated by the stepwise regression direction until a preset regression condition is satisfied, thereby obtaining multiple candidate photolithography process models, includes:

[0017] Determining a reference photolithography process model according to the photolithography process model in a first direction indicated by the stepwise regression direction;

[0018] Determine a model item to be added for the first regression from the photolithography process model, and determine a candidate photolithography process model for the first regression based on the reference photolithography process model and the model item to be added;

[0019] For each regression process after the second one, determining a model item to be added for each regression from the photolithography process model; wherein the model item to be added for each regression is different;

[0020] According to the model item to be newly added and the last candidate photolithography process model, the candidate photolithography process model for current regression is determined until a first preset regression end condition is met, thereby obtaining multiple candidate photolithography process models.

[0021] In one embodiment, determining the candidate photolithography process models obtained each time in the stepwise regression process according to the photolithography process model in the direction indicated by the stepwise regression direction until a preset regression condition is satisfied, thereby obtaining multiple candidate photolithography process models, includes:

[0022] According to the second direction indicated by the stepwise regression direction, for each regression process, obtaining the selection index value of each model item in the previous candidate photolithography process model;

[0023] Determining, from all the model items, the model items to be deleted that meet the preset conditions according to the selection index value;

[0024] The model item to be deleted is deleted from the previous candidate lithography process model to obtain the candidate lithography process model of the current regression processing, until the first preset regression end condition is met, and multiple candidate lithography process models are obtained; wherein, the previous candidate lithography process model of the first regression processing is the lithography process model.

[0025] In one embodiment, determining a target lithography process model from a plurality of candidate lithography process models according to the model selection index value includes:

[0026] According to the model selection index value, the candidate photolithography process model corresponding to the model selection index value with the smallest value is determined as the target photolithography process model.

[0027] In one embodiment, determining the lithography process window according to the constraint condition and the target lithography process model includes:

[0028] Determining a contour map of the target lithography process model according to the constraint conditions;

[0029] A target region with the largest area is determined in the contour map according to the constraint conditions, and the target region is determined as a photolithography process window.

[0030] In one embodiment, the method further comprises:

[0031] Analyzing the photolithography process window to determine a first length of the photolithography process window along a focal length axis and a second length of the photolithography process window along an exposure energy axis;

[0032] determining a focal depth based on the first length;

[0033] determining a target focal length and a target exposure energy according to the center coordinates of the photolithography process window;

[0034] An exposure energy margin is determined according to the second length and the target exposure energy.

[0035] In a second aspect, the present application further provides a lithography process window analysis device based on stepwise regression, comprising:

[0036] A data acquisition module, used to acquire photoresist pattern image data and determine the constraints required for the photolithography process window;

[0037] A data determination module, configured to determine a photolithography process model satisfying the relationship between the photoresist pattern characteristics and the exposure energy and focal length, and a stepwise regression direction of the photolithography process model;

[0038] a stepwise regression module, configured to perform stepwise regression on the photolithography process model according to the photoresist pattern image data in a direction indicated by the stepwise regression direction, to obtain a plurality of candidate photolithography process models and a model selection index value of each candidate photolithography process model;

[0039] A model screening module, configured to determine a target photolithography process model from a plurality of candidate photolithography process models according to the model selection index value;

[0040] The photolithography process window determination module is used to determine the photolithography process window according to the constraint conditions and the target photolithography process model.

[0041] In a third aspect, the present application further provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0042] Obtaining photoresist pattern image data and determining the constraints required for the photolithography process window;

[0043] Determining a photolithography process model that satisfies the relationship between photoresist pattern characteristics and exposure energy and focal length, and a stepwise regression direction of the photolithography process model;

[0044] Performing stepwise regression on the photolithography process model according to the photoresist pattern image data in the direction indicated by the stepwise regression direction to obtain a plurality of candidate photolithography process models and a model selection index value of each candidate photolithography process model;

[0045] determining a target photolithography process model from a plurality of candidate photolithography process models according to the model selection index value;

[0046] A photolithography process window is determined according to the constraint conditions and the target photolithography process model.

[0047] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the following steps are implemented:

[0048] Obtaining photoresist pattern image data and determining the constraints required for the photolithography process window;

[0049] Determining a photolithography process model that satisfies the relationship between photoresist pattern characteristics and exposure energy and focal length, and a stepwise regression direction of the photolithography process model;

[0050] Performing stepwise regression on the photolithography process model according to the photoresist pattern image data in the direction indicated by the stepwise regression direction to obtain a plurality of candidate photolithography process models and a model selection index value of each candidate photolithography process model;

[0051] determining a target photolithography process model from a plurality of candidate photolithography process models according to the model selection index value;

[0052] A photolithography process window is determined according to the constraint conditions and the target photolithography process model.

[0053] In a fifth aspect, the present application further provides a computer program product, comprising a computer program, which, when executed by a processor, implements the following steps:

[0054] Obtaining photoresist pattern image data and determining the constraints required for the photolithography process window;

[0055] Determining a photolithography process model that satisfies the relationship between photoresist pattern characteristics and exposure energy and focal length, and a stepwise regression direction of the photolithography process model;

[0056] Performing stepwise regression on the photolithography process model according to the photoresist pattern image data in the direction indicated by the stepwise regression direction to obtain a plurality of candidate photolithography process models and a model selection index value of each candidate photolithography process model;

[0057] determining a target photolithography process model from a plurality of candidate photolithography process models according to the model selection index value;

[0058] A photolithography process window is determined according to the constraint conditions and the target photolithography process model.

[0059] The above-mentioned lithography process window analysis method, apparatus, computer device, computer-readable storage medium, and computer program product based on stepwise regression obtain photoresist pattern image data and determine the constraints required for the lithography process window; determine a lithography process model that satisfies the relationship between photoresist pattern features and exposure energy and focal length, and a stepwise regression direction of the lithography process model; perform stepwise regression on the lithography process model based on the photoresist pattern image data in the direction indicated by the stepwise regression direction to obtain multiple candidate lithography process models and model selection index values ​​for each of the candidate lithography process models; determine a target lithography process model from the multiple candidate lithography process models based on the model selection index values; determine the lithography process window based on the constraints and the target lithography process model, and determine the target lithography process model from the multiple candidate lithography process models, thereby determining the lithography process window. This method can eliminate low-confidence items in the model that should not exist, improve the confidence of the model and improve the noise resistance, and thus can correctly reflect the relationship between input and output. On this basis, the accuracy of FEM data analysis results is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the following briefly introduces the drawings required for use in the embodiments of the present application or related technical descriptions. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying any creative work.

[0061] Figure 1 2. A diagram illustrating an application environment of a lithography process window analysis method based on stepwise regression in one embodiment;

[0062] Figure 2 1 is a flow chart of a lithography process window analysis method based on stepwise regression in one embodiment;

[0063] Figure 3 is a schematic diagram of photoresist pattern image data in one embodiment;

[0064] Figure 4 A schematic diagram of a flow chart and a contour map for determining a target photolithography process model in one embodiment;

[0065] Figure 5 is a schematic diagram of a photolithography process window in one embodiment;

[0066] Figure 6 206 is a flow chart of step 206 in one embodiment;

[0067] Figure 71 is a schematic diagram of a process of stepwise regression according to a first direction in one embodiment;

[0068] Figure 8 is a schematic diagram of a process of stepwise regression according to the second direction in one embodiment;

[0069] Figure 9 1 is a schematic flow chart of a method for analyzing a photolithography process window in one embodiment;

[0070] Figure 10 is a schematic diagram of a photolithography process window in one embodiment;

[0071] Figure 11 is a schematic diagram of a photolithography process window in another embodiment;

[0072] Figure 12 is a structural block diagram of a lithography process window analysis device based on stepwise regression in one embodiment;

[0073] Figure 13 FIG. 1 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0074] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0075] Photolithography is a crucial step in semiconductor device manufacturing. It uses exposure and development to create geometric structures on a photoresist layer. Etching then transfers the pattern from the photomask to the substrate. The photolithography process window refers to the range of exposure dose and defocus that ensures the mask pattern is accurately replicated on the silicon wafer during the photolithography process. Determining the photolithography process window is crucial for ensuring the quality and efficiency of chip manufacturing, as it directly affects whether the pattern on the photoresist meets the desired requirements. To achieve this goal, all parameters in the photolithography process must be controlled within a narrow range, known as the process window. Process window analysis is determined using exposure focus-energy matrix (FEM) data. Focus and exposure energy are adjusted during the exposure process to achieve optimal imaging results. The focus-energy matrix (FEM) data is a set of data that includes the photoresist pattern and measured data on the FEM wafer.

[0076] An accurate lithography process window analysis requires a high-confidence model with good noise immunity to produce accurate results when processing actual FEM data. Low model confidence or weak noise immunity directly impacts the accuracy of FEM data analysis, and thus the precise determination of the process window. Therefore, when conducting lithography process window analysis, it is crucial to select or develop a model that combines high confidence with good noise immunity.

[0077] Related art FEM analysis uses regression analysis to fit FEM data to a defined polynomial function model, thereby determining the functional relationship between the photoresist pattern's critical dimension (CD) and exposure energy and focal length. This functional relationship is plotted as a contour map. Based on the required CD range for the photoresist pattern, a region can be identified within the contour map where all CDs meet the requirements. Within this region, the largest rectangle or ellipse is found. The range of exposure energy and focal length corresponding to this rectangle or ellipse represents the photolithography process window.

[0078] However, current lithography process window analysis uses a fixed polynomial function model for fitting. However, this analysis method cannot exclude low-confidence terms in the model that should not exist, resulting in a lack of confidence in the results. Furthermore, this analysis method is weak in noise immunity, resulting in low accuracy when processing FEM data with high noise levels. In other words, the accuracy of lithography process window analysis in related art is low.

[0079] To address this technical problem, through the constraints required by the lithography process window, the lithography process model that satisfies the relationship between the photoresist pattern features and the exposure energy and focal length is regressed stepwise according to the photoresist pattern image data in the direction indicated by the stepwise regression direction to obtain multiple candidate lithography process models and model selection index values ​​for each candidate lithography process model; based on the model selection index values, the target lithography process model is determined from the multiple candidate lithography process models; based on the constraints and the target lithography process model, the lithography process window is determined. This lithography process window analysis method based on stepwise regression can autonomously select the optimal model, provide high confidence and high noise resistance, and thus improve the accuracy of data analysis.

[0080] The lithography process window analysis method based on stepwise regression provided in the embodiment of the present application can be applied to Figure 1In the application environment shown, the terminal 102 communicates with the server 104 via a network. The data storage system can store data that the server 104 needs to process. The data storage system can be integrated on the server 104, or placed on a cloud or other network server. The terminal 102 obtains photoresist pattern image data and the constraints required to determine the photolithography process window; determines the photolithography process model that satisfies the relationship between the photoresist pattern features and the exposure energy and focal length, as well as the stepwise regression direction of the photolithography process model; performs stepwise regression on the photolithography process model based on the photoresist pattern image data in the direction indicated by the stepwise regression direction to obtain multiple candidate photolithography process models and model selection index values ​​for each candidate photolithography process model; determines a target photolithography process model from the multiple candidate photolithography process models based on the model selection index values; and determines the photolithography process window based on the constraints and the target photolithography process model. The terminal 102 can be, but is not limited to, various personal computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. IoT devices can be smart speakers, smart TVs, smart air conditioners, smart car devices, projection equipment, etc. Portable wearable devices can be smart watches, smart bracelets, head-mounted devices, etc. Head-mounted devices can be virtual reality (VR) devices, augmented reality (AR) devices, smart glasses, etc. Server 104 can be an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing cloud computing services.

[0081] In an exemplary embodiment, Figure 2 As shown in the figure, a lithography process window analysis method based on stepwise regression is provided. Figure 1 The terminal in FIG is taken as an example to illustrate the method, including the following steps 202 to 210. Among them:

[0082] Step 202: Acquire photoresist pattern image data and determine the constraints required for the photolithography process window.

[0083] The photoresist pattern image data includes, but is not limited to, exposure energy and focal length. Constraints can be determined based on actual needs. For example, the constraints can include target photoresist pattern features and their corresponding allowable ranges. For example, if the target photoresist pattern feature is a target critical dimension (CD), the corresponding constraints include the target CD and the upper and lower limits of the allowable CD variation.

[0084] It's important to note that photoresist image data is typically stored in an Excel spreadsheet, containing scanning electron microscope (SEM) images of multiple photoresist patterns and their corresponding measurement results. Actual photoresist patterns may contain defects, such as cracks, collapses, and adhesions, which can be visually detected in SEM images. However, automated machine measurement doesn't account for these potential defects. While measuring these defective patterns will yield accurate results, the resulting data doesn't reflect the true situation. These are considered "bad data points." This is like using a machine to measure the diameter of a circle. If the pattern is a perfect circle, the result will be accurate. However, if the pattern is a rectangle, the machine will still measure it as a circle and produce a value, but this value is no longer relevant to the actual pattern. Therefore, to ensure data accuracy and reliability, the raw photoresist pattern image data must be preprocessed to remove bad data points. This preprocessing method can include, but is not limited to, automated processing and discarding of data using machine learning or image recognition, or manual data processing and deletion.

[0085] For example, the original photoresist pattern image data is obtained, pre-processed to obtain the photoresist pattern image data with poor data points removed, and the constraints required for the photolithography process window are determined. Figure 3 1 is a schematic diagram of photoresist graphic image data in one embodiment, wherein the left portion of the figure is a FEM photoresist image, and the right portion is pre-processed FEM data, including critical dimensions, focal length, and exposure energy.

[0086] Step 204 : determining a photolithography process model that satisfies the relationship between the photoresist pattern characteristics and the exposure energy and focal length, and a stepwise regression direction of the photolithography process model.

[0087] The lithography process model can be a model of multiple specific model items, and the specific model items are associated with exposure energy and focal length. The lithography process model can be expressed as or Or any other equivalent form of polynomial function, where the dependent variable CD is the critical dimension of the photoresist pattern feature, the independent variable E is the exposure energy, the independent variable F is the focal length, and for any i∈{0,1,…M}, j∈{0,1,…N}, a ij is the unknown coefficient. When i=j=0, a ij E i F j =a 00It is called a constant term. When the undetermined coefficient of a term is not 0, it is said to be included in the model. It is understandable that the photolithography process model can also be other functional forms used to explain the relationship between photoresist pattern characteristics and exposure energy and focal length, which is not limited here.

[0088] Stepwise regression can be done in either a first or second direction. The first direction is the positive direction, which can be understood as starting with a model containing only a constant term, performing a least squares fit on the photoresist pattern image data, and then selecting an item from the full model and adding it to the model to obtain a new model. The second direction is the negative direction, which can be understood as starting with the full model, performing a least squares fit on the photoresist pattern image data, and then removing an item from the model to obtain a new model. The model before removing the item is called the old model.

[0089] It should be noted that when specifying the direction of stepwise regression, it is also necessary to determine the model selection criteria. The selection criteria can be, but are not limited to, any of the following: p-value, AIC, BIC, adjusted R-squared, PrESS, AICc, and Mallow's Cp. A p-value threshold must also be specified. The p-value is a statistic used to measure the significance of model terms. It is a positive number between 0 and 1, calculated after fitting is complete. The smaller the p-value, the more significant the model term.

[0090] Akaike Information Criterion (AIC), Bayesian Information Criterion (BIC), Prediction Error Sum of Squares (PrESS), and Akaike Information Criterion with correction (AICc) are statistics used to measure the quality of the fitting results. They are calculated after the fitting is completed. The smaller the better.

[0091] Adjusted R-squared: A statistic used to measure the quality of the fitting results. It is a positive number between 0 and 1, calculated after the fitting is completed. The larger the better.

[0092] Mallow's Cp: A statistic used to measure the quality of the fitting results. It is calculated after the fitting is completed. The smaller it is and the closer it is to the number of items included in the model, the better.

[0093] Step 206 , performing stepwise regression on the photolithography process model according to the photoresist pattern image data in the direction indicated by the stepwise regression direction, to obtain a plurality of candidate photolithography process models and a model selection index value of each candidate photolithography process model.

[0094] It is understandable that the direction indicated by the stepwise regression direction is different depending on the initial photolithography process model used for the stepwise regression. This embodiment is described by taking the model selection index value as the p value as an example.

[0095] Exemplarily, according to the direction indicated by the stepwise regression direction, an initial lithography process model corresponding to each stepwise regression direction is determined based on the lithography process model. Based on the initial lithography process model, a candidate lithography process model obtained each time during the stepwise regression process is determined based on the lithography process model. For each candidate lithography process model, the candidate lithography process model is fitted according to the photoresist pattern image data to obtain a selection index value for each model item in each candidate lithography process model until a preset regression condition is satisfied, thereby obtaining multiple candidate lithography process models. Based on the selection index value for each model item in each candidate lithography process model, a model selection index value for each candidate lithography process model is determined. Wherein, when the stepwise regression direction is the first direction, the preset regression condition may be that the new model includes all items included in the full model, or that the p-values ​​of all new items after fitting are greater than a specified p-value threshold. When the stepwise regression direction is the second direction, the preset regression condition may be that the new model only includes a constant term, or that the p-values ​​of all items in the new model after fitting are not greater than a specified p-value threshold.

[0096] Step 208 : determining a target photolithography process model from a plurality of candidate photolithography process models according to the model selection index value.

[0097] The selection criterion for determining the target photolithography process model may be to determine the model with the smallest model selection index value as the target photolithography process model. For example, if the selection criterion is a p-value criterion, then a target photolithography process model is selected from the model records: the model with the smallest maximum p-value in the model stepwise regression record, where the maximum p-value of the model refers to the p-value of the model term with the largest p-value among all model terms included in the model. For example, if the selection criterion is one of the AIC criterion, the BIC criterion, the PrESS criterion, and the AICc criterion, then a model with the smallest corresponding AIC, BIC, PrESS, and AICc in the model records is selected from the model records.

[0098] The selection criterion for determining the target lithography process model may also be to determine the model with the largest model selection index value as the target lithography process model. For example, if the selection criterion is the adjusted R-square criterion, a model is selected from the model regression record: the adjusted R-square of the model is the largest in the model record.

[0099] The selection criteria for determining the target lithography process model may also be that the number of model items of the candidate lithography process model is less than or equal to half of the full model, and the model selection index value of the candidate lithography process model is the smallest in the model record, or the model selection index value of the model in the model record is closest to the number of model items of itself. For example, if the selection criteria is the Mallow's Cp criteria, then a model is selected from the model record: the number of items contained in the model is less than or equal to half of the full model, and the Mallow's Cp of the model is the smallest in the model record, or the Mallow's Cp of the model in the model record is closest to the number of model items of itself. It can be understood that the number of model items closest to itself here can be the difference between Mallow's Cp and the number of model items within a preset numerical range.

[0100] Exemplarily, a model selection index value is determined, and a target lithography process model is determined from a plurality of candidate lithography process models according to a selection criterion corresponding to each model selection index value. Figure 4 In an exemplary embodiment, the optimal model, i.e., the target lithography process model, is selected from the model records based on specified model selection criteria. A contour plot of the model is then plotted based on the constraints, with the horizontal and vertical axes representing the focal length F and exposure energy E, respectively. The horizontal and vertical axes in this plot represent inputs; each point corresponds to an E and F input, and can also correspond to a CD output. All points corresponding to the same CD are plotted and connected with a line. Since different E and F inputs can produce the same CD output, a contour plot is formed. The constraints are the upper and lower limits of the critical dimension (CD). Contour lines corresponding to the upper and lower limits are then drawn, respectively. Since multiple contour lines can correspond to the same CD value, the points in the region between these two or two groups of contour lines (i.e., the allowable critical dimension range) have CD values ​​greater than the lower limit and less than the upper limit.

[0101] Step 210 : determining the photolithography process window according to the constraint conditions and the target photolithography process model.

[0102] For example, according to the specified model selection criteria, the optimal model is selected from the model records, and the contour map of the model is drawn according to the constraints. According to the constraints, the maximum area is drawn in the contour map to obtain the lithography process window. The target lithography process model takes the focal length F and exposure energy E as input and the critical dimension CD as output. Given any E and F values, the CD value can be known. Based on this binary function, the contour map can be drawn, such as Figure 5 As shown in an exemplary embodiment, according to the constraint conditions, a maximum area rectangle or ellipse is drawn in the contour map to obtain a schematic diagram of the lithography process window, where the horizontal and vertical axes are the focal length F and the exposure energy E, respectively.

[0103] The above-mentioned lithography process window analysis method based on stepwise regression obtains photoresist pattern image data and determines the constraints required for the lithography process window; determines the lithography process model that satisfies the relationship between the photoresist pattern characteristics and the exposure energy and focal length, as well as the stepwise regression direction of the lithography process model; performs stepwise regression on the lithography process model according to the photoresist pattern image data in the direction indicated by the stepwise regression direction to obtain multiple candidate lithography process models and model selection index values ​​for each candidate lithography process model; determines a target lithography process model from the multiple candidate lithography process models according to the model selection index values; determines the lithography process window according to the constraints and the target lithography process model, determines the target lithography process model from the multiple candidate lithography process models, and then determines the lithography process window. This method can exclude items with low confidence that should not exist in the model, improve the confidence of the model and improve the noise resistance, and thus can correctly reflect the relationship between input and output. On this basis, the accuracy of the FEM data analysis results is improved.

[0104] The lithography process window ensures that the pattern on the mask can be accurately copied to the silicon wafer. In order to ensure the reliability of the lithography process window, the confidence of the model needs to be improved.

[0105] In an exemplary embodiment, Figure 6 As shown, step 206 includes steps 602 to 606. Among them:

[0106] Step 602 : determining candidate photolithography process models obtained each time in the stepwise regression process according to the photolithography process model in the direction indicated by the stepwise regression direction, until a preset regression condition is met, thereby obtaining a plurality of candidate photolithography process models.

[0107] It is understandable that the directions indicated by the stepwise regression directions are different, and the corresponding preset regression conditions are also different. Two different stepwise regression directions are provided below.

[0108] Method 1: According to the first direction indicated by the stepwise regression direction, the baseline photolithography process model is determined according to the photolithography process model; the model items to be added for the first regression are determined from the photolithography process model, and the candidate photolithography process model for the first regression is determined according to the baseline photolithography process model and the model items to be added; for each regression processing after the second time, the model items to be added for each regression are determined from the photolithography process model; wherein, the model items to be added for each regression are different; the candidate photolithography process model for the current regression is determined according to the model items to be added and the previous candidate photolithography process model, until the first preset regression end condition is met, and multiple candidate photolithography process models are obtained.

[0109] The reference photolithography process model may also be understood as an initial photolithography process model, and the reference photolithography process model may be a model containing only constant terms.

[0110] For example, regression begins with a model containing only constant terms, performs a least squares fit on this model using the photoresist pattern image data, and selects a new model term from the full model and adds it to the model to obtain a new model. The model before the new term is added is called the old model. The method for selecting which new term to add as the new model term can be: first, each term in the full model that is not included in the old model is added to obtain multiple new models, then the photoresist pattern image data is subjected to a least squares fit on each of these new models, and the p-value of each new term in each new model after fitting is obtained. The case where the p-value of the new term is the smallest and the p-value is not greater than a specified p-value threshold is selected. This process is repeated until the new model contains all terms included in the full model, or the p-values ​​of all new terms after fitting are greater than a specified p-value threshold, and the regression is terminated to obtain multiple candidate photolithography process models. Wherein, least squares fitting: regression analysis is performed using the least squares method to fit the data to a function model to solve for the unknown coefficients in the function model. After fitting is completed, the unknown coefficients of each item in the function model, the p-value of each item, and many other statistics can be obtained.

[0111] like Figure 7 The figure is a schematic diagram of a stepwise regression process in a first direction in an exemplary embodiment. Add the model item A1 to be added to the baseline lithography process model to obtain CD=a 00 +A1, in the next regression, add the new model item A2 on the basis of the previous one, and get CD=a 00 +A1+A2, and so on. At the (M+1)(N+1)-2th regression, we get CD=a 00 +A1+A2+…+A (M+1)(N+1)-2 , in the (M+1)(N+1)-1th regression, the full model is obtained, namely:

[0112] In this way, model items are added one by one to obtain multiple candidate photolithography process models, that is, model records. (M+1)(N+1)-1 Represents a model term.

[0113] What needs to be understood is that until the new model contains all the items contained in the full model, the construction process of each intermediate model cannot be omitted, because until the model construction is completed, it is not known whether the model is good or not, so the full model cannot be directly used as the final new model. In this way, we can gradually regress and screen out models with high confidence and exclude models with low confidence to ensure the model with the best performance, thereby ensuring the reliability of the lithography process window. On this basis, it has high noise resistance, and when processing noisy data, the processing results are more accurate than existing methods.

[0114] Method 2: According to the second direction indicated by the stepwise regression direction, for each regression processing, obtain the selection index value of each model item in the previous candidate lithography process model; determine the model items to be deleted that meet the preset conditions from all model items based on the selection index value; delete the model items to be deleted from the previous candidate lithography process model to obtain the candidate lithography process model of the current regression processing, until the second preset regression end condition is met, and multiple candidate lithography process models are obtained; among them, the previous candidate lithography process model of the first regression processing is the lithography process model.

[0115] Among them, the method of selecting which item to eliminate here (i.e., the model item to be deleted) is: it can refer to the item in the model item that has the largest index value and is greater than the specified p-value threshold after the last candidate lithography process model is fitted. The second preset regression end condition can be that the new model only contains constant terms, or the p-values ​​of all model items in the new model after fitting are not greater than the specified p-value threshold. It should be noted that the deletion here can refer to excluding it from the consideration range of fitting, so that the undetermined coefficient of this item in the fitting result is zero. The last candidate lithography process model of the first regression processing is the lithography process model, which can also be understood as the full model, for example:

[0116]

[0117] For example, starting with a full model, i.e., the photolithography process model, the photoresist pattern image data is fitted to this model using a least squares method. One item is then selected from this model and removed to obtain a new model. The model before the item is removed is referred to as the old model. This process of removing items from the old model is repeated until the new model contains only constant items, or the p-values ​​of all items in the new model after fitting are no greater than a specified p-value threshold. This results in multiple candidate photolithography process models.

[0118] like Figure 8 The figure shows a flow chart of stepwise regression in the second direction in an exemplary embodiment. Eliminate the model items one by one, and in the (M+1)(N+1)-2th regression, we get CD=a 00+A1+A2+…+A (M+1)(N+1)-2 , and so on, we get model 2 CD=a 00 +A1+A2 and Model 1 CD=a 00 +A1+A2, in this way, model items are eliminated one by one to obtain multiple candidate lithography process models, that is, model records. In this way, it is possible to gradually regress and screen out models with high confidence and eliminate models with low confidence, ensuring the model with the best performance, thereby ensuring the reliability of the lithography process window. On this basis, it has high noise resistance and, when processing noisy data, the processing results are more accurate than existing methods. In addition, this method does not require every remaining item in the entire model to be put in for trial, which reduces the number of calculations and improves the utilization of computing resources.

[0119] Step 604 : For each candidate photolithography process model, fit the candidate photolithography process model according to the photoresist pattern image data to obtain a selection index value of each model item in each candidate photolithography process model.

[0120] The fitting method can be, but is not limited to, the least squares method. It should be noted that each model fitting process obtains the specific form of the model, that is, the terms and coefficients of each model (function), and also obtains relevant data for determining the target lithography process model and drawing contour maps.

[0121] Step 606 : Determine the model selection index value of each candidate lithography process model according to the selection index value.

[0122] Among them, determining the model selection index value according to the selection index value can be to determine the selection index value with the largest value among all model items in the candidate lithography process model as the model selection index value of the candidate lithography process model. There is no limitation on the method of determining the model selection index value according to the selection index value.

[0123] Exemplarily, for each candidate photolithography process model, the maximum selection index value of all model items in the candidate photolithography process model is determined as the model selection index value of the candidate photolithography process model.

[0124] In the above embodiment, using the photoresist pattern image data to fit the candidate photolithography process model can ensure that the selected model is closer to the actual process, improve the prediction accuracy of the model, and further improve the photolithography process to achieve higher-precision chip manufacturing.

[0125] The target lithography process model is determined by performing stepwise regression on the model. The following provides a method for determining the lithography process window. In an exemplary embodiment, the lithography process window is determined based on the constraint conditions and the target lithography process model, including:

[0126] Determine a contour map of a target photolithography process model according to the constraint conditions; determine a target region with the largest area in the contour map according to the constraint conditions, and determine the target region as a photolithography process window.

[0127] Among them, the method of determining the contour map can be implemented by existing methods and will not be elaborated here. The target area can be that when the lithography machine fluctuates, the parameters of the lithography machine are still within the target area, that is, the lithography process can still produce qualified photoresist patterns. In other words, this target area defines the fault tolerance range of the lithography conditions. As long as the fluctuation is within this range, the lithography quality will not be affected, thereby ensuring the stability of the production process and the yield rate of the chip. The target area can be elliptical or rectangular.

[0128] The process window obtained by the above analysis method can be closer to the ideal, noise-free, interference-free, and real process window, and has high noise resistance.

[0129] It should be noted that the lithography process window is crucial to ensuring the stability of the lithography process and the yield rate of chips. Through process window analysis, it can be ensured that even in the case of parameter fluctuations, the lithography process can still produce chips that meet the requirements.

[0130] In an illustrative example, Figure 9 As shown, a method for analyzing a lithography process window is provided, comprising steps 902 to 908, wherein:

[0131] Step 902 : Analyze the photolithography process window to determine a first length of the photolithography process window along the focal length axis and a second length of the photolithography process window along the exposure energy axis.

[0132] Step 904: Determine the focus depth according to the first length.

[0133] Step 906 : determining a target focal length and a target exposure energy according to the center coordinates of the photolithography process window.

[0134] Step 908: Determine the exposure energy margin according to the second length and the target exposure energy.

[0135] Exposure energy margin is a measure of the photoresist's tolerance to exposure energy variations. A larger exposure energy margin means the photoresist is less sensitive to fluctuations in exposure energy, thereby improving the process's fault tolerance.

[0136] For example, when the lithography process window is a rectangle, the first length of the rectangle along the focal length axis is recorded as the depth of focus (DOF), the second length of the rectangle along the exposure energy axis is recorded as ΔE, and the focal length and exposure energy corresponding to the center point of the rectangle are recorded as the target focal length and target exposure energy, respectively. The exposure energy margin (Exposure Latitude, EL) is calculated as follows: EL = 100% × ΔE / target exposure energy, as shown in FIG. Figure 10 shown.

[0137] When the lithography process window is an ellipse, the axis length of the ellipse along the focal length axis is recorded as the depth of focus (DOF), the axis length of the ellipse along the exposure energy axis is recorded as ΔE, and the focal length and exposure energy corresponding to the center point of the ellipse are recorded as the target focal length and target exposure energy, that is, the optimal focal length and optimal exposure energy, respectively. The exposure energy margin (Exposure Latitude, EL) is calculated as follows: EL = 100% × ΔE / target exposure energy, such as Figure 11 shown.

[0138] In the above manner, by analyzing the photolithography process window and obtaining the above parameters, the photolithography process can be optimized, controlled and improved, thereby improving the quality of photolithography production and reducing production costs.

[0139] In an exemplary embodiment, a lithography process window analysis method based on stepwise regression includes the following steps: obtaining photoresist pattern image data and determining the constraints required for the lithography process window; determining a lithography process model that satisfies the relationship between the photoresist pattern features and the exposure energy and focal length, and the stepwise regression direction of the lithography process model.

[0140] When the direction indicated by the stepwise regression direction is the first direction, the baseline photolithography process model is determined according to the photolithography process model; the model items to be added for the first regression are determined from the photolithography process model, and the candidate photolithography process model for the first regression is determined according to the baseline photolithography process model and the model items to be added; for each regression processing after the second time, the model items to be added for each regression are determined from the photolithography process model; wherein the model items to be added for each regression are different each time; the candidate photolithography process model for the current regression is determined according to the model items to be added and the previous candidate photolithography process model, until the first preset regression end condition is met, and multiple candidate photolithography process models are obtained.

[0141] When the direction indicated by the stepwise regression direction is the second direction, for each regression processing, the selection index value of each model item in the previous candidate lithography process model is obtained; according to the selection index value, the model items to be deleted that meet the preset conditions are determined from all model items; the model items to be deleted are deleted from the previous candidate lithography process model to obtain the candidate lithography process model of the current regression processing, until the second preset regression end condition is met, and multiple candidate lithography process models are obtained; among them, the previous candidate lithography process model of the first regression processing is the lithography process model.

[0142] For each candidate photolithography process model, the candidate photolithography process model is fitted according to the photoresist graphic image data to obtain the selection index value of each model item in each candidate photolithography process model; according to the model selection index value, the candidate photolithography process model corresponding to the model selection index value with the smallest numerical value is determined as the target photolithography process model; according to the constraint conditions and the target photolithography process model, the photolithography process window is determined; the photolithography process window is analyzed to determine the first length of the photolithography process window along the focal length axis and the second length of the photolithography process window along the exposure energy axis; the focus depth is determined according to the first length; the target focal length and target exposure energy are determined according to the center coordinates of the photolithography process window; the exposure energy margin is determined according to the second length and the optimal exposure energy.

[0143] It should be noted that the specific implementation of this embodiment can be achieved through the above-mentioned limitation method, which will not be elaborated here. Compared with the existing photolithography process window analysis method, the above-mentioned process window analysis method based on stepwise regression screens out models with high confidence, excludes models with low confidence, and uses statistics to select the optimal model. This method also has high noise resistance and can obtain better results than existing methods when processing noisy data, so that the process window obtained by analysis can be closer to the "ideal, noise-free, interference-free, real" process window, further improving the quality of photolithography production and reducing production costs.

[0144] It should be understood that, although the steps in the flowcharts of the above embodiments are shown in sequence as indicated by the arrows, these steps are not necessarily performed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and these steps can be performed in other orders. Moreover, at least a portion of the steps in the flowcharts of the above embodiments may include multiple steps or multiple stages, and these steps or stages are not necessarily performed at the same time, but can be performed at different times. The execution order of these steps or stages is not necessarily to be performed in sequence, but can be performed in turn or alternately with other steps or at least a portion of steps or stages in other steps.

[0145] Based on the same inventive concept, the embodiments of the present application further provide a stepwise regression-based lithography process window analysis device for implementing the aforementioned stepwise regression-based lithography process window analysis method. The implementation solution provided by the device is similar to the implementation solution described in the aforementioned method. Therefore, the specific limitations in the embodiments of one or more lithography process window analysis devices provided below can be found in the above-mentioned limitations on the stepwise regression-based lithography process window analysis method, and will not be repeated here.

[0146] In an exemplary embodiment, Figure 12 As shown, a lithography process window analysis device based on stepwise regression is provided, comprising: a data acquisition module 1202, a data determination module 1204, a stepwise regression module 1206, a model screening module 1208 and a lithography process window determination module 1210, wherein:

[0147] The data acquisition module 1202 is used to acquire the photoresist pattern image data and determine the constraint conditions required for the photolithography process window.

[0148] The data determination module 1204 is used to determine a photolithography process model that satisfies the relationship between the photoresist pattern characteristics and the exposure energy and focal length, as well as a stepwise regression direction of the photolithography process model.

[0149] The stepwise regression module 1206 is configured to perform stepwise regression on the photolithography process model according to the photoresist pattern image data in the direction indicated by the stepwise regression direction to obtain multiple candidate photolithography process models and a model selection index value of each candidate photolithography process model.

[0150] The model screening module 1208 is configured to determine a target photolithography process model from a plurality of candidate photolithography process models according to a model selection index value.

[0151] The photolithography process window determination module 1210 is configured to determine the photolithography process window according to the constraint conditions and the target photolithography process model.

[0152] The above-mentioned lithography process window analysis device based on stepwise regression obtains photoresist pattern image data and determines the constraints required for the lithography process window; determines the lithography process model that satisfies the relationship between the photoresist pattern characteristics and the exposure energy and focal length, and the stepwise regression direction of the lithography process model; performs stepwise regression on the lithography process model according to the photoresist pattern image data in the direction indicated by the stepwise regression direction to obtain multiple candidate lithography process models and model selection index values ​​of each candidate lithography process model; determines a target lithography process model from the multiple candidate lithography process models according to the model selection index values; determines the lithography process window according to the constraints and the target lithography process model, determines the target lithography process model from the multiple candidate lithography process models, and then determines the lithography process window. This method can exclude items with low confidence that should not exist in the model, improve the confidence of the model and improve the anti-noise ability, and thus can correctly reflect the relationship between input and output. On this basis, the accuracy of the FEM data analysis results is improved.

[0153] In an exemplary embodiment, the stepwise regression module 1206 is further configured to determine, according to the lithography process model, a candidate lithography process model obtained each time in the stepwise regression process, in a direction indicated by the stepwise regression direction, until a preset regression condition is satisfied, thereby obtaining a plurality of candidate lithography process models;

[0154] For each candidate photolithography process model, fitting the candidate photolithography process model according to the photoresist pattern image data to obtain a selection index value of each model item in each candidate photolithography process model;

[0155] According to the selection index value, the model selection index value of each candidate photolithography process model is determined.

[0156] In an exemplary embodiment, the stepwise regression module 1206 is further configured to determine the reference photolithography process model according to the photolithography process model in a first direction indicated by the stepwise regression direction;

[0157] Determine a model item to be added for the first regression from the lithography process model, and determine a candidate lithography process model for the first regression based on the baseline lithography process model and the model item to be added;

[0158] For each regression process after the second one, determining a model item to be added for each regression from the photolithography process model; wherein the model item to be added for each regression is different;

[0159] According to the model item to be newly added and the last candidate photolithography process model, the candidate photolithography process model for current regression is determined until a first preset regression end condition is met, thereby obtaining multiple candidate photolithography process models.

[0160] In an exemplary embodiment, the stepwise regression module 1206 is further configured to obtain, for each regression process, a selection index value of each model item in the previous candidate lithography process model according to a second direction indicated by the stepwise regression direction;

[0161] Determine the model items to be deleted that meet the preset conditions from all model items according to the selection index value;

[0162] Delete the model items to be deleted from the previous candidate lithography process model to obtain the candidate lithography process model of the current regression processing, until the second preset regression end condition is met, and obtain multiple candidate lithography process models; among them, the previous candidate lithography process model of the first regression processing is the lithography process model.

[0163] In an exemplary embodiment, the model screening module 1208 is configured to determine, according to the model selection index value, a candidate photolithography process model corresponding to a minimum model selection index value as a target photolithography process model.

[0164] In an exemplary embodiment, the photolithography process window determination module 1210 is further configured to determine a contour map of a target photolithography process model according to the constraint conditions;

[0165] A target region with the largest area is determined in the contour map according to the constraint conditions, and the target region is determined as the photolithography process window.

[0166] In an exemplary embodiment, the photolithography process window analysis device further includes an analysis module, the analysis module being configured to analyze the photolithography process window to determine a first length of the photolithography process window along a focal length axis and a second length of the photolithography process window along an exposure energy axis;

[0167] determining a depth of focus based on the first length;

[0168] Determine the target focal length and target exposure energy according to the center coordinates of the photolithography process window;

[0169] An exposure energy margin is determined based on the second length and the target exposure energy.

[0170] Each module in the aforementioned stepwise regression-based lithography process window analysis device can be implemented in whole or in part through software, hardware, or a combination thereof. Each module can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in the computer device in the form of software, so that the processor can call and execute the corresponding operations of each module.

[0171] In an exemplary embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as shown in FIG. Figure 13As shown. The computer device includes a processor, a memory, an input / output interface, a communication interface, a display unit and an input device. The processor, the memory and the input / output interface are connected via a system bus, and the communication interface, the display unit and the input device are connected to the system bus via the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be implemented through WIFI, a mobile cellular network, near field communication (NFC) or other technologies. When the computer program is executed by the processor, a lithography process window analysis method based on stepwise regression is implemented. The display unit of the computer device is used to form a visually visible picture, which can be a display screen, a projection device or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a button, trackball or touchpad set on the computer device casing, or an external keyboard, touchpad or mouse.

[0172] Those skilled in the art will understand that Figure 13 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0173] In one embodiment, a computer device is further provided, including a memory and a processor. The memory stores a computer program, and the processor implements the steps in the above method embodiments when executing the computer program.

[0174] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0175] In one embodiment, a computer program product is provided, including a computer program, which implements the steps in the above method embodiments when executed by a processor.

[0176] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0177] Those skilled in the art will appreciate that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, database or other media used in the embodiments provided in this application may include at least one of non-volatile memory and volatile memory. Non-volatile memory may include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory may include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in the various embodiments provided herein may include at least one of a relational database and a non-relational database. Non-relational databases may include, but are not limited to, distributed databases based on blockchains. The processor involved in the various embodiments provided herein may be, but are not limited to, a general-purpose processor, a central processing unit (CPU), a graphics processing unit (GPU), a digital signal processor (DSP), a programmable logic unit (PLC), a data processing logic unit based on quantum computing, an artificial intelligence (AI) processor, and the like.

[0178] The technical features of the above embodiments can be combined arbitrarily. In order to make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0179] The above embodiments merely illustrate several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that a person skilled in the art may make various modifications and improvements without departing from the spirit of the present invention, all of which fall within the scope of protection of the present application. Therefore, the scope of protection of the present application shall be determined by the appended claims.

Claims

1. A lithography process window analysis method based on stepwise regression, characterized in that: The method comprises: Obtaining photoresist pattern image data and determining the constraints required for the photolithography process window; Determining a photolithography process model that satisfies the relationship between photoresist pattern characteristics and exposure energy and focal length, and a stepwise regression direction of the photolithography process model; Performing stepwise regression on the photolithography process model according to the photoresist pattern image data in the direction indicated by the stepwise regression direction to obtain a plurality of candidate photolithography process models and a model selection index value of each candidate photolithography process model; determining a target photolithography process model from a plurality of candidate photolithography process models according to the model selection index value; A photolithography process window is determined according to the constraint conditions and the target photolithography process model.

2. The method according to claim 1, characterized in that The stepwise regression is performed on the photolithography process model according to the photoresist pattern image data in the direction indicated by the stepwise regression direction to obtain a plurality of candidate photolithography process models and a model selection index value of each candidate photolithography process model, including: Determining the candidate photolithography process models obtained each time in the stepwise regression process according to the photolithography process model in the direction indicated by the stepwise regression direction until a preset regression condition is met, thereby obtaining a plurality of candidate photolithography process models; For each candidate photolithography process model, fitting the candidate photolithography process model according to the photoresist pattern image data to obtain a selection index value of each model item in each candidate photolithography process model; According to the selection index value, a model selection index value of each candidate photolithography process model is determined.

3. The method according to claim 2, characterized in that The step of determining the candidate photolithography process models obtained each time in the stepwise regression process according to the photolithography process model in the direction indicated by the stepwise regression direction until a preset regression condition is satisfied, thereby obtaining a plurality of candidate photolithography process models, including: Determining a reference photolithography process model according to the photolithography process model in a first direction indicated by the stepwise regression direction; Determine a model item to be added for the first regression from the photolithography process model, and determine a candidate photolithography process model for the first regression based on the reference photolithography process model and the model item to be added; For each regression process after the second one, determining a model item to be added for each regression from the photolithography process model; wherein the model item to be added for each regression is different; According to the model item to be newly added and the last candidate photolithography process model, the candidate photolithography process model for current regression is determined until a first preset regression end condition is met, thereby obtaining multiple candidate photolithography process models.

4. The method according to claim 2, characterized in that The step of determining the candidate photolithography process models obtained each time in the stepwise regression process according to the photolithography process model in the direction indicated by the stepwise regression direction until a preset regression condition is satisfied, thereby obtaining a plurality of candidate photolithography process models, including: According to the second direction indicated by the stepwise regression direction, for each regression process, obtaining the selection index value of each model item in the previous candidate photolithography process model; Determining, from all the model items, the model items to be deleted that meet the preset conditions according to the selection index value; Delete the model item to be deleted from the previous candidate lithography process model to obtain the candidate lithography process model of the current regression processing, until the second preset regression end condition is met, and obtain multiple candidate lithography process models; among which, the previous candidate lithography process model of the first regression processing is the lithography process model.

5. The method according to any one of claims 1 to 4, characterized in that The step of determining a target photolithography process model from a plurality of candidate photolithography process models according to the model selection index value includes: According to the model selection index value, the candidate photolithography process model corresponding to the model selection index value with the smallest value is determined as the target photolithography process model.

6. The method according to claim 5, characterized in that Determining the lithography process window according to the constraint conditions and the target lithography process model includes: Determining a contour map of the target lithography process model according to the constraint conditions; A target region with the largest area is determined in the contour map according to the constraint conditions, and the target region is determined as a photolithography process window.

7. The method according to claim 6, characterized in that The method further comprises: Analyzing the photolithography process window to determine a first length of the photolithography process window along a focal length axis and a second length of the photolithography process window along an exposure energy axis; determining a focal depth based on the first length; determining a target focal length and a target exposure energy according to the center coordinates of the photolithography process window; An exposure energy margin is determined according to the second length and the target exposure energy.

8. A lithography process window analysis device based on stepwise regression, characterized in that: The device comprises: A data acquisition module, used to acquire photoresist pattern image data and determine the constraints required for the photolithography process window; A data determination module, configured to determine a photolithography process model satisfying the relationship between the photoresist pattern characteristics and the exposure energy and focal length, and a stepwise regression direction of the photolithography process model; a stepwise regression module, configured to perform stepwise regression on the photolithography process model according to the photoresist pattern image data in a direction indicated by the stepwise regression direction, to obtain a plurality of candidate photolithography process models and a model selection index value for each of the candidate photolithography process models; A model screening module, configured to determine a target photolithography process model from a plurality of candidate photolithography process models according to the model selection index value; The photolithography process window determination module is used to determine the photolithography process window according to the constraint conditions and the target photolithography process model.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images