A light source mask co-optimization method and apparatus

Abstract

The present application relates to the field of integrated circuit manufacturing, in particular to a light source mask collaborative optimization method and device, obtaining an initial light source and an initial mask pattern; the initial mask pattern is a binary image; the adjusted initial light source and the adjusted initial mask pattern corresponding to the minimum function value of the light source mask collaborative optimization loss function are the optimized light source and the optimized mask pattern; the light source mask collaborative optimization loss function includes a main pattern edge error term and a sub-resolution auxiliary pattern printing term; a simulation lithography image is obtained, and it is judged whether there is a sub-resolution auxiliary pattern printing in the simulation lithography image; when there is a printed sub-resolution auxiliary pattern, output the latest optimized mask pattern as a process iteration mask pattern; calculate the proportion value; when the proportion value does not exceed the upper limit value, output the optimized light source and the process iteration mask pattern. The present application can ensure that there is no sub-resolution auxiliary pattern printing and a larger process window is retained.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of integrated circuit manufacturing, and in particular to a method and device for source mask optimization. BACKGROUND

[0002] Source mask optimization (SMO) is a technique to co-optimizing the illumination source and the mask pattern for a given set of target patterns, in order to obtain the best process window. As an important resolution enhancement technique, SMO is widely used in advanced nodes of integrated circuit manufacturing.

[0003] In SMO, for non-dense target patterns, the optimized mask pattern usually contains sub-resolution assisting features (SRAF). Adding SRAF can effectively improve the process window. Generally speaking, larger SRAF can reduce the edge placement error (EPE) under defocus and dose bias conditions, but will increase the EPE under mask dimension bias conditions, and will also greatly increase the probability of printing SRAF on the wafer after lithography. On the other hand, if smaller SRAF is used, although the probability of printing SRAF on the wafer after lithography is reduced, the process window may not be large enough to meet the process requirements.

[0004] In other words, the SMO in the prior art does not fully consider the risk of printing SRAF on the wafer after lithography, so when the optimized source and mask pattern are used for actual lithography process verification, the situation of printing SRAF on the wafer after lithography often occurs. On the other hand, if the optimized source is used and verification is performed based on smaller SRAF, the process window may not be large enough to meet the process requirements.

[0005] Therefore, how to ensure a large process window of the SMO method without printing SRAF is a problem to be solved by those skilled in the art. SUMMARY

[0006] The purpose of the present application is to provide a method and device for source mask optimization to solve the problem that the prior art cannot have a large process window without printing SRAF.

[0007] To solve the above technical problems, the present application provides a method for source mask optimization, comprising:

[0008] receiving a lithography target pattern;

[0009] gridding the lithography target pattern to obtain a lithography target image;

[0010] obtaining an initial light source and an initial mask pattern according to the lithography target image; the initial mask pattern is a binary image;

[0011] performing iterative optimization on the initial light source and the initial mask pattern based on the lithography target image by using a preset light source and mask collaborative optimization loss function, and determining that the adjusted initial light source and the adjusted initial mask pattern corresponding to a minimum function value of the light source and mask collaborative optimization loss function are an optimized light source and an optimized mask pattern; the light source and mask collaborative optimization loss function includes a main pattern edge error term and a sub-resolution auxiliary pattern printing term;

[0012] calculating a first process window according to the optimized light source and the optimized mask pattern;

[0013] performing simulation on the optimized mask pattern under the optimized light source to obtain a simulation lithography image, and determining whether there is a sub-resolution auxiliary pattern printed in the simulation lithography image; when there is a printed sub-resolution auxiliary pattern, reducing the size of the corresponding pattern in the optimized mask pattern to obtain an updated optimized mask pattern, and then performing simulation on the updated optimized mask pattern until there is no sub-resolution auxiliary pattern printed in the simulation lithography image corresponding to the optimized mask pattern, and outputting the latest optimized mask pattern as a process iteration mask pattern;

[0014] calculating a second process window according to the optimized light source and the process iteration mask pattern;

[0015] calculating a proportion value of a difference between the first process window and the second process window to the first process window, and determining whether the proportion value exceeds a preset upper limit value;

[0016] when the proportion value does not exceed the upper limit value, outputting the optimized light source and the process iteration mask pattern.

[0017] Optionally, in the light source and mask collaborative optimization method, the light source and mask collaborative optimization loss function is a sum of the main pattern edge error term and the sub-resolution auxiliary pattern printing term; and the sub-resolution auxiliary pattern printing term includes a sub-resolution weight coefficient.

[0018] After determining whether the proportion value exceeds the preset upper limit value, the method further includes:

[0019] When the ratio value exceeds the qualified upper limit value, the sub-resolution weight coefficient is increased to obtain an updated light source and mask co-optimization loss function;

[0020] Based on the lithography target image, the initial light source and the initial mask pattern are iteratively optimized by using the updated light source and mask co-optimization loss function to obtain an updated optimized light source and an updated optimized mask pattern, and a corresponding first process window, a corresponding process iteration mask pattern, a corresponding second process window, and a corresponding ratio value are calculated. Then, it is determined whether the updated ratio value exceeds the preset qualified upper limit value. When the updated ratio value exceeds the qualified upper limit value, the sub-resolution weight coefficient is increased to obtain an updated light source and mask co-optimization loss function, until the ratio value does not exceed the qualified upper limit value.

[0021] Optionally, in the light source and mask co-optimization method, based on the lithography target image, the initial light source and the initial mask pattern are iteratively optimized by using a preset light source and mask co-optimization loss function, including:

[0022] The initial light source and the initial mask pattern are iteratively optimized by the following formula:

[0023] ;

[0024] wherein, w p is a weight coefficient of the process condition p corresponding to the initial light source, w x is a weight coefficient of the evaluation point x on the lithography target pattern corresponding to the initial mask pattern, EPE p,x is an edge placement error of the evaluation point x corresponding position under the process condition p, β is the sub-resolution weight coefficient, R S j (p) is an image value of the sub-resolution auxiliary pattern printing grid point p under the process condition p, T s is a printing threshold of the sub-resolution auxiliary pattern.

[0025] Optionally, in the light source and mask co-optimization method, the size of the printed sub-resolution auxiliary pattern in the corresponding pattern in the optimized mask pattern is reduced to obtain an updated optimized mask pattern, including:

[0026] The size of the printed sub-resolution auxiliary pattern in the corresponding pattern in the optimized mask pattern is reduced to obtain an adjusted mask pattern;

[0027] Based on the lithography target image, the adjustment loss function of the to-be-adjusted mask pattern under the optimized light source is optimized, and the to-be-adjusted mask pattern corresponding to the minimum function value of the adjustment loss function is determined as the updated optimized mask pattern:

[0028] ;

[0029] wherein w' p is a weight coefficient of the process condition p corresponding to the optimized light source, w' x is a weight coefficient of the evaluation point x on the lithography target pattern corresponding to the to-be-adjusted mask pattern, EPE' p,x is an edge placement error of the evaluation point x corresponding position under the process condition p.

[0030] Optionally, in the light source and mask collaborative optimization method, the initial light source and the initial mask pattern are obtained according to the lithography target image, comprising:

[0031] Based on the lithography target image, the initial light source and the initial mask image are obtained by optimizing the initial loss function as follows:

[0032] ;

[0033] wherein w'' p is a weight coefficient of the process condition p corresponding to the initial light source, w'' x is a weight coefficient of the evaluation point x on the lithography target pattern corresponding to the initial mask image, R M i (p) is an image value of the grid point i of the simulated lithography after image of the initial mask image under the process condition p, O i is an image value of the grid point i of the lithography target image;

[0034] The initial mask image is binarized to obtain the initial mask pattern.

[0035] Optionally, in the light source and mask collaborative optimization method, the qualified upper limit value is not greater than 5%.

[0036] Optionally, in the light source and mask collaborative optimization method, the simulation in the iterative optimization of the initial light source and the initial mask pattern, and / or the simulation of the optimized mask pattern, is performed by a pre-trained sub-resolution auxiliary graphic exposure model.

[0037] The training method of the sub-resolution auxiliary graphic exposure model comprises:

[0038] establishing an uncalibrated sub-resolution assist feature exposure model; the sub-resolution assist feature exposure model is capable of determining a main pattern image plane light intensity distribution image from an inputted to-be-processed mask pattern through a preset first cross transfer coefficient, calculating a main pattern image plane photoresist image corresponding to the main pattern image plane light intensity distribution image through a preset photoresist model term and a corresponding linear coefficient, and determining a main pattern critical dimension from the main pattern image plane photoresist image through a first threshold value; the sub-resolution assist feature exposure model is capable of determining a sub-resolution assist feature image plane light intensity distribution image from the inputted to-be-processed mask pattern through a preset second cross transfer coefficient, calculating a sub-resolution assist feature image plane photoresist image corresponding to the sub-resolution assist feature image plane light intensity distribution image through the photoresist model term and the linear coefficient, and determining sub-resolution assist feature printing information from the sub-resolution assist feature image plane photoresist image through a second threshold value;

[0039] manufacturing a test mask according to a preset modeling pattern set, and acquiring corresponding acquired photoresist images of the test mask under different illumination light sources; the modeling pattern set comprises main patterns and sub-resolution assist features of multiple different sizes;

[0040] determining main pattern critical dimensions and sub-resolution assist feature printing information corresponding to each of the acquired photoresist images;

[0041] calibrating the first cross transfer coefficient, the second cross transfer coefficient, the first threshold value, the second threshold value, the photoresist model term and the corresponding linear coefficient by using the main pattern critical dimensions and the sub-resolution assist feature printing information corresponding to each of the illumination light sources and the modeling pattern set, so that differences between the main pattern critical dimensions and the sub-resolution assist feature printing information obtained by inputting the modeling pattern set as a to-be-processed mask pattern into the sub-resolution assist feature exposure model and the main pattern critical dimensions and the sub-resolution assist feature printing information of the acquired photoresist images under corresponding illumination light sources are less than a preset calibration threshold value.

[0042] Optionally, in the light source mask cooperative optimization method, the sub-resolution assist feature exposure model determines the main pattern image plane light intensity distribution image through the following formula:

[0043] ;

[0044] wherein, I M i (x, y) is the main pattern image plane light intensity distribution image, M(x, y) is a rasterized mask image corresponding to the to-be-processed mask pattern, TCC i (ZM is the first cross transfer coefficient, is a convolution operation;

[0045] The sub-resolution assist feature patterning exposure model determines a sub-resolution assist feature image plane light intensity distribution image by:

[0046] ;

[0047] wherein I S i (x,y) is the sub-resolution assist feature image plane light intensity distribution image, M(x,y) is the rasterized mask image corresponding to the mask pattern to be processed, TCC i (Z S ) is the second cross transfer coefficient, is a convolution operation;

[0048] The sub-resolution assist feature patterning exposure model calculates a main feature image plane photoresist image corresponding to the main feature image plane light intensity distribution image by:

[0049] ;

[0050] wherein R M i (x,y) is the main feature image plane photoresist image, F j is the jth photoresist model term, c j is a linear coefficient corresponding to the jth photoresist model term, I M i (x,y) is the main feature image plane light intensity distribution image, T M is the first threshold value;

[0051] The sub-resolution assist feature patterning exposure model calculates a sub-resolution assist feature image plane photoresist image corresponding to the sub-resolution assist feature image plane light intensity distribution image by:

[0052] ;

[0053] wherein R S i (x,y) is the sub-resolution assist feature image plane photoresist image, F j is the jth photoresist model term, c j is a linear coefficient corresponding to the jth photoresist model term, I S i (x,y) is the sub-resolution assist feature image plane light intensity distribution image, T S is the second threshold value.

[0054] Optionally, in the light source mask co-optimization method, the first cross- transfer coefficient, the second cross-transfer coefficient, the first threshold value, the second threshold value, the photoresist model term and the corresponding linear coefficient are calibrated by using the main pattern critical dimension corresponding to each of the illumination light sources and the sub-resolution assist pattern printing information of the modeling pattern set, comprising:

[0055] The calibration of the first cross-transfer coefficient, the second cross-transfer coefficient, the first threshold value, the second threshold value, the photoresist model term and the corresponding linear coefficient is completed when the function value of the correction loss function in the following formula is the minimum:

[0056] ;

[0057] Wherein, Cost is the function value of the correction loss function, is the difference between the main pattern critical dimension corresponding to the kth main pattern in the modeling pattern set in the collected photoresist image and the main pattern critical dimension corresponding to the main pattern in the main pattern image plane photoresist image, w k is the pattern weight coefficient of the kth main pattern in the modeling pattern set, is the printing difference of the nth sub-resolution assist pattern in the modeling pattern set, w n is the pattern weight coefficient of the nth sub-resolution assist pattern in the modeling pattern set;

[0058] When the printing of the sub-resolution assist pattern of the collected photoresist image is the same as the printing of the simulation result of the sub-resolution assist pattern exposure model:

[0059] ;

[0060] When the printing of the sub-resolution assist pattern of the collected photoresist image is different from the printing of the simulation result of the sub-resolution assist pattern exposure model:

[0061] ;

[0062] Wherein, T S is the second threshold value, R S max is the maximum image value of the corresponding sub-resolution assist pattern in the sub-resolution assist pattern image plane photoresist image.

[0063] A light source mask co-optimization device, comprising:

[0064] A receiving module for receiving a lithography target pattern;

[0065] a gridding module configured to grid the lithography target pattern to obtain a lithography target image;

[0066] a binarization module configured to obtain an initial light source and an initial mask pattern according to the lithography target image; the initial mask pattern is a binarized image;

[0067] a first iteration module configured to perform iterative optimization on the initial light source and the initial mask pattern based on the lithography target image by using a preset light source and mask collaborative optimization loss function, and determine an adjusted initial light source and an adjusted initial mask pattern corresponding to a minimum function value of the light source and mask collaborative optimization loss function as an optimized light source and an optimized mask pattern; the light source and mask collaborative optimization loss function comprises a main pattern edge error term and a sub-resolution auxiliary pattern printing term;

[0068] a first window module configured to calculate a first process window according to the optimized light source and the optimized mask pattern;

[0069] a second iteration module configured to perform simulation on the optimized mask pattern under the optimized light source to obtain a simulated lithography image, and determine whether there is a sub-resolution auxiliary pattern printed in the simulated lithography image; when there is a printed sub-resolution auxiliary pattern, the size of the corresponding pattern in the optimized mask pattern is reduced to obtain an updated optimized mask pattern, and the updated optimized mask pattern is simulated again until there is no sub-resolution auxiliary pattern printed in the simulated lithography image corresponding to the optimized mask pattern, and the latest optimized mask pattern is output as a process iteration mask pattern;

[0070] a second window module configured to calculate a second process window according to the optimized light source and the process iteration mask pattern;

[0071] a ratio determination module configured to calculate a ratio value of a difference between the first process window and the second process window to the first process window, and determine whether the ratio value exceeds a preset upper limit value of qualification;

[0072] an output module configured to output the optimized light source and the process iteration mask pattern when the ratio value does not exceed the upper limit value of qualification.

[0073] The method comprises the steps of: receiving a photolithography target pattern; rasterizing the photolithography target pattern to obtain a photolithography target image; obtaining an initial light source and an initial mask pattern according to the photolithography target image; the initial mask pattern is a binary image; based on the photolithography target image, an initial light source and an initial mask pattern are iteratively optimized by using a preset light source and mask cooperative optimization loss function, and the adjusted initial light source and the adjusted initial mask pattern corresponding to the minimum function value of the light source and mask cooperative optimization loss function are the optimized light source and the optimized mask pattern; the light source and mask cooperative optimization loss function comprises a main pattern edge error term and a sub-resolution auxiliary pattern printing term; a first process window is calculated according to the optimized light source and the optimized mask pattern; under the optimized light source, the optimized mask pattern is simulated to obtain a simulation lithography image, and it is judged whether there is a sub-resolution auxiliary pattern printing in the simulation lithography image; when there is a printed sub-resolution auxiliary pattern, the size of the corresponding pattern in the optimized mask pattern is reduced, an updated optimized mask pattern is obtained, and the updated optimized mask pattern is simulated again until there is no sub-resolution auxiliary pattern printing in the simulation lithography image corresponding to the optimized mask pattern, and the latest optimized mask pattern is output as a process iteration mask pattern; a second process window is calculated according to the optimized light source and the process iteration mask pattern; the difference between the first process window and the second process window is calculated, and the proportion of the difference in the first process window is calculated; and it is judged whether the proportion value exceeds a preset upper limit value; when the proportion value does not exceed the upper limit value, the optimized light source and the process iteration mask pattern are output.

[0074] The method comprises the steps of: receiving a photolithography target pattern; rasterizing the photolithography target pattern to obtain a photolithography target image; obtaining an initial light source and an initial mask pattern according to the photolithography target image; the initial mask pattern is a binary image; based on the photolithography target image, an initial light source and an initial mask pattern are iteratively optimized by using a preset light source and mask cooperative optimization loss function, and the adjusted initial light source and the adjusted initial mask pattern corresponding to the minimum function value of the light source and mask cooperative optimization loss function are the optimized light source and the optimized mask pattern; the light source and mask cooperative optimization loss function comprises a main pattern edge error term and a sub-resolution auxiliary pattern printing term; a first process window is calculated according to the optimized light source and the optimized mask pattern; under the optimized light source, the optimized mask pattern is simulated to obtain a simulation lithography image, and it is judged whether there is a sub-resolution auxiliary pattern printing in the simulation lithography image; when there is a printed sub-resolution auxiliary pattern, the size of the corresponding pattern in the optimized mask pattern is reduced, an updated optimized mask pattern is obtained, and the updated optimized mask pattern is simulated again until there is no sub-resolution auxiliary pattern printing in the simulation lithography image corresponding to the optimized mask pattern, and the latest optimized mask pattern is output as a process iteration mask pattern; a second process window is calculated according to the optimized light source and the process iteration mask pattern; the difference between the first process window and the second process window is calculated, and the proportion of the difference in the first process window is calculated; and it is judged whether the proportion value exceeds a preset upper limit value; when the proportion value does not exceed the upper limit value, the optimized light source and the process iteration mask pattern are output. BRIEF DESCRIPTION OF DRAWINGS

[0075] In order to make the technical scheme of the present application or prior art clearer, the accompanying drawings needed in the description of the embodiments or prior art will be briefly introduced. Obviously, the accompanying drawings in the description below only show some embodiments of the present application, and all other embodiments obtained by a person of ordinary skill in the art without creative effort based on the accompanying drawings are within the protection scope of the present application.

[0076] Figure 1 A flowchart of a specific embodiment of the light source mask collaborative optimization method provided by the present application is shown in the figure.

[0077] Figure 2 A flowchart of a specific embodiment of the light source mask collaborative optimization method provided by the present application is shown in the figure.

[0078] Figure 3 A process structure diagram of a specific embodiment of the light source mask collaborative optimization method provided by the present application is shown in the figure.

[0079] Figure 4 A structure diagram of a specific embodiment of the light source mask collaborative optimization device provided by the present application is shown in the figure.

[0080] Reference signs:

[0081] 100-receiving module; 200-rasterizing module; 300-binarizing module; 400-first iteration module; 500-first window module; 600-second iteration module; 700-second window module; 800-ratio judging module; 900-output module. DETAILED DESCRIPTION

[0082] In order to make the technical scheme of the present application or prior art clearer, the accompanying drawings needed in the description of the embodiments or prior art will be briefly introduced. Obviously, the accompanying drawings in the description below only show some embodiments of the present application, and all other embodiments obtained by a person of ordinary skill in the art without creative effort based on the accompanying drawings are within the protection scope of the present application.

[0083] The core of the present application is to provide a light source mask collaborative optimization method, a flowchart of a specific embodiment of which is shown in the figure. Figure 1 as embodiment one, which includes:

[0084] S101: receiving a lithography target pattern.

[0085] The lithography target pattern is the pattern needed to be obtained on the wafer, which can only be the contour of the pattern or a complete image with image values, and the present application does not make any limitation here.

[0086] S102: rasterize the lithography target pattern to obtain a lithography target image.

[0087] Rasterizing the lithography target pattern refers to dividing the lithography target pattern in a matrix plane by a grid with a preset size, so as to obtain the lithography target image in units of grid.

[0088] S103: obtaining an initial light source and an initial mask pattern according to the lithography target image; the initial mask pattern is a binary image.

[0089] The initial light source and the initial mask pattern can be a preset light source and mask pattern fixed according to the experience of workers, or a light source and mask pattern calculated according to the lithography target image.

[0090] As a preferred embodiment, obtaining an initial light source and an initial mask pattern according to the lithography target image comprises:

[0091] A1: obtaining an initial light source and an initial mask image by optimizing an initial loss function of formula (1) based on the lithography target image:

[0092] ; (1)

[0093] wherein, w’’ p is a weight coefficient of a process condition p corresponding to the initial light source, w’’ x is a weight coefficient of an evaluation point x on the lithography target pattern corresponding to the initial mask image, R M i (p) is an image value of a grid point i of a simulated post-etching image of the initial mask image under a process condition p, O i is an image value of a grid point i of the lithography target image, and CF3 is a function value of the initial loss function.

[0094] A2: binaryzation of the initial mask image to obtain an initial mask pattern.

[0095] Since the initial mask pattern is obtained by binaryzation, it must be a polygon mask pattern.

[0096] The initial loss function in the embodiment assigns different weights to each grid point, so that the difference between the image value obtained by simulation of each grid point and the target image value (i.e. the value of the corresponding grid point on the lithography target image) needs to be multiplied by the weight coefficient corresponding to the grid point, and the weight coefficient corresponding to each process condition is also considered. The initial light source and the initial mask image obtained by formula (1) are close to the optimized light source and the process-iterated mask pattern, which can reduce the number of subsequent iterations and improve the universality of the application, so that the application can adapt to various use scenarios.

[0097] S104: Based on the lithography target image, the initial light source and the initial mask pattern are iteratively optimized by using a preset light source-mask collaborative optimization loss function, and the adjusted initial light source and the adjusted initial mask pattern corresponding to the minimum function value of the light source-mask collaborative optimization loss function are determined as the optimized light source and the optimized mask pattern; the light source-mask collaborative optimization loss function includes a main pattern edge error term and a sub-resolution auxiliary pattern printing term.

[0098] In the prior art, the initial light source and the initial mask pattern also need to be iteratively optimized, but the iteration in the prior art only considers the edge error of the main pattern, and does not consider whether the printing of the sub-resolution auxiliary pattern meets the design expectation. The light source-mask collaborative optimization loss function provided in the application simultaneously considers the edge error of the main pattern (corresponding to the main pattern edge error term) and whether the printing of the sub-resolution auxiliary pattern meets the design expectation (corresponding to the sub-resolution auxiliary pattern printing term).

[0099] In each iteration of the iterative optimization in this step, the mask pattern obtained in the last iteration is simulated under the light source condition obtained in the last iteration to obtain a corresponding simulated lithography simulation result image, and then the lithography simulation result image is compared with the lithography target image, and the mask pattern obtained in the last iteration is adjusted according to the comparison result to obtain the mask pattern after this iteration.

[0100] S105: Calculate a first process window according to the optimized light source and the optimized mask pattern.

[0101] The first process window is a process window corresponding to the combination of the light source and the mask pattern considering the printing of the sub-resolution auxiliary pattern, which is obtained according to the computer simulation result.

[0102] S106: Simulate the optimization mask pattern under the optimization light source to obtain a simulation lithography image, and determine whether there is a sub-resolution auxiliary pattern printed in the simulation lithography image; when there is a printed sub-resolution auxiliary pattern, reduce the size of the corresponding pattern in the optimization mask pattern to obtain an updated optimization mask pattern, and then simulate the updated optimization mask pattern until there is no sub-resolution auxiliary pattern printed in the simulation lithography image corresponding to the optimization mask pattern, and output the latest optimization mask pattern as a process iteration mask pattern.

[0103] In this step, the optimization mask pattern is repeatedly simulated multiple times. After each simulation, as long as the simulation lithography image obtained still has the sub-resolution auxiliary pattern printed, the optimization mask pattern is adjusted and simulated again. It should be noted that each adjustment only reduces the size of the printed sub-resolution auxiliary pattern, and other sub-resolution auxiliary patterns are not changed. The single size reduction of the printed sub-resolution auxiliary pattern can be a preset fixed value, or can be determined according to the size of the printed pattern of the sub-resolution auxiliary pattern on the simulation lithography image. The present application does not limit this.

[0104] Further, reducing the size of the corresponding pattern of the printed sub-resolution auxiliary pattern in the optimization mask pattern to obtain an updated optimization mask pattern comprises:

[0105] B1: Reduce the size of the corresponding pattern of the printed sub-resolution auxiliary pattern in the optimization mask pattern to obtain a to-be-adjusted mask pattern.

[0106] B2: Based on the lithography target image, the to-be-adjusted mask pattern is optimized under the optimization light source by an adjustment loss function of formula (2) to determine that the to-be-adjusted mask pattern corresponding to the minimum function value of the adjustment loss function is an updated optimization mask pattern:

[0107] ; (2)

[0108] Wherein, w' p is the weight coefficient of the process condition p corresponding to the optimization light source, w' x is the weight coefficient of the evaluation point x on the lithography target pattern corresponding to the to-be-adjusted mask pattern, EPE' p,x is the edge placement error of the corresponding position of the evaluation point x under the process condition p (the edge placement error here is the edge placement error between the simulation lithography image and the lithography target image described in the foregoing), and CF2 is the function value of the adjustment loss function.

[0109] In the preferred embodiment, when the sub-resolution auxiliary pattern is printed, not only the printed sub-resolution auxiliary pattern is reduced, but also the main pattern in the mask pattern to be adjusted is adjusted, because the size of the sub-resolution auxiliary pattern is changed, which will inevitably affect the printed edge position of the main pattern in the mask pattern to be adjusted in the photolithography simulation, and therefore further increasing the adjustment of the main pattern in the optimized mask pattern based on EPE can further reduce the edge placement error between the simulation photolithography image of the final optimized mask pattern and the photolithography target image, and improve the simulation accuracy and precision.

[0110] S107: Calculate a second process window according to the optimized light source and the process iteration mask pattern.

[0111] The second process window is a process window corresponding to the combination of the light source and the mask pattern after iteration, which is determined not to print the sub-resolution auxiliary pattern. Since there is a smaller sub-resolution auxiliary pattern in the mask pattern after iteration, the second process window is necessarily smaller than the first process window.

[0112] S108: Calculate the proportion of the difference between the first process window and the second process window to the first process window, and determine whether the proportion exceeds a preset upper limit value.

[0113] As a preferred embodiment, the upper limit value is not greater than 5%, that is, as long as the proportion of the difference between the first process window and the second process window to the first process window is within 5%, the effect of avoiding printing the sub-resolution auxiliary pattern can be achieved without greatly damaging the process window.

[0114] Further, the light source and mask collaborative optimization loss function is the sum of the main pattern edge error term and the sub-resolution auxiliary pattern printing term; the sub-resolution auxiliary pattern printing term includes a sub-resolution weight coefficient.

[0115] Correspondingly, after determining whether the proportion exceeds the preset upper limit value, it further includes:

[0116] C1: When the proportion exceeds the upper limit value, increase the sub-resolution weight coefficient to obtain an updated light source and mask collaborative optimization loss function.

[0117] The greater the sub-resolution weight coefficient is, the smaller the probability of printing the sub-resolution auxiliary pattern on the optimized mask pattern under the irradiation of the optimized light source when the function value of the light source and mask collaborative optimization loss function takes the minimum value.

[0118] C2: based on the lithography target image, using the updated light source mask collaborative optimization loss function, iteratively optimizing the initial light source and the initial mask pattern, obtaining an updated optimized light source and an updated optimized mask pattern, and calculating a corresponding first process window, a corresponding process iteration mask pattern, a corresponding second process window, and a corresponding scale value, and then determining whether the updated scale value exceeds a preset upper limit value of eligibility. When the updated scale value exceeds the upper limit value of eligibility, the sub-resolution weight coefficient is increased to obtain an updated light source mask collaborative optimization loss function, until the scale value does not exceed the upper limit value of eligibility.

[0119] In this step, the updated light source mask collaborative optimization loss function is first used to recalculate the optimized light source and the optimized mask pattern corresponding to the initial light source and the initial mask pattern. After obtaining the updated optimized light source and the updated optimized mask pattern, the steps S105 to S108 in the foregoing are performed again until the scale value does not exceed the upper limit value of eligibility.

[0120] Through this step, the sub-resolution weight coefficient corresponding to the light source mask collaborative optimization loss function when the process window loss rate (reflected in the scale value) is qualified can be calculated.

[0121] Further, based on the lithography target image, the initial light source and the initial mask pattern are iteratively optimized using a preset light source mask collaborative optimization loss function, including:

[0122] The initial light source and the initial mask pattern are iteratively optimized by the following formula (3):

[0123] ; (3)

[0124] wherein w p is the weight coefficient of the process condition p corresponding to the initial light source, w x is the weight coefficient of the evaluation point x on the lithography target pattern corresponding to the initial mask pattern, EPE p,x is the edge placement error of the evaluation point x corresponding position under the process condition p, β is the sub-resolution weight coefficient, R S j (p) is the image value of the sub-resolution auxiliary pattern printing grid point p under the process condition p, T s is the printing threshold value of the sub-resolution auxiliary pattern, and CF1 is the function value of the light source mask collaborative optimization loss function.

[0125] The edge placement error in formula (3) in the embodiment is the edge placement error at the evaluation point x of the image obtained by simulating the initial mask pattern under the initial light source compared with the lithography target image, and the evaluation point is a point arranged on the main pattern of the lithography target image. The specific arrangement position selection method can refer to the prior art, and the present application is not limited herein.

[0126] It should be noted that R S j (p) is the image value of the printed grid point p of the sub-resolution auxiliary pattern under the process condition p, that is, if the non-printed grid point in the sub-resolution auxiliary pattern is not included in this item, and since only the printed grid point is calculated, R S j The value of (p) must be greater than T s .

[0127] In the preferred embodiment, formula (3) is given, which takes the average of the edge placement error of the evaluation point position and the difference between the image value of the printed grid point of the sub-resolution auxiliary pattern and the printing threshold. This can greatly reduce the operation difficulty and improve the operation efficiency. Of course, the calculation method of the function value of the loss function of the light source mask cooperative optimization based on formula (3) can be changed, such as only taking the absolute value of EPE p,x without squaring in the main pattern edge error term, and only calculating the absolute value of the difference between the image value of the printed grid point of the sub-resolution auxiliary pattern and the printing threshold without squaring in the sub-resolution auxiliary pattern printing term. Only the positive numbers are required, and the present application will not be repeated here.

[0128] S109: When the proportion value does not exceed the qualified upper limit value, output the optimized light source and the process iteration mask pattern.

[0129] When the proportion value does not exceed the qualified upper limit value, it means that during the process of iterating from the optimized mask pattern to determine the process iteration mask pattern that does not print the sub-resolution auxiliary pattern, the lost process window is small enough. It can be considered that the light source and mask pattern combination with the largest process window is obtained under the condition of ensuring that there is no sub-resolution auxiliary pattern printing.

[0130] The application provides a light source and mask cooperative optimization method, which comprises the following steps: receiving a lithography target pattern; rasterizing the lithography target pattern to obtain a lithography target image; obtaining an initial light source and an initial mask pattern according to the lithography target image; the initial mask pattern is a binary image; based on the lithography target image, an initial light source and an initial mask pattern are iteratively optimized by using a preset light source and mask cooperative optimization loss function, and the adjusted initial light source and the adjusted initial mask pattern corresponding to the minimum function value of the light source and mask cooperative optimization loss function are the optimized light source and the optimized mask pattern; the light source and mask cooperative optimization loss function comprises a main pattern edge error term and a sub-resolution auxiliary pattern printing term; a first process window is calculated according to the optimized light source and the optimized mask pattern; the optimized mask pattern is simulated under the optimized light source to obtain a simulation lithography image, and it is judged whether there is a sub-resolution auxiliary pattern printing in the simulation lithography image; when there is a printed sub-resolution auxiliary pattern, the size of the corresponding pattern in the optimized mask pattern is reduced, an updated optimized mask pattern is obtained, and the updated optimized mask pattern is simulated again until there is no sub-resolution auxiliary pattern printing in the simulation lithography image corresponding to the optimized mask pattern, and the latest optimized mask pattern is output as a process iteration mask pattern; a second process window is calculated according to the optimized light source and the process iteration mask pattern; the difference between the first process window and the second process window is calculated, and the proportion of the difference in the first process window is calculated; it is judged whether the proportion value exceeds a preset upper limit value; when the proportion value does not exceed the upper limit value, the optimized light source and the process iteration mask pattern are output. The light source and mask cooperative optimization loss function comprising the sub-resolution auxiliary pattern printing term is used to iteratively optimize the initial light source and the initial mask pattern, that is, the risk of sub-resolution auxiliary pattern printing is accurately added to the loss function, and then the size of the sub-resolution auxiliary pattern in the optimized mask pattern is further selectively reduced, and the loss of the process window in the size reduction process of the sub-resolution auxiliary pattern is limited by the upper limit value, so that the optimal combination of the light source and the mask pattern is obtained under the condition that there is no sub-resolution auxiliary pattern printing, and the precision and practicability of the light source and mask cooperative optimization are greatly improved.

[0131] The application further provides a training method of a sub-resolution auxiliary pattern exposure model, and a corresponding flowchart is shown in FIG. Figure 2 The simulation in the iterative optimization of the initial light source and the initial mask pattern and / or the simulation of the optimized mask pattern can be performed by using the sub-resolution auxiliary pattern exposure model trained by the training method of the sub-resolution auxiliary pattern exposure model.

[0132] As a specific embodiment, the training method of the sub-resolution auxiliary pattern exposure model comprises the following steps:

[0133] S201: establishing an uncalibrated sub-resolution auxiliary pattern exposure model; the sub-resolution auxiliary pattern exposure model can first determine a main pattern image plane light intensity distribution image according to an inputted to-be-processed mask pattern through a preset first cross transfer coefficient, then calculate a main pattern image plane photoresist image corresponding to the main pattern image plane light intensity distribution image through a preset photoresist model term and a corresponding linear coefficient, and finally determine a main pattern critical dimension from the main pattern image plane photoresist image through a first threshold value; the sub-resolution auxiliary pattern exposure model can first determine a sub-resolution auxiliary pattern image plane light intensity distribution image according to the inputted to-be-processed mask pattern through a preset second cross transfer coefficient, then calculate a sub-resolution auxiliary pattern image plane photoresist image corresponding to the sub-resolution auxiliary pattern image plane light intensity distribution image through the photoresist model term and the linear coefficient, and finally determine a sub-resolution auxiliary pattern printing information from the sub-resolution auxiliary pattern image plane photoresist image through a second threshold value.

[0134] The sub-resolution auxiliary pattern exposure model can calculate light intensity distribution images of two planes, one of which is a photoresist plane corresponding to the main pattern, and the other of which is a photoresist plane corresponding to the sub-resolution auxiliary pattern. Therefore, after obtaining the two light intensity distribution images, the corresponding photoresist images are calculated, the simulated post-lithography profile is obtained by using the main pattern image plane photoresist image and the corresponding first threshold value, and the critical dimension of the main pattern, i.e., the main pattern critical dimension, is obtained naturally from the profile. As for the sub-resolution auxiliary pattern image plane photoresist image, the corresponding simulated post-lithography profile can also be obtained through the second threshold value. Of course, the profile of the sub-resolution auxiliary pattern printing is not very helpful, and the sub-resolution auxiliary pattern is mainly used to determine whether the sub-resolution auxiliary pattern is printed, i.e., the sub-resolution auxiliary pattern printing information, which includes not only the "yes" / "no" state information representing the printing state, but also the image value corresponding to the printed sub-resolution auxiliary pattern.

[0135] As a specific embodiment, the sub-resolution auxiliary pattern exposure model determines the main pattern image plane light intensity distribution image through the following formula (4):

[0136] ; (4)

[0137] wherein, I M i (x, y) is the main pattern image plane light intensity distribution image, M(x, y) is a rasterized mask image corresponding to the to-be-processed mask pattern, TCCi (Z M ) represents the first cross-transfer coefficient. This is a convolution operation.

[0138] The sub-resolution assisted graphic exposure model determines the light intensity distribution image of the sub-resolution assisted graphic imaging plane using the following formula (5):

[0139] ; (5)

[0140] Among them, I S i (x,y) represents the light intensity distribution image of the sub-resolution auxiliary image imaging plane, M(x,y) represents the rasterized mask image corresponding to the mask pattern to be processed, and TCC i (Z S ) is the second cross-transfer coefficient. This is a convolution operation.

[0141] The rasterized mask image is the image obtained by rasterizing the mask pattern to be processed according to a preset grid size. The above two equations calculate the light intensity distribution images corresponding to the two planes at different height positions of the photoresist using different cross-transfer coefficients. (Refer to...) Figure 3 , Figure 3 Z in Chinese M Z represents the position of the photoresist plane corresponding to the light intensity distribution image of the main pattern imaging plane. S This indicates the position of the photoresist plane corresponding to the sub-resolution auxiliary graphic imaging plane light intensity distribution image.

[0142] The sub-resolution assisted pattern exposure model calculates the main pattern imaging plane photoresist image corresponding to the main pattern imaging plane light intensity distribution image using the following formula (6):

[0143] ; (6)

[0144] Among them, R M i (x,y) represents the main graphic imaging planar photoresist image, F j For the photoresist model term described in the j-th term, c j I represents the linear coefficient corresponding to the photoresist model term described in the j-th term. M i (x, y) represents the light intensity distribution image of the main graphic imaging plane, T M The first threshold;

[0145] The sub-resolution auxiliary pattern exposure model calculates the sub-resolution auxiliary pattern image plane photoresist image corresponding to the sub-resolution auxiliary pattern image plane light intensity distribution image by the following formula (7):

[0146] (7)

[0147] wherein, R S i (x,y) is the sub-resolution auxiliary pattern image plane photoresist image, F j is the jth photoresist model term, c j is the linear coefficient corresponding to the jth photoresist model term, I S i (x,y) is the sub-resolution auxiliary pattern image plane light intensity distribution image, T S is the second threshold value.

[0148] The photoresist model terms and the corresponding linear coefficients of each term in formula (6) and formula (7) are the same, and the rest can be referred to the foregoing, which will not be expanded here.

[0149] S202: According to the preset modeling pattern set, a test mask is made, and the corresponding collected photoresist images of the test mask under different illumination light sources are obtained; the modeling pattern set includes a plurality of main patterns and sub-resolution auxiliary patterns of different sizes.

[0150] The modeling pattern set includes main patterns and sub-resolution auxiliary patterns of different geometric shapes and sizes, and in this step, the test mask is placed under a plurality of different illumination light sources, and the corresponding photoresist images are collected.

[0151] S203: Determine the main pattern critical dimension and sub-resolution auxiliary pattern printing information corresponding to each of the collected photoresist images.

[0152] In the photoresist images corresponding to each of the illumination light sources, the critical dimensions of various main patterns and the printing conditions of various sub-resolution auxiliary patterns corresponding to each of the illumination light sources are collected.

[0153] S204: calibrate the first cross-over transfer coefficient, the second cross-over transfer coefficient, the first threshold value, the second threshold value, the photoresist model term and the corresponding linear coefficient by using the main pattern critical dimension and the sub-resolution assist pattern print-out information corresponding to each of the illumination sources of the modeling pattern set, so that the difference between the main pattern critical dimension and the sub-resolution assist pattern print-out information of the modeling pattern set as a to-be-processed mask pattern input into the sub-resolution assist pattern exposure model and the main pattern critical dimension and the sub-resolution assist pattern print-out information of the collected photoresist image under the corresponding illumination source is less than a preset calibration threshold value.

[0154] Further, the step comprises:

[0155] When the function value of the correction loss function in the following formula (8), formula (9), formula (10) is the minimum, the calibration of the first cross-over transfer coefficient, the second cross-over transfer coefficient, the first threshold value, the second threshold value, the photoresist model term and the corresponding linear coefficient is completed:

[0156] ; (8)

[0157] Wherein, Cost is the function value of the correction loss function, is the difference between the main pattern critical dimension corresponding to the kth main pattern in the modeling pattern set in the collected photoresist image and the main pattern critical dimension corresponding to the kth main pattern in the main pattern image plane photoresist image, w k is the pattern weight coefficient of the kth main pattern in the modeling pattern set, is the print-out difference of the nth sub-resolution assist pattern in the modeling pattern set, w n is the pattern weight coefficient of the nth sub-resolution assist pattern in the modeling pattern set;

[0158] When the sub-resolution assist pattern print-out condition of the collected photoresist image is the same as the print-out condition of the simulation result of the sub-resolution assist pattern exposure model:

[0159] ; (9)

[0160] When the sub-resolution assist pattern print-out condition of the collected photoresist image is different from the print-out condition of the simulation result of the sub-resolution assist pattern exposure model:

[0161] ; (10)

[0162] Wherein, T S is the second threshold value, R Smax corresponding sub-resolution assist feature in the sub-resolution assist feature image plane photoresist image.

[0163] In the specific embodiment, the loss function for calibrating the above parameters is further defined, which is specifically divided into two categories, one is the difference of the critical dimension of the main pattern, and the other is the printing difference of the sub-resolution assist feature, and is supplemented by the pattern weight coefficient of each pattern weight coefficient, and the sum of the weight of the total height difference is taken as the loss value of the function, which is intuitive and simple, that is, the smaller the total difference between the simulation result and the actual result, the more accurate the correction of the cross transfer coefficient, the photoresist height model term and the linear coefficient, and it is also closer to the actual situation, of course, other ways can also be used to correct the above parameters, which is not limited in the present application.

[0164] The light source mask cooperative optimization device provided by the embodiment of the present application will be introduced below, and the light source mask cooperative optimization device described below can be correspondingly referred to the light source mask cooperative optimization method described above.

[0165] Figure 4 The structure block diagram of the light source mask cooperative optimization device provided by the embodiment of the present application is referred to Figure 4 The light source mask cooperative optimization device can include:

[0166] The receiving module 100 is configured to receive a lithography target pattern.

[0167] The gridding module 200 is configured to grid the lithography target pattern to obtain a lithography target image.

[0168] The binarization module 300 is configured to obtain an initial light source and an initial mask pattern according to the lithography target image; the initial mask pattern is a binarized image.

[0169] The first iteration module 400 is configured to perform iterative optimization on the initial light source and the initial mask pattern based on the lithography target image by using a preset light source mask cooperative optimization loss function, and determine the adjusted initial light source and the adjusted initial mask pattern corresponding to the minimum function value of the light source mask cooperative optimization loss function as an optimized light source and an optimized mask pattern; the light source mask cooperative optimization loss function includes a main pattern edge error term and a sub-resolution assist feature printing term.

[0170] The first window module 500 is configured to calculate a first process window according to the optimized light source and the optimized mask pattern.

[0171] The second iteration module 600 is configured to simulate the optimized mask pattern under the optimized light source to obtain a simulation lithography image, and determine whether there is a sub-resolution auxiliary pattern printed in the simulation lithography image; when there is a printed sub-resolution auxiliary pattern, the size of the printed sub-resolution auxiliary pattern in the corresponding pattern in the optimized mask pattern is reduced to obtain an updated optimized mask pattern, and the updated optimized mask pattern is simulated again until there is no sub-resolution auxiliary pattern printed in the simulation lithography image corresponding to the optimized mask pattern, and the latest optimized mask pattern is output as a process iteration mask pattern.

[0172] The second window module 700 is configured to calculate a second process window according to the optimized light source and the process iteration mask pattern.

[0173] The ratio determining module 800 is configured to calculate a ratio value of a difference between the first process window and the second process window to the first process window, and determine whether the ratio value exceeds a preset upper limit value of qualification.

[0174] The output module 900 is configured to output the optimized light source and the process iteration mask pattern when the ratio value does not exceed the upper limit value of qualification.

[0175] As a specific embodiment, the light source and mask collaborative optimization loss function is a sum of the main pattern edge error term and the sub-resolution auxiliary pattern printing term; the sub-resolution auxiliary pattern printing term includes a sub-resolution weight coefficient.

[0176] The ratio determining module 800 further includes:

[0177] The weight increasing unit is configured to increase the sub-resolution weight coefficient to obtain an updated light source and mask collaborative optimization loss function when the ratio value exceeds the upper limit value of qualification.

[0178] The weight recalculation unit is configured to perform iterative optimization on the initial light source and the initial mask pattern based on the lithography target image by using the updated light source and mask collaborative optimization loss function to obtain an updated optimized light source and an updated optimized mask pattern, and calculate a corresponding first process window, a corresponding process iteration mask pattern, a corresponding second process window, and a corresponding ratio value, and then determine whether the updated ratio value exceeds the upper limit value of qualification; when the updated ratio value exceeds the upper limit value of qualification, the sub-resolution weight coefficient is increased to obtain an updated light source and mask collaborative optimization loss function, until the ratio value does not exceed the upper limit value of qualification.

[0179] As a specific embodiment, the first iteration module 400 includes:

[0180] An initial loss calculation unit is used to iteratively optimize the initial light source and the initial mask pattern using the following formula:

[0181] ;

[0182] Among them, w p w is the weighting coefficient of the process condition p corresponding to the initial light source. x EPE is the weighting coefficient of the initial mask pattern corresponding to the evaluation point x on the photolithographic target pattern. p,x The edge placement error of the evaluation point x under process condition p is β, where β is the sub-resolution weighting coefficient, and R is the edge placement error of the evaluation point x under process condition p. S j (p) represents the image value of the sub-resolution auxiliary graphic printing grid point p under process condition p, T s The printing threshold for the sub-resolution auxiliary graphic.

[0183] In one specific implementation, the second iteration module 600 includes:

[0184] The size of the corresponding graphic in the optimized mask pattern is reduced by shrinking the size of the sub-resolution auxiliary graphic to be printed, to obtain the mask pattern to be adjusted.

[0185] Based on the photolithographic target image, under the optimized light source, the mask pattern to be adjusted is optimized using the following adjustment loss function, and the mask pattern to be adjusted corresponding to the minimum function value of the adjustment loss function is determined as the updated optimized mask pattern:

[0186] ;

[0187] Among them, w' p w' is the weighting coefficient of the process condition p corresponding to the optimized light source. x EPE' is the weighting coefficient of the mask pattern to be adjusted corresponding to the evaluation point x on the photolithographic target pattern. p,x The edge placement error of the position corresponding to the evaluation point x under process condition p.

[0188] As one specific implementation, the binarization module 300 includes:

[0189] The initial loss calculation unit is used to optimize the initial light source and initial mask image based on the lithographic target image using the initial loss function of the following formula:

[0190] ;

[0191] Among them, w'' p w'' is the weighting coefficient of the process condition p corresponding to the initial light source.x is a weight coefficient of the initial mask image corresponding to an evaluation point x on the lithography target pattern, R M i (p) is an image value of a simulation post-lithography image of the initial mask image at grid point i under process condition p, O i is an image value of grid point i of the lithography target image;

[0192] a binary assignment unit configured to binarize the initial mask image to obtain an initial mask pattern.

[0193] As a specific embodiment, the simulation in the iterative optimization of the initial light source and the initial mask pattern, and / or the simulation of the optimized mask pattern, is performed by a pre-trained sub-resolution assist feature exposure model;

[0194] The training device of the sub-resolution assist feature exposure model comprises:

[0195] a modeling module configured to establish an uncalibrated sub-resolution assist feature exposure model; the sub-resolution assist feature exposure model is capable of determining a main pattern image plane light intensity distribution image according to an input to-be-processed mask pattern through a preset first cross transfer coefficient, calculating a main pattern image plane photoresist image corresponding to the main pattern image plane light intensity distribution image through a preset photoresist model term and a corresponding linear coefficient, and determining a main pattern critical dimension from the main pattern image plane photoresist image through a first threshold value; the sub-resolution assist feature exposure model is capable of determining a sub-resolution assist feature image plane light intensity distribution image according to an input to-be-processed mask pattern through a preset second cross transfer coefficient, calculating a sub-resolution assist feature image plane photoresist image corresponding to the sub-resolution assist feature image plane light intensity distribution image through the photoresist model term and the linear coefficient, and determining sub-resolution assist feature printing information from the sub-resolution assist feature image plane photoresist image through a second threshold value;

[0196] a test acquisition module configured to fabricate a test mask according to a preset modeling pattern set, and acquire an acquired photoresist image corresponding to the test mask under different illumination light sources; the modeling pattern set comprises main patterns and sub-resolution assist features of multiple different sizes;

[0197] a test parameter determination module configured to determine main pattern critical dimensions and sub-resolution assist feature printing information corresponding to each of the acquired photoresist images;

[0198] a calibration module, configured to calibrate the first cross-talk coefficient, the second cross-talk coefficient, the first threshold value, the second threshold value, the photoresist model term and the corresponding linear coefficient by using the main pattern critical dimension and the sub-resolution assist pattern printing information corresponding to each of the illumination sources and the modeling pattern set, so that a difference between the main pattern critical dimension and the sub-resolution assist pattern printing information obtained by inputting the modeling pattern set as a to-be-processed mask pattern into the sub-resolution assist pattern exposure model and the main pattern critical dimension and the sub-resolution assist pattern printing information of the collected photoresist image under the corresponding illumination source is less than a preset calibration threshold value.

[0199] As a specific embodiment, the modeling module comprises:

[0200] a main pattern image plane light intensity distribution unit, configured to make the sub-resolution assist pattern exposure model determine a main pattern image plane light intensity distribution image according to the following formula:

[0201] ;

[0202] wherein, I M i (x, y) is the main pattern image plane light intensity distribution image, M(x, y) is a rasterized mask image corresponding to the to-be-processed mask pattern, TCC i (Z M ) is the first cross-talk coefficient, is a convolution operation;

[0203] a sub-resolution assist pattern image plane light intensity distribution unit, configured to make the sub-resolution assist pattern exposure model determine a sub-resolution assist pattern image plane light intensity distribution image according to the following formula:

[0204] ;

[0205] wherein, I S i (x, y) is the sub-resolution assist pattern image plane light intensity distribution image, M(x, y) is a rasterized mask image corresponding to the to-be-processed mask pattern, TCC i (Z S ) is the second cross-talk coefficient, is a convolution operation;

[0206] a main pattern image plane photoresist unit, configured to make the sub-resolution assist pattern exposure model calculate a main pattern image plane photoresist image corresponding to the main pattern image plane light intensity distribution image according to the following formula:

[0207] ;

[0208] wherein R M i (x,y) is the main pattern formed image plane photoresist image, F j is the jth photoresist model term, c j is the linear coefficient corresponding to the jth photoresist model term, I M i (x,y) is the main pattern formed image plane light intensity distribution image, T M is the first threshold value;

[0209] The sub-resolution assist pattern formed image plane photoresist unit is configured to make the sub-resolution assist pattern exposure model calculate the sub-resolution assist pattern formed image plane photoresist image corresponding to the sub-resolution assist pattern formed image plane light intensity distribution image by the following formula:

[0210] ;

[0211] wherein R S i (x,y) is the sub-resolution assist pattern formed image plane photoresist image, F j is the jth photoresist model term, c j is the linear coefficient corresponding to the jth photoresist model term, I S i (x,y) is the sub-resolution assist pattern formed image plane light intensity distribution image, T S is the second threshold value.

[0212] As a specific embodiment, the calibration module comprises:

[0213] The correction loss calculation unit is configured to complete calibration of the first cross transfer coefficient, the second cross transfer coefficient, the first threshold value, the second threshold value, the photoresist model term and the corresponding linear coefficient when the function value of the correction loss function in the following formula is the minimum:

[0214] ;

[0215] wherein Cost is the function value of the correction loss function, is the difference between the main pattern critical dimension corresponding to the kth main pattern in the modeling pattern set in the collected photoresist image and the main pattern critical dimension corresponding to the kth main pattern in the main pattern formed image plane photoresist image, w k is the pattern weight coefficient of the kth main pattern in the modeling pattern set, is the printing difference of the nth sub-resolution assist pattern in the modeling pattern set, wn a pattern weight coefficient of an nth sub-resolution assist pattern in the set of modeled patterns;

[0216] when the printed-out condition of the sub-resolution assist pattern in the collected photoresist image is the same as the printed-out condition of the simulation result of the sub-resolution assist pattern exposure model:

[0217] ;

[0218] when the printed-out condition of the sub-resolution assist pattern in the collected photoresist image is different from the printed-out condition of the simulation result of the sub-resolution assist pattern exposure model:

[0219] ;

[0220] wherein T S is the second threshold value, R S max is the maximum image value of the corresponding sub-resolution assist pattern in the photoresist image plane of the sub-resolution assist pattern.

[0221] The application provides a light source and mask cooperative optimization device, which comprises a receiving module 100, a griding module 200, a binaryzation module 300, a first iteration module 400, a first window module 500, a second iteration module 600, a second window module 700, a ratio judgment module 800 and an output module 900. The receiving module 100 is used for receiving a lithography target pattern. The griding module 200 is used for griding the lithography target pattern to obtain a lithography target image. The binaryzation module 300 is used for obtaining an initial light source and an initial mask pattern according to the lithography target image. The initial mask pattern is a binaryzated image. The first iteration module 400 is used for iteratively optimizing the initial light source and the initial mask pattern based on the lithography target image by using a preset light source and mask cooperative optimization loss function, and determining that the adjusted initial light source and the adjusted initial mask pattern corresponding to the minimum function value of the light source and mask cooperative optimization loss function are an optimized light source and an optimized mask pattern. The light source and mask cooperative optimization loss function comprises a main pattern edge error term and a sub-resolution auxiliary pattern printing term. The first window module 500 is used for calculating a first process window according to the optimized light source and the optimized mask pattern. The second iteration module 600 is used for simulating the optimized mask pattern under the optimized light source to obtain a simulation lithography image, and judging whether there is a sub-resolution auxiliary pattern printing in the simulation lithography image. When there is a printed sub-resolution auxiliary pattern, the size of the corresponding pattern in the optimized mask pattern is reduced to obtain an updated optimized mask pattern, and the updated optimized mask pattern is simulated again until there is no sub-resolution auxiliary pattern printing in the simulation lithography image corresponding to the optimized mask pattern, and the latest optimized mask pattern is output as a process iteration mask pattern. The second window module 700 is used for calculating a second process window according to the optimized light source and the process iteration mask pattern. The ratio judgment module 800 is used for calculating the difference value of the first process window and the second process window accounts for the proportion of the first process window, and judging whether the proportion value exceeds a preset qualified upper limit value. The output module 900 is used for outputting the optimized light source and the process iteration mask pattern when the proportion value does not exceed the qualified upper limit value. The application uses the light source and mask cooperative optimization loss function comprising the sub-resolution auxiliary pattern printing term to iteratively optimize the initial light source and the initial mask pattern, that is, the risk of sub-resolution auxiliary pattern printing is accurately added to the loss function, and then the size of the sub-resolution auxiliary pattern in the optimized mask pattern is further selectively reduced, and the qualified upper limit value is used to limit the loss of the process window in the size reduction process of the sub-resolution auxiliary pattern, so that the optimal light source and mask pattern combination is obtained under the condition that there is no sub-resolution auxiliary pattern printing, and the precision and practicability of the light source and mask cooperative optimization are greatly improved.

[0222] The light source mask cooperative optimization device of the embodiment is used to realize the light source mask cooperative optimization method described above, and therefore the specific embodiments in the light source mask cooperative optimization device can be seen from the embodiment part of the light source mask cooperative optimization method in the foregoing, for example, the receiving module 100, the rasterization module 200, the binarization module 300, the first iteration module 400, the first window module 500, the second iteration module 600, the second window module 700, the ratio judgment module 800, and the output module 900 are respectively used to realize steps S101, S102, S103, S104, S105, S106, S107, S108, and S109 in the light source mask cooperative optimization method described above, and therefore the specific embodiments can be referred to the description of the respective embodiment parts, and details are not described herein again.

[0223] The application further provides a light source mask cooperative optimization device, comprising:

[0224] a memory for storing a computer program;

[0225] Processor, for executing the computer program to realize the steps of the light source mask collaborative optimization method. The light source mask collaborative optimization method provided by the application receives a lithography target pattern; rasterizes the lithography target pattern to obtain a lithography target image; obtains an initial light source and an initial mask pattern according to the lithography target image; the initial mask pattern is a binary image; based on the lithography target image, an initial light source and an initial mask pattern are iteratively optimized by using a preset light source mask collaborative optimization loss function, and the adjusted initial light source and the adjusted initial mask pattern corresponding to the minimum function value of the light source mask collaborative optimization loss function are the optimized light source and the optimized mask pattern; the light source mask collaborative optimization loss function includes a main pattern edge error term and a sub-resolution auxiliary pattern printing term; calculate a first process window according to the optimized light source and the optimized mask pattern; simulate the optimized mask pattern under the optimized light source to obtain a simulation lithography image, and judge whether there is a sub-resolution auxiliary pattern printing in the simulation lithography image; when there is a printed sub-resolution auxiliary pattern, the size of the corresponding pattern in the optimized mask pattern is reduced, an updated optimized mask pattern is obtained, and the updated optimized mask pattern is simulated again until there is no sub-resolution auxiliary pattern printing in the simulation lithography image corresponding to the optimized mask pattern, and the latest optimized mask pattern is output as a process iteration mask pattern; calculate a second process window according to the optimized light source and the process iteration mask pattern; calculate the difference between the first process window and the second process window, and judge whether the proportion of the difference in the first process window exceeds a preset upper limit value; when the proportion does not exceed the upper limit value, the optimized light source and the process iteration mask pattern are output. The light source mask collaborative optimization loss function including the sub-resolution auxiliary pattern printing term is used to iteratively optimize the initial light source and the initial mask pattern, that is, the risk of sub-resolution auxiliary pattern printing is accurately added to the loss function, and then the size of the sub-resolution auxiliary pattern in the optimized mask pattern is further selectively reduced, and the loss of the process window in the size reduction process of the sub-resolution auxiliary pattern is limited by the upper limit value. Under the condition of ensuring that there is no sub-resolution auxiliary pattern printing, the optimal combination of light source and mask pattern is obtained, which greatly improves the accuracy and practicality of light source mask collaborative optimization.

[0226] The application further provides a computer readable storage medium, and the computer readable storage medium stores a computer program, and the computer program is executed by a processor to implement steps of any of the light source and mask co-optimization methods described above. The light source and mask co-optimization method provided by the application includes the following steps: receiving a lithography target pattern; rasterizing the lithography target pattern to obtain a lithography target image; obtaining an initial light source and an initial mask pattern according to the lithography target image; the initial mask pattern is a binary image; based on the lithography target image, an initial light source and an initial mask pattern are iteratively optimized by using a preset light source and mask co-optimization loss function, and the adjusted initial light source and the adjusted initial mask pattern corresponding to the minimum function value of the light source and mask co-optimization loss function are the optimized light source and the optimized mask pattern; the light source and mask co-optimization loss function includes a main pattern edge error term and a sub-resolution auxiliary pattern printing term; a first process window is calculated according to the optimized light source and the optimized mask pattern; under the optimized light source, the optimized mask pattern is simulated to obtain a simulation lithography image, and it is judged whether there is a sub-resolution auxiliary pattern printing in the simulation lithography image; when there is a printed sub-resolution auxiliary pattern, the size of the corresponding pattern in the optimized mask pattern is reduced, an updated optimized mask pattern is obtained, and the updated optimized mask pattern is simulated again until there is no sub-resolution auxiliary pattern printing in the simulation lithography image corresponding to the optimized mask pattern, and the latest optimized mask pattern is output as a process iteration mask pattern; a second process window is calculated according to the optimized light source and the process iteration mask pattern; a difference between the first process window and the second process window is calculated, and a proportion of the difference in the first process window is calculated; and it is judged whether the proportion value exceeds a preset upper limit value; when the proportion value does not exceed the upper limit value, the optimized light source and the process iteration mask pattern are output. The light source and mask co-optimization loss function including the sub-resolution auxiliary pattern printing term is used to iteratively optimize the initial light source and the initial mask pattern, that is, the risk of sub-resolution auxiliary pattern printing is accurately added to the loss function, and then the size of the sub-resolution auxiliary pattern in the optimized mask pattern is further selectively reduced, and the loss of the process window in the size reduction process of the sub-resolution auxiliary pattern is limited by the upper limit value, so that the optimal combination of the light source and the mask pattern is obtained under the condition that there is no sub-resolution auxiliary pattern printing, and the precision and practicability of the light source and mask co-optimization are greatly improved.

[0227] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to in the method section.

[0228] It should be noted that, in this specification, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0229] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0230] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein can be implemented directly by hardware, a software module executed by a processor, or a combination of both. The software module can be located in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art.

[0231] The light source mask cooperative optimization method, device, equipment and storage medium provided by the present application are described in detail. The principles and implementation modes of the present application are described by applying specific examples. The above description of the embodiments is only used to help understand the method of the present application and its core idea. It should be pointed out that, for ordinary skilled persons in the technical field, some improvements and modifications can be made to the present application without departing from the principles of the present application, and these improvements and modifications also fall within the protection scope of the present application.

Claims

1. A method for collaborative optimization of light source masks, characterized in that, include: Receive the photolithography target pattern; The photolithographic target pattern is rasterized to obtain a photolithographic target image; Based on the lithographic target image, obtain the initial light source and the initial mask pattern; The initial mask pattern is a binary image; Based on the lithographic target image, the initial light source and the initial mask pattern are iteratively optimized using a preset light source-mask collaborative optimization loss function. The adjusted initial light source and the adjusted initial mask pattern corresponding to the minimum function value of the light source-mask collaborative optimization loss function are determined as the optimized light source and the optimized mask pattern. The light source mask collaborative optimization loss function includes a main graphic edge error term and a sub-resolution auxiliary graphic printing term; Calculate the first process window based on the optimized light source and the optimized mask pattern; Under the optimized light source, the optimized mask pattern is simulated to obtain a simulated lithography image, and it is determined whether a sub-resolution auxiliary pattern is printed in the simulated lithography image. When there is a printed sub-resolution auxiliary pattern, the size of the corresponding pattern in the optimized mask pattern is reduced to obtain an updated optimized mask pattern. Then, the updated optimized mask pattern is simulated until no sub-resolution auxiliary pattern is printed in the simulated lithography image corresponding to the optimized mask pattern. The latest optimized mask pattern is then output as the process iteration mask pattern. The second process window is calculated based on the optimized light source and the process iteration mask pattern; Calculate the ratio of the difference between the first process window and the second process window to the first process window, and determine whether the ratio exceeds a preset upper limit value for acceptance. When the ratio value does not exceed the qualified upper limit value, the optimized light source and the process iteration mask pattern are output; The light source mask collaborative optimization loss function is the sum of the main graphic edge error term and the sub-resolution auxiliary graphic printing term; The sub-resolution auxiliary graphic printing item includes a sub-resolution weighting coefficient; After determining whether the ratio value exceeds the preset acceptable upper limit, the process further includes: When the ratio exceeds the qualified upper limit, the sub-resolution weight coefficient is increased to obtain the updated light source mask collaborative optimization loss function; Based on the lithography target image, the initial light source and the initial mask pattern are iteratively optimized using the updated light source mask collaborative optimization loss function to obtain an updated optimized light source and an updated optimized mask pattern. The corresponding first process window, the corresponding process iteration mask pattern, the corresponding second process window, and the corresponding ratio value are calculated. Then, it is determined whether the updated ratio value exceeds the preset qualified upper limit value. When the updated ratio value exceeds the qualified upper limit value, the sub-resolution weight coefficient is increased to obtain the updated light source mask collaborative optimization loss function until the ratio value does not exceed the qualified upper limit value.

2. The light source mask collaborative optimization method as described in claim 1, characterized in that, Based on the lithographic target image, the initial light source and the initial mask pattern are iteratively optimized using a preset light source-mask collaborative optimization loss function, including: The initial light source and the initial mask pattern are iteratively optimized using the following formula: ； Among them, w p w is the weighting coefficient of the process condition p corresponding to the initial light source. x EPE is the weighting coefficient of the initial mask pattern corresponding to the evaluation point x on the photolithographic target pattern. p,x The edge placement error of the evaluation point x under process condition p is β, where β is the sub-resolution weighting coefficient, and R is the edge placement error of the evaluation point x under process condition p. S j (p) represents the image value of grid point j printed in the sub-resolution auxiliary graphic under process condition p, T s The printing threshold for the sub-resolution auxiliary graphic.

3. The light source mask collaborative optimization method as described in claim 1, characterized in that, The size of the corresponding graphic in the optimized mask pattern of the printed sub-resolution auxiliary graphic is reduced to obtain an updated optimized mask pattern, including: The size of the corresponding graphic in the optimized mask pattern is reduced by shrinking the size of the sub-resolution auxiliary graphic to be printed, to obtain the mask pattern to be adjusted. Based on the photolithographic target image, under the optimized light source, the mask pattern to be adjusted is optimized using the following adjustment loss function, and the mask pattern to be adjusted corresponding to the minimum function value of the adjustment loss function is determined as the updated optimized mask pattern: ； Among them, w' p w' is the weighting coefficient of the process condition p corresponding to the optimized light source. x EPE' is the weighting coefficient of the mask pattern to be adjusted corresponding to the evaluation point x on the photolithographic target pattern. p,x The edge placement error of the position corresponding to the evaluation point x under process condition p.

4. The light source mask collaborative optimization method as described in claim 1, characterized in that, Based on the photolithographic target image, the initial light source and initial mask pattern are obtained, including: Based on the lithographic target image, the initial light source and initial mask image are obtained by optimizing using the initial loss function as shown in the following formula: ； Among them, w'' p w'' is the weighting coefficient of the process condition p corresponding to the initial light source. i R is the weight coefficient of the initial mask image corresponding to the evaluation point i on the photolithographic target pattern. M i (p) represents the image value of grid point i in the simulated photolithographic image of the initial mask image under process condition p. i The image value of grid point i in the lithographic target image; The initial mask image is binarized to obtain the initial mask pattern.

5. The light source mask collaborative optimization method as described in claim 1, characterized in that, The upper limit for acceptable values ​​is no greater than 5%.

6. The light source mask collaborative optimization method as described in claim 1, characterized in that, The simulation of iterative optimization of the initial light source and the initial mask pattern, and / or the simulation of the optimized mask pattern, are performed using a pre-trained sub-resolution assisted image exposure model. The training method for the sub-resolution assisted image exposure model includes: An uncalibrated sub-resolution auxiliary pattern exposure model is established. This model first determines the light intensity distribution image of the main pattern imaging plane based on the input mask pattern to be processed using a preset first cross-transfer coefficient. Then, it calculates the main pattern imaging plane photoresist image corresponding to the light intensity distribution image of the main pattern imaging plane using a preset photoresist model term and the corresponding linear coefficient. Finally, it determines the key dimensions of the main pattern from the main pattern imaging plane photoresist image using a first threshold. The sub-resolution auxiliary pattern exposure model also first determines the light intensity distribution image of the sub-resolution auxiliary pattern imaging plane based on the input mask pattern to be processed using a preset second cross-transfer coefficient. Then, it calculates the sub-resolution auxiliary pattern imaging plane photoresist image corresponding to the light intensity distribution image of the sub-resolution auxiliary pattern imaging plane using the photoresist model term and the linear coefficient. Finally, it determines the sub-resolution auxiliary pattern printing information from the sub-resolution auxiliary pattern imaging plane photoresist image using a second threshold. A test mask is fabricated based on a preset modeling pattern set, and the corresponding photoresist images of the test mask under different lighting sources are acquired; the modeling pattern set includes a variety of main patterns and sub-resolution auxiliary patterns of different sizes; Determine the key dimensions of the main graphic and the sub-resolution auxiliary graphic printing information corresponding to each of the acquired photoresist images; Using the modeling pattern set and the key dimensions of the main graphics corresponding to each of the illumination sources, as well as the sub-resolution auxiliary graphic printing information, the first cross-transfer coefficient, the second cross-transfer coefficient, the first threshold, the second threshold, the photoresist model term, and the corresponding linear coefficient are calibrated. This ensures that the key dimensions of the main graphics and the sub-resolution auxiliary graphic printing information obtained by inputting the modeling pattern set as a mask pattern to be processed into the sub-resolution auxiliary graphic exposure model are less than the difference between the key dimensions of the main graphics and the sub-resolution auxiliary graphic printing information of the acquired photoresist image under the corresponding illumination source and the difference between these two values ​​is less than a preset calibration threshold.

7. The light source mask collaborative optimization method as described in claim 6, characterized in that, The sub-resolution assisted graphic exposure model determines the light intensity distribution image of the main graphic imaging plane using the following formula: ； Among them, I M i (x,y) is the light intensity distribution image of the main graphic imaging plane, M(x,y) is the rasterized mask image corresponding to the mask pattern to be processed, and TCC i (Z M ) represents the first cross-transfer coefficient. This is a convolution operation; The sub-resolution assisted graphic exposure model determines the light intensity distribution image of the sub-resolution assisted graphic imaging plane using the following formula: ； Among them, I S i (x,y) represents the light intensity distribution image of the sub-resolution auxiliary image imaging plane, M(x,y) represents the rasterized mask image corresponding to the mask pattern to be processed, and TCC i (Z S ) is the second cross-transfer coefficient. This is a convolution operation; The sub-resolution assisted pattern exposure model calculates the main pattern imaging plane photoresist image corresponding to the main pattern imaging plane light intensity distribution image using the following formula: ； Among them, R M i (x,y) represents the main graphic imaging planar photoresist image, F j For the photoresist model term described in the j-th term, c j I represents the linear coefficient corresponding to the photoresist model term described in the j-th term. M i (x, y) represents the light intensity distribution image of the main graphic imaging plane, T M The first threshold; The sub-resolution assisted pattern exposure model calculates the sub-resolution assisted pattern imaging plane photoresist image corresponding to the sub-resolution assisted pattern imaging plane light intensity distribution image using the following formula: ； Among them, R S i (x,y) represents the sub-resolution auxiliary graphic imaging planar photoresist image, F j For the photoresist model term described in the j-th term, c j I represents the linear coefficient corresponding to the photoresist model term described in the j-th term. S i (x,y) represents the light intensity distribution image of the sub-resolution auxiliary image imaging plane, T S This is the second threshold.

8. The light source mask collaborative optimization method as described in claim 6, characterized in that, Using the modeling pattern set and the key dimensions of the main graphics corresponding to each of the lighting sources, as well as the sub-resolution auxiliary graphic printing information, the first cross-transfer coefficient, the second cross-transfer coefficient, the first threshold, the second threshold, the photoresist model term, and the corresponding linear coefficients are calibrated, including: When the value of the correction loss function in the following formula is minimized, the calibration of the first cross-transfer coefficient, the second cross-transfer coefficient, the first threshold, the second threshold, the photoresist model term, and the corresponding linear coefficient is completed: ； Where Cost is the value of the correction loss function. w is the difference between the key dimension of the k-th principal pattern in the modeling pattern set corresponding to the principal pattern in the acquired photoresist image and the key dimension of the principal pattern corresponding to the principal pattern in the photoresist image of the principal pattern imaging plane. k The graphic weight coefficient of the k-th principal graphic in the modeling pattern set. w represents the printing difference of the nth sub-resolution auxiliary graphic in the modeling pattern set. n The graphic weight coefficient is the nth sub-resolution auxiliary graphic in the modeling pattern set; When the printing result of the sub-resolution auxiliary pattern of the acquired photoresist image is the same as the printing result of the simulation result of the sub-resolution auxiliary pattern exposure model: ； When the printing result of the sub-resolution auxiliary pattern of the acquired photoresist image differs from the printing result of the simulation result of the sub-resolution auxiliary pattern exposure model: ； Among them, T S R is the second threshold. S max The maximum image value of the corresponding sub-resolution auxiliary pattern in the photoresist image of the sub-resolution auxiliary pattern imaging plane.

9. A light source mask collaborative optimization device, characterized in that, include: The receiving module is used to receive the photolithographic target pattern; A rasterization module is used to rasterize the lithographic target pattern to obtain a lithographic target image; The binarization module is used to obtain the initial light source and the initial mask pattern based on the photolithography target image; The initial mask pattern is a binary image; The first iteration module is used to iteratively optimize the initial light source and the initial mask pattern based on the photolithography target image using a preset light source mask collaborative optimization loss function, and determine the adjusted initial light source and the adjusted initial mask pattern corresponding to the minimum function value of the light source mask collaborative optimization loss function as the optimized light source and the optimized mask pattern. The light source mask collaborative optimization loss function includes a main graphic edge error term and a sub-resolution auxiliary graphic printing term; The first window module is used to calculate the first process window based on the optimized light source and the optimized mask pattern; The second iteration module is used to simulate the optimized mask pattern under the optimized light source to obtain a simulated lithography image, and determine whether a sub-resolution auxiliary pattern is printed in the simulated lithography image; when a sub-resolution auxiliary pattern is printed, the size of the corresponding pattern in the optimized mask pattern is reduced to obtain an updated optimized mask pattern, and then the updated optimized mask pattern is simulated until no sub-resolution auxiliary pattern is printed in the simulated lithography image corresponding to the optimized mask pattern, and the latest optimized mask pattern is output as the process iteration mask pattern. The second window module is used to calculate the second process window based on the optimized light source and the process iteration mask pattern. The ratio judgment module is used to calculate the ratio of the difference between the first process window and the second process window to the first process window, and to determine whether the ratio exceeds a preset qualified upper limit value. The output module is used to output the optimized light source and the process iteration mask pattern when the ratio value does not exceed the qualified upper limit value; The ratio determination module further includes: The weighting unit is used to increase the sub-resolution weighting coefficient when the ratio value exceeds the qualified upper limit value, so as to obtain the updated light source mask collaborative optimization loss function. The weighted recalculation unit is used to iteratively optimize the initial light source and the initial mask pattern based on the lithography target image and using the updated light source mask collaborative optimization loss function to obtain an updated optimized light source and an updated optimized mask pattern. It also calculates the corresponding first process window, the corresponding process iteration mask pattern, the corresponding second process window, and the corresponding ratio value. Then, it determines whether the updated ratio value exceeds a preset qualified upper limit value. When the updated ratio value exceeds the qualified upper limit value, it increases the sub-resolution weight coefficient to obtain the updated light source mask collaborative optimization loss function until the ratio value does not exceed the qualified upper limit value.

Images