A photoresist model optimization method

Abstract

The present application relates to the technical field of photoetching modeling, in particular to a photoetching model optimization method. The photoetching model optimization method provides an initial photoetching model, sets an initial measurement line, moves the initial measurement line in a preset range to obtain offset measurement lines, obtains the offset measurement line with the maximum grid point dependent error among all the offset measurement lines, compares the signal corresponding to each photoetching item of the offset measurement line with the signal corresponding to each photoetching item of the initial measurement line to obtain the signal with the greatest impact on the grid point dependent error, limits the signal, and obtains an initial calibration photoetching model. The signal with the greatest impact on the grid point dependent error is found by comparison, and the signal with the greatest impact is limited, so that the impact of the signal on the model is reduced during the calibration model process, thereby improving the accuracy of the model and solving the problem of insufficient model accuracy in the prior art.

Description

Technical Field

[0001] This invention relates to the field of photolithography modeling technology, and in particular to a photolithography model optimization method. Background Technology

[0002] The most crucial, complex, and costly process in integrated circuit manufacturing is lithography. Simply put, lithography refers to etching a designed pattern onto a silicon wafer according to a specific scale using light imaging. Computational lithography uses computer calculations to simulate the actual process, thereby effectively optimizing the process, saving costs, and shortening the development cycle. The lithography model is the foundation of computational lithography simulation, and the accuracy of the model directly determines the accuracy of the computational lithography simulation.

[0003] During modeling, continuous images are first sampled and quantized to convert them into digital images for computer processing. The grid quantization error during image pixelation is defined as grid-dependent error. Specifically, when the same image is located at different positions across the entire chip, insufficient sampling and inconsistent quantization assignments lead to differences in simulation results. During optical proximity correction, the same target image will yield different corrected images (post-opc).

[0004] Existing photoresist models suffer from large grid-dependent errors, resulting in insufficient accuracy. Summary of the Invention

[0005] To address the problem of insufficient accuracy in existing photoresist models, this invention provides a photoresist model optimization method.

[0006] The solution to the technical problem of this invention is to provide a photoresist model optimization method for reducing grid-point dependency error, characterized by comprising the following steps:

[0007] Provide an initial photoresist model;

[0008] Set an initial measurement line, and move the initial measurement line at least twice along a direction parallel to the initial measurement line at a preset interval within a preset range of the initial photoresist model to obtain the corresponding offset measurement line, and calculate the grid point dependency error of each offset measurement line.

[0009] Select the offset measurement line with the largest grid point dependency error, and output the signals corresponding to the photoresist terms of the offset measurement line and the initial measurement line respectively to determine the signal with the greatest impact on the grid point dependency error;

[0010] The signal that has the greatest impact on grid-dependent errors is restricted, and an initial calibrated photoresist model is obtained.

[0011] Preferably, the method further includes the following steps:

[0012] For the initial calibrated photoresist model, the grid dependency error is calculated again following the steps described above.

[0013] If the grid dependency error meets the preset conditions, the initially calibrated photoresist model is used as the final photoresist model.

[0014] If the grid-point dependency error does not meet the preset conditions, the signal that has the greatest impact on the grid-point dependency error will be restricted again, and the initial calibration photoresist model will be recalibrated.

[0015] Preferably, the preset condition is set as the target numerical range of the grid-dependent error or the number of recalibrations.

[0016] Preferably, the signal is the signal corresponding to each photoresist item decomposed by computer from the photoresist model. The signal includes the photoresist item and its corresponding coefficient. The method of limiting the signal is to limit the size of its coefficient.

[0017] Preferably, the preset interval is set to 0.5-5nm.

[0018] Preferably, the grid-point dependency error is the difference between the simulated value of the offset measurement line and the simulated value of the initial measurement line, and the offset measurement line with the largest difference is the offset measurement line with the largest grid-point dependency error.

[0019] Preferably, the simulated value of the measurement line is obtained by truncating the total signal of each measurement line obtained by linearly superimposing all signals of the offset measurement line or the initial measurement line.

[0020] Preferably, the formula for defining the preset range is as follows: , where pixel represents the pixel size of the initial optical model and shift represents the offset range of the offset measurement line.

[0021] Preferably, the specific steps for determining the signal that has the greatest impact on grid-point dependency error include:

[0022] Compare the signals associated with each photoresist term corresponding to the offset measurement line with the largest grid-dependent error and the initial measurement line;

[0023] The difference between each signal of the offset measurement line with the largest error and the initial measurement line is calculated.

[0024] Based on the above signal differences, the signal with the greatest impact on grid-point dependency error is determined. The signal with the largest absolute value of the difference has the greatest impact, that is, the photoresist term corresponding to this signal has the greatest impact on grid-point dependency error.

[0025] Preferably, the coefficient of the photoresist term is limited by the multiple relationship between the signal difference between the signal that has the greatest impact on grid-dependent error and the signal corresponding to the initial measurement line, and the value that the signal difference is expected to reach.

[0026] Compared with the prior art, the photoresist model optimization method of the present invention has the following advantages:

[0027] 1. The photoresist model optimization method of the present invention first provides an initial photoresist model, then sets initial measurement lines, moves the initial measurement lines within a preset range to obtain offset measurement lines, and finds the offset measurement line with the largest grid point dependency error among all offset measurement lines. The signal corresponding to each photoresist item of this offset measurement line is compared with the signal corresponding to each photoresist item of the initial measurement line to obtain the signal with the greatest impact on the grid point dependency error. This signal is then constrained to obtain the initial calibrated photoresist model. By calculating and comparing to find the signal with the greatest impact on the grid point dependency error and constraining the design of the signal with the greatest impact, the influence of the signal on the model during the calibration process is reduced, thereby improving the model's accuracy and solving the problem of insufficient model accuracy in existing technologies.

[0028] 2. The photoresist model optimization method of the present invention further includes the following steps: calculating the grid dependency error again for the initially calibrated photoresist model according to the above steps; if the grid dependency error meets the preset conditions, then using the initially calibrated photoresist model as the final photoresist model; if the grid dependency error does not meet the preset conditions, then restricting the signal that has the greatest impact on the grid dependency error again and recalibrating the initially calibrated photoresist model. By recalibrating the model after optimization and then judging whether the preset conditions are met, the design is prevented from still not meeting the customer's requirements after optimization. At the same time, it can be automatically implemented, improving the efficiency and accuracy of optimization.

[0029] 3. The preset conditions of the photoresist model optimization method of the present invention are set to the target numerical range of the grid-dependent error or the number of recalibrations. Setting the preset conditions to the target numerical range of the grid-dependent error can improve the model accuracy improved by a single optimization; setting the preset conditions to the number of recalibrations can reduce the number of optimization resettings and improve optimization efficiency.

[0030] 4. The signal in the photoresist model optimization method of this invention is the signal corresponding to each photoresist item decomposed from the photoresist model by computer. The signal includes the photoresist item and its corresponding coefficient. The method of limiting the signal is to limit the magnitude of its coefficient. The influence of the photoresist item on the grid dependency error is expressed by calculating the signal difference between the initial measurement line and the offset measurement line. When the signal of the offset measurement line with the largest grid dependency error is subtracted from the signal of the initial measurement line, it can be directly and clearly seen which photoresist item has the greatest influence on the grid dependency error, thus improving efficiency. At the same time, the influence of the photoresist item on the model can be directly limited by limiting the magnitude of its coefficient, which is simple and fast.

[0031] 5. The preset interval of the photoresist model optimization method of the present invention is set to 0.5-5nm. Existing photoresist processes are all at the nanometer level. Setting the preset interval to 0.5-5nm can ensure both the accuracy and efficiency of detection, thereby achieving accurate and rapid detection results.

[0032] 6. The grid-point dependency error of the photoresist model optimization method of the present invention is the difference between the simulated value of the offset measurement line and the simulated value of the initial measurement line. The offset measurement line with the largest difference is the offset measurement line with the largest grid-point dependency error. The design of calculating the grid-point dependency error by using the difference of simulated values ​​makes the grid-point dependency error data more quantifiable, clearer, and easier to determine which grid-point dependency error is the largest.

[0033] 7. The photoresist model optimization method of the present invention obtains the simulated value of each measurement line by truncating the total signal of each line obtained by linearly superimposing all signals of the initial measurement line or the offset measurement line. By truncating the total signal after linearly superimposing all signals, the grid-dependent error at the measurement line is determined by the simulated value, making the judgment quantifiable and easy to make.

[0034] 8. The formula for defining the preset range of the photoresist model optimization method of the present invention is as follows: Pixel represents the pixel size of the initial optical model, and shift represents the offset range of the offset measurement line. The design, which offsets the initial photoresist model by half a pixel size to the left and right of the initial measurement line, facilitates comparison and improves comparison efficiency.

[0035] 9. The specific steps of the photoresist model optimization method of the present invention to determine the signal that contributes most to the grid-point dependency error include: comparing the signals related to each photoresist term corresponding to the offset measurement line with the largest grid-point dependency error and the initial measurement line; calculating the signal difference between the offset measurement line with the largest error and the initial measurement line; and determining the signal with the greatest impact on the grid-point dependency error based on the above signal difference, where the largest absolute value of the difference indicates the greatest impact, i.e., the photoresist term corresponding to that signal has the greatest impact on the grid-point dependency error. By using the signal difference between the offset measurement line with the largest grid-point dependency error and the initial measurement line to determine which photoresist term has a greater impact on the grid-point dependency error, the accuracy of the model can be improved by limiting the signal size in subsequent model calibration.

[0036] 10. The photoresist model optimization method of the present invention limits the coefficients of the photoresist term by establishing a multiple relationship between the signal difference between the signal with the greatest influence of grid-dependent error and the signal corresponding to the initial measurement line, and the desired value of that signal difference. By limiting the coefficients of the photoresist term through the multiple by which the signal difference decreases, erroneous optimization directions are reduced, and optimization efficiency is improved. Attached Figure Description

[0037] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0038] Figure 1 This is a flowchart of the photoresist model optimization method provided in the first embodiment of the present invention.

[0039] Figure 2 This is a detailed step diagram of the photoresist model optimization method provided in the first embodiment of the present invention.

[0040] Figure 3 This is a schematic diagram of the truncation operation of the photoresist model optimization method provided in the first embodiment of the present invention.

[0041] Figure 4 This is a flowchart of the photoresist model optimization method provided in the first embodiment of the present invention.

[0042] Figure 5 This is the initial grid-dependent error map of the example operation of this invention.

[0043] Figure 6 This is a comparison chart of photoresist errors in the example operation of this invention.

[0044] Figure 7This is a schematic diagram illustrating the operation of limiting the photoresist term coefficient in the example operation of the present invention.

[0045] Figure 8 This is an optimized grid-dependent error map of the example operation of this invention. Detailed Implementation

[0046] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0047] In the embodiments provided by this invention, it should be understood that "B corresponding to A" means that B is associated with A, and B can be determined based on A. However, it should also be understood that determining B based on A does not mean determining B solely based on A; B can also be determined based on A and / or other information.

[0048] It should be understood that the phrase "one embodiment" or "an embodiment" throughout the specification means that a specific feature, structure, or characteristic related to the embodiment is included in at least one embodiment of the invention. Therefore, "in one embodiment" or "in an embodiment" appearing throughout the specification does not necessarily refer to the same embodiment. Furthermore, these specific features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. Those skilled in the art should also recognize that the embodiments described in the specification are optional embodiments, and the actions and modules involved are not necessarily essential to the invention.

[0049] In various embodiments of the present invention, it should be understood that the sequence number of each process does not necessarily imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0050] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It is particularly important to note that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0051] Please see Figure 1 The first embodiment of the present invention provides a photoresist model optimization method, including the following steps:

[0052] Includes the following steps:

[0053] S1: Provides the initial photoresist model;

[0054] S2: Set the initial measurement line, move the initial measurement line at least twice along a direction parallel to the initial measurement line at a preset interval within a preset range of the initial photoresist model to obtain the corresponding offset measurement line, and calculate the grid point dependency error of each offset measurement line.

[0055] S3: Select the offset measurement line with the largest grid point dependency error, and output the signals corresponding to the photoresist terms of the offset measurement line and the initial measurement line respectively to determine the signal with the greatest impact on the grid point dependency error;

[0056] S4: Limit the signal that has the greatest impact on grid-dependent errors and obtain the initial calibrated photoresist model.

[0057] By calculating and comparing, the signal with the greatest impact on grid-point dependency error is found, and the design of the signal with the greatest impact is restricted. This reduces the influence of the signal on the model during the calibration process, thereby improving the model's accuracy and solving the problem of insufficient accuracy of existing models.

[0058] Furthermore, in the first embodiment of the present invention, the preset interval is set to 0.5-5 nm. Existing photoresist processes are all at the nanometer level. Setting the preset interval to 0.5-5 nm can ensure both the accuracy and efficiency of detection, thereby achieving accurate and rapid detection.

[0059] Understandably, the preset interval size is related to the photoresist model pixel. The first embodiment of the present invention uses a 28nm process with a preset interval of 1nm. When the process pixel is larger, 2nm, 3nm, 4nm or 5nm can be used as the preset interval. When the pixel is smaller, 0.5nm can be used as the preset interval.

[0060] Specifically, the formula for defining the preset range is as follows: Pixel represents the pixel size of the initial optical model, and shift represents the offset range of the offset measurement line. By offsetting the initial measurement line by half the pixel size of the initial photoresist model in opposite directions, comparison is facilitated and comparison efficiency is improved. For example, in the first embodiment of this invention, the pixel size is 28nm, and the starting point is set to 0, then the preset range is -14nm to 14nm.

[0061] Further, please refer to Figure 2 The photoresist model optimization method 1 of the present invention further includes the following steps:

[0062] S501: Calculate the grid dependency error again for the initial calibrated photoresist model following the steps above.

[0063] S502: If the grid dependency error meets the preset conditions, the initially calibrated photoresist model is used as the final photoresist model.

[0064] S503: If the grid point dependency error does not meet the preset conditions, then restrict the signal that has the greatest impact on the grid point dependency error again and recalibrate the initial calibration photoresist model.

[0065] By optimizing and recalibrating the model, and then determining whether the design meets the preset conditions, it is possible to prevent the optimized model from still not meeting the customer's requirements in terms of accuracy. This process can be automated, improving the efficiency and accuracy of optimization.

[0066] Specifically, the preset conditions are set to the target numerical range of the grid-dependent error or the number of recalibrations. Setting the preset conditions to grid-dependent error can improve the model accuracy gained in a single optimization; setting the preset conditions to the number of recalibrations can reduce the number of optimization resettings and improve optimization efficiency.

[0067] Furthermore, the signal is the signal corresponding to each photoresist item decomposed from the photoresist model by computer. The signal includes the photoresist item and its corresponding coefficient. The method of limiting the signal is to limit the magnitude of its coefficient. The influence of the photoresist item on the grid dependency error is expressed by calculating the signal difference between the initial measurement line and the offset measurement line. When the signal of the offset measurement line with the largest grid dependency error is subtracted from the signal of the initial measurement line, it can be directly and clearly seen which photoresist item has the greatest impact on the grid dependency error, thus improving efficiency. At the same time, the influence of the photoresist item on the model can be directly limited by limiting the magnitude of its coefficient, which is simple and fast.

[0068] Understandably, different photoresist terms can be used to characterize different physical and chemical effects, such as molecular diffusion and acid-base neutralization. The number of photoresist terms in each photoresist model is not fixed. A photoresist term is a mathematical expression, and the signal is the signal representing a single photoresist term extracted from the established photoresist model using a computer algorithm. This signal includes the photoresist term and a coefficient, which is a property of the established model itself. A single signal has no physical meaning.

[0069] Please see Figure 3 The resist image (i.e., mathematical image) is obtained by linearly superimposing all the signals. Then, the resist image is truncated by a truncation function (resist threshold) to obtain the simulation value (i.e., the size of the simulated model). The truncation function varies depending on the requirements of different models.

[0070] Understandably, grid point dependency error is the difference between the simulated value of the offset measurement line and the simulated value of the initial measurement line. The offset measurement line with the largest difference is the offset measurement line with the largest grid point dependency error. Calculating grid point dependency error using the difference in simulated values ​​makes the grid point dependency error data more quantifiable, clearer, and easier to determine which grid point has the largest dependency error.

[0071] Furthermore, the photoresist model optimization method of the present invention obtains the simulated value of the measurement line by truncating the total signal of each line obtained by linearly superimposing all signals of the initial measurement line or the offset measurement line. By truncating the total signal after linearly superimposing all signals, the grid-dependent error at the measurement line is determined by the simulated value, making the judgment quantifiable and easy to make.

[0072] Specifically, please refer to Figure 4 The specific steps for identifying photoresist terms that have a significant impact on grid-dependent errors include:

[0073] S101: Compare the signal associated with each photoresist term corresponding to the offset measurement line with the initial measurement line and the measurement line with the offset measurement line with the largest grid-dependent error;

[0074] S102: Calculate the signal difference between the offset measurement line with the largest error and the initial measurement line;

[0075] S103: Based on the above signal differences, determine the signal that has the greatest impact on grid point dependency error. The signal with the largest absolute value of the difference has the greatest impact, that is, the photoresist term corresponding to the signal has the greatest impact on grid point dependency error.

[0076] By using the difference between the signal of the offset measurement line with the largest grid-dependent error and the initial measurement line, we can determine which photoresist term has a greater impact on the grid-dependent error. This allows us to improve the model's accuracy in subsequent model calibration by limiting the signal magnitude.

[0077] Specifically, the photoresist model optimization method of the present invention limits the coefficients of the photoresist term by establishing a multiple relationship between the signal difference between the signal with the greatest influence of grid-dependent error and the signal corresponding to the initial measurement line, and the desired value of that signal difference. By limiting the coefficients of the photoresist term through the multiple by which the signal difference decreases, erroneous optimization directions are reduced, and optimization efficiency is improved.

[0078] Specifically, if the signal difference between the signal that has the greatest impact on grid-point dependency error and the signal corresponding to the initial measurement line is 5.0, and we want to limit this signal difference to within 1.0, then the coefficient can be set to 1 / 5, or 0.2.

[0079] For example, a node model using a 28nm process is selected for practical operation. The center point of the initial measurement line is selected as the starting point and set to shift 0. The initial measurement line is offset by 14nm on both sides of the starting point at a preset interval of 1nm, resulting in 28 offset measurement lines. The offset measurement lines are represented by shift i, where i is the nth line of the offset.

[0080] like Figure 5 As shown, it can be clearly seen that the offset measurement line grid dependency error is largest when the offset distance is 7nm. The signals corresponding to all photoresist terms at shift 7 and shift 0 are output, and the signal change difference is calculated, as shown below. Figure 6 As shown, "term" represents the photoresist term, and "term1," "term2," "term3," "term4," and "term5" represent five different photoresist terms. It can be seen that "term4" has the greatest impact on grid-dependent errors. The coefficient for "term4" is set to 0.045. After recalibrating the model, the above steps are repeated. Figure 7 As shown, after repeating the above steps three times, the coefficients for term3, term4, and term5 are all restricted. The coefficient for term3 is restricted to 0.057, the coefficient for term4 is restricted to 0.045, and the coefficient for term5 is restricted to 0.05.

[0081] The optimized grid dependency error is as follows: Figure 8 As shown, it can be seen that the grid-point dependency error of the model after limiting the photoresist term coefficient has decreased from the original maximum value of 4.7 to less than 1.2. At the same time, the grid-point dependency errors of other offset measurement lines have also decreased synchronously, which fully verifies the effectiveness of this scheme in reducing grid-point dependency error.

[0082] Compared with the prior art, the photoresist model optimization method of the present invention has the following advantages:

[0083] 1. The photoresist model optimization method of the present invention first provides an initial photoresist model, then sets initial measurement lines, moves the initial measurement lines within a preset range to obtain offset measurement lines, finds the offset measurement line with the largest grid point dependency error among all offset measurement lines, compares the signal corresponding to each photoresist item of the offset measurement line with the signal corresponding to each photoresist item of the initial measurement line to obtain the signal with the greatest impact on the grid point dependency error, restricts this signal, and obtains the initial calibrated photoresist model. By calculating and comparing to find the signal with the greatest impact on the grid point dependency error and restricting the design of the signal with the greatest impact, the influence of the signal on the model during the calibration process is reduced, thereby improving the accuracy of the model and solving the problem of insufficient model accuracy in the prior art. 2. The photoresist model optimization method of the present invention further includes the following steps: calculating the grid point dependency error again for the initial calibrated photoresist model according to the above steps; if the grid point dependency error meets the preset conditions, the initial calibrated photoresist model is used as the final photoresist model; if the grid point dependency error does not meet the preset conditions, the signal with the greatest impact on the grid point dependency error is restricted again, and the initial calibrated photoresist model is recalibrated. By optimizing and recalibrating the model, and then determining whether the design meets the preset conditions, it is possible to prevent the optimized model from still not meeting the customer's requirements in terms of accuracy. This process can be automated, improving the efficiency and accuracy of optimization.

[0084] 3. The preset conditions of the photoresist model optimization method of the present invention are set to the target numerical range of the grid-dependent error or the number of recalibrations. Setting the preset conditions to the target numerical range of the grid-dependent error can improve the model accuracy improved by a single optimization; setting the preset conditions to the number of recalibrations can reduce the number of optimization resettings and improve optimization efficiency.

[0085] 4. The signal in the photoresist model optimization method of this invention is the signal corresponding to each photoresist item decomposed from the photoresist model by computer. The signal includes the photoresist item and its corresponding coefficient. The method of limiting the signal is to limit the magnitude of its coefficient. The influence of the photoresist item on the grid dependency error is expressed by calculating the signal difference between the initial measurement line and the offset measurement line. When the signal of the offset measurement line with the largest grid dependency error is subtracted from the signal of the initial measurement line, it can be directly and clearly seen which photoresist item has the greatest influence on the grid dependency error, thus improving efficiency. At the same time, the influence of the photoresist item on the model can be directly limited by limiting the magnitude of its coefficient, which is simple and fast.

[0086] 5. The preset interval of the photoresist model optimization method of the present invention is set to 0.5-5nm. Existing photoresist processes are all at the nanometer level. Setting the preset interval to 0.5-5nm can ensure both the accuracy and efficiency of detection, thereby achieving accurate and rapid detection results.

[0087] 6. The grid-point dependency error of the photoresist model optimization method of the present invention is the difference between the simulated value of the offset measurement line and the simulated value of the initial measurement line. The offset measurement line with the largest difference is the offset measurement line with the largest grid-point dependency error. The design of calculating the grid-point dependency error by using the difference of simulated values ​​makes the grid-point dependency error data more quantifiable, clearer, and easier to determine which grid-point dependency error is the largest.

[0088] 7. The photoresist model optimization method of the present invention obtains the simulated value of each measurement line by truncating the total signal of each line obtained by linearly superimposing all signals of the initial measurement line or the offset measurement line. By truncating the total signal after linearly superimposing all signals, the grid-dependent error at the measurement line is determined by the simulated value, making the judgment quantifiable and easy to make.

[0089] 8. The formula for defining the preset range of the photoresist model optimization method of the present invention is as follows: Pixel represents the pixel size of the initial optical model, and shift represents the offset range of the offset measurement line. The design, which offsets the initial photoresist model by half a pixel size to the left and right of the initial measurement line, facilitates comparison and improves comparison efficiency.

[0090] 9. The specific steps of the photoresist model optimization method of the present invention to determine the signal that contributes most to the grid-point dependency error include: comparing the signals related to each photoresist term corresponding to the offset measurement line with the largest grid-point dependency error and the initial measurement line; calculating the signal difference between the offset measurement line with the largest error and the initial measurement line; and determining the signal with the greatest impact on the grid-point dependency error based on the above signal difference, where the largest absolute value of the difference indicates the greatest impact, i.e., the photoresist term corresponding to that signal has the greatest impact on the grid-point dependency error. By using the signal difference between the offset measurement line with the largest grid-point dependency error and the initial measurement line to determine which photoresist term has a greater impact on the grid-point dependency error, the accuracy of the model can be improved by limiting the signal size in subsequent model calibration.

[0091] 10. The photoresist model optimization method of the present invention limits the coefficients of the photoresist term by establishing a multiple relationship between the signal difference between the signal with the greatest influence of grid-dependent error and the signal corresponding to the initial measurement line, and the desired value of that signal difference. By limiting the coefficients of the photoresist term through the multiple by which the signal difference decreases, erroneous optimization directions are reduced, and optimization efficiency is improved.

[0092] The above provides a detailed description of a photoresist model optimization method disclosed in the embodiments of the present invention. Specific examples have been used to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention. Any modifications, equivalent substitutions, and improvements made within the principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A photoresist model optimization method for reducing grid-dependent errors, characterized in that, Includes the following steps: Provide an initial photoresist model; Set an initial measurement line, and move the initial measurement line at least twice along a direction parallel to the initial measurement line within a preset range of the initial photoresist model at preset intervals to obtain the corresponding offset measurement line, and calculate the grid point dependency error of each offset measurement line. Select the offset measurement line with the largest grid point dependency error, and output the signals corresponding to the photoresist terms of the offset measurement line and the initial measurement line respectively to determine the signal with the greatest impact on the grid point dependency error; The signal that has the greatest impact on grid-dependent errors is restricted, and an initial calibrated photoresist model is obtained.

2. The photoresist model optimization method as described in claim 1, characterized in that: It also includes the following steps: For the initial calibrated photoresist model, the grid dependency error is calculated again following the steps described above. If the grid dependency error meets the preset conditions, the initially calibrated photoresist model is used as the final photoresist model. If the grid-point dependency error does not meet the preset conditions, the signal that has the greatest impact on the grid-point dependency error will be restricted again, and the initial calibration photoresist model will be recalibrated.

3. The photoresist model optimization method as described in claim 2, characterized in that: The preset conditions are set as either the target numerical range of the grid-dependent error or the number of recalibrations.

4. The photoresist model optimization method as described in claim 1, characterized in that: The signal is the signal corresponding to each photoresist item extracted by computer from the photoresist model. The signal includes the photoresist item and its corresponding coefficient. The method of limiting the signal is to limit the size of its coefficient.

5. The photoresist model optimization method as described in claim 1, characterized in that: The preset interval is set to 0.5-5nm.

6. The photoresist model optimization method as described in claim 1, characterized in that: The grid point dependency error is the difference between the simulated value of the offset measurement line and the simulated value of the initial measurement line. The offset measurement line with the largest difference is the offset measurement line with the largest grid point dependency error.

7. The photoresist model optimization method as described in claim 5, characterized in that: The simulated value of the offset measurement line or the initial measurement line is obtained by truncating the total signal of each point obtained by linearly superimposing all signals of the offset measurement line or the initial measurement line.

8. The photoresist model optimization method as described in claim 1, characterized in that: The preset range is defined by the formula -pixel / 2≤shift≤pixel / 2, where pixel represents the pixel size of the initial photoresist model and shift represents the offset range of the offset measurement line.

9. The photoresist model optimization method as described in claim 4, characterized in that: The specific steps for determining the signal that has the greatest impact on grid-point dependency error include: Compare the signals associated with each photoresist term corresponding to the offset measurement line with the largest grid-dependent error and the initial measurement line; The difference between each signal of the offset measurement line with the largest error and the initial measurement line is calculated. Based on the above signal differences, the signal with the greatest impact on grid-point dependency error is determined. The signal with the largest absolute value of the difference has the greatest impact, that is, the photoresist term corresponding to this signal has the greatest impact on grid-point dependency error.

10. The photoresist model optimization method as described in claim 9, characterized in that: The coefficient of the photoresist term is limited by the multiple relationship between the signal difference between the signal that has the greatest impact on grid-dependent error and the signal corresponding to the initial measurement line, and the value that the signal difference is expected to reach.

Images