A mask pattern correction method and correction system

Abstract

The present invention provides a method and system for correcting reticle patterns. The method comprises the following steps: setting multiple main patterns and multiple auxiliary patterns, collecting wafer data after photolithography of the main and auxiliary patterns; searching the reticle for a main pattern with the same characteristic dimensions as the auxiliary pattern as a reference pattern; obtaining the scale ratio of the reference pattern when it is transferred from the reticle to the wafer, and obtaining virtual size data of the auxiliary pattern transferred to the wafer based on the scale ratio and the size information of the auxiliary pattern on the reticle; and using the virtual size data and the wafer data as modeling parameters of an optical proximity correction model, iteratively establishing the optical proximity correction model until the correction error is less than a threshold. The present invention provides a method and system for correcting reticle patterns, which can improve the accuracy of pattern transfer during photolithography, thereby improving the yield rate of semiconductor manufacturing.

Description

Technical Field

[0001] The present invention relates to the technical field of semiconductor manufacturing, and in particular to a mask pattern correction method and correction system. Background Art

[0002] Photolithography is a key process in semiconductor manufacturing, used to transfer designed circuit patterns from a mask onto a wafer. Limited by the diffraction of light, the resolution, process window, and optical corrections directly impact the performance of this process. Due to light interference and diffraction, the actual pattern transferred to the wafer will be distorted. Therefore, optical proximity correction technology is required to correct the mask pattern so that the distorted pattern conforms to the actual design.

[0003] On a layout, the lithography process window for sparsely populated patterns is significantly smaller than that for densely populated patterns, limiting the overall process window. Therefore, adding auxiliary patterns can help balance the overall process window. Adding auxiliary patterns can complicate optical proximity effect correction, as correction modeling cannot account for the dimensional characteristics of the auxiliary patterns. Transfer defects from the auxiliary patterns can pose significant yield risks to semiconductor manufacturing. Summary of the Invention

[0004] The object of the present invention is to provide a mask pattern correction method and correction system, which can improve the pattern transfer accuracy of photolithography, thereby improving the semiconductor manufacturing yield.

[0005] To solve the above technical problems, the present invention is achieved through the following technical solutions:

[0006] The present invention provides a method for correcting a pattern of a mask, comprising the following steps:

[0007] Setting a plurality of main patterns and a plurality of auxiliary patterns, and collecting wafer data of the main patterns and the auxiliary patterns after photolithography;

[0008] Searching for the main pattern having the same feature size as the auxiliary pattern on the mask as a reference pattern;

[0009] Obtaining a scaling ratio of the reference pattern when it is transferred from the mask to the wafer, and obtaining virtual size data of the auxiliary pattern when it is transferred to the wafer according to the scaling ratio and size information of the auxiliary pattern on the mask; and

[0010] The virtual size data and the wafer data are used as modeling parameters of an optical proximity correction model, and the optical proximity correction model is iteratively established until a correction error is less than a threshold.

[0011] In one embodiment of the present invention, setting the widths of the plurality of auxiliary graphics comprises the following steps:

[0012] According to the optical signal intensity of the auxiliary pattern in the photolithography process, the pattern width when the optical signal intensity is close to the exposure threshold is used as the visible width of the auxiliary pattern;

[0013] Setting a test width range, the test width range covering the display width; and

[0014] A plurality of the auxiliary patterns are provided, wherein widths of the plurality of the auxiliary patterns traverse the test width range.

[0015] In one embodiment of the present invention, the step of obtaining the virtual size data is based on the following formula:

[0016] ADI CD2=(ADI CD1 / mask CD1)×mask CD2;

[0017] Among them, ADI CD2 is the virtual size data, ADI CD1 is the measured size of the main pattern after being transferred to the wafer, mask CD1 is the designed size of the main pattern on the mask, and mask CD2 is the designed size of the auxiliary pattern on the mask.

[0018] In one embodiment of the present invention, after obtaining the virtual size data, the virtual size data of the plurality of auxiliary graphics are adjusted until the virtual size data is positively correlated with the image exposure degree of the auxiliary graphics, the width of the auxiliary graphics, and the width of the reference graphics.

[0019] In one embodiment of the present invention, in the step of adjusting the virtual size data of the auxiliary graphic, an electron microscope image of the auxiliary graphic is obtained, and the occupied area data of the auxiliary graphic on the electron microscope image is output, wherein the occupied area data is used to characterize the image exposure degree, and the occupied area data is positively correlated with the image exposure degree.

[0020] In one embodiment of the present invention, in the step of adjusting the virtual size data of the auxiliary pattern, the virtual size data is adjusted with reference to the widths of the auxiliary pattern and the reference pattern on the mask.

[0021] In one embodiment of the present invention, the step of establishing the optical proximity correction model includes:

[0022] Acquiring an initial model of the optical proximity correction model, wherein parameters involved in the initial model include the virtual size data of the auxiliary pattern;

[0023] Setting parameter value ranges for parameters involved in the primary model, and taking values ​​from the parameter value ranges to obtain the primary model with determined parameters;

[0024] Acquire simulated lithography data of the pattern size data of the mask after being processed by the primary model;

[0025] Obtaining a deviation value between the simulated lithography data and the wafer data, taking the first-generation model with the smallest deviation value as the second-generation model, and determining whether the deviation value of the second-generation model is less than a preset threshold; and

[0026] Adjust the parameter value range and the model framework combination of the first-generation model, and loop through the steps of obtaining the second-generation model until the deviation value of the second-generation model is less than the preset threshold.

[0027] In one embodiment of the present invention, in the step of obtaining the second-generation model, values ​​are taken from the parameter value range in multiple rounds, and the model with the smallest deviation value is selected from the multiple first-generation models in each round as the second-generation model, until the deviation value of the second-generation model is less than the preset threshold.

[0028] In one embodiment of the present invention, when the parameter value ranges of all parameters are traversed and the deviation value of the second-generation model is greater than or equal to the preset threshold, the model framework combination of the first-generation model is adjusted.

[0029] The present invention provides a mask pattern correction system, comprising:

[0030] A graphic database storing a plurality of main graphics and a plurality of auxiliary graphics;

[0031] A photolithography simulation module is used to simulate the photolithography process of transferring the pattern on the mask to the wafer and obtain wafer data after the pattern is transferred;

[0032] a pattern reference module, configured to search on the mask for the main pattern having the same feature size as the auxiliary pattern as a reference pattern;

[0033] a data calculation module, configured to obtain a scaling ratio of the reference pattern when it is transferred from the mask to the wafer, and obtain virtual size data of the auxiliary pattern when it is transferred to the wafer according to the scaling ratio and size information of the auxiliary pattern on the mask; and

[0034] The modeling module is configured to use the virtual size data and the wafer data as modeling parameters of an optical proximity correction model, and iteratively establish the optical proximity correction model until a correction error is less than a threshold.

[0035] As described above, the present invention provides a mask pattern correction method and correction system, and its unexpected technical effect is that: when the auxiliary pattern transfer cannot be quantitatively studied, the present invention can quantify the exposure degree of the auxiliary pattern through virtual size data, so that the auxiliary pattern is taken into account in the optical proximity correction to avoid the auxiliary pattern being transferred to the wafer. And the present invention can dynamically adjust the virtual size data of the auxiliary pattern, so that the virtual size data can accurately reflect the degree to which the auxiliary pattern is exposed. The present invention can calibrate the auxiliary pattern by the visible width, thereby improving the parameter accuracy and parameter comprehensiveness of the wafer data and the virtual size data. The pattern correction method and correction system provided by the present invention are based on virtual data driven modeling and multi-parameter collaborative optimization. While improving the photolithography process window through the auxiliary pattern, it can strictly avoid the auxiliary pattern from being developed on the wafer, thereby significantly improving the yield of the photolithography process.

[0036] Of course, any product implementing the present invention does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0038] Figure 1 Schematic diagram of the distribution of main graphics and auxiliary graphics in one embodiment of the present invention.

[0039] Figure 2 Schematic diagram of the contour curve after the auxiliary graphic is developed in a computer according to one embodiment of the present invention.

[0040] Figure 3 FIG. 1 is a schematic diagram of an electron microscope in which an auxiliary pattern is transferred onto a wafer in one embodiment of the present invention.

[0041] Figure 4 FIG. 4 is a flow chart of a graphics correction method according to an embodiment of the present invention.

[0042] Figure 5 FIG. 1 is a flowchart of step S100 in one embodiment of the present invention.

[0043] Figure 6 FIG. 1 is a schematic diagram of optical signal intensity for obtaining the auxiliary pattern display width according to an embodiment of the present invention.

[0044] Figure 7 An electron microscope image of the first level of auxiliary patterns in one embodiment of the present invention.

[0045] Figure 8 This is an electron microscope image of the second level of auxiliary patterns in one embodiment of the present invention.

[0046] Figure 9 This is an electron microscope image of the third level of auxiliary patterns in one embodiment of the present invention.

[0047] Figure 10 This is an electron microscope image of the fourth level of auxiliary patterns in one embodiment of the present invention.

[0048] Figure 11 FIG. 4 is a flow chart of step S400 in one embodiment of the present invention.

[0049] Figure 12 FIG. 4 is a structural diagram of an image correction system according to an embodiment of the present invention.

[0050] In the figure: A, main graphic; B, auxiliary graphic; 10, graphic correction system; 11, graphic database; 12, simulated lithography module; 13, graphic reference module; 14, data calculation module; 15, data adjustment module; 16, modeling module. DETAILED DESCRIPTION

[0051] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.

[0052] See also Figures 1 to 3As shown, the reticle pattern correction method and correction system provided by the present invention are used to correct the design pattern of the reticle to ensure that after the pattern on the reticle is transferred to the wafer, the transferred pattern on the wafer conforms to the integrated circuit design. Pattern conformity to the integrated circuit design includes pattern size, pattern shape, and relationships between patterns. Pattern size can be reflected in the critical dimensions (CD) of patterns such as lines and contact holes designed on the reticle, such as line width and hole diameter. Pattern shape can be reflected in the geometric contours of the pattern, such as whether corners are right angles, whether line ends are blunt, and whether holes are round. Relationships between patterns can be reflected in the relative position and spacing between patterns, the alignment accuracy of patterns on different layers, such as the spacing consistency between dense patterns and isolated patterns, and the alignment accuracy between metal lines and vias. In this embodiment, main pattern A is the pattern designed by the designer on the reticle that needs to be transferred to the wafer. The auxiliary pattern B is related to the main pattern A and is used to balance the process window of the dense main pattern A and the sparse main pattern A. In this embodiment, the auxiliary pattern B is, for example, a sub-resolution assist feature (SRAF). Figure 1 The left half shows a dense main pattern A, while the right half shows a sparse main pattern A. Auxiliary patterns B are set between adjacent sparse main patterns A, so that each position on the wafer has a balanced light window. Figure 2 A schematic diagram of the layout pattern corresponding to a defective lithography technology is shown. Figure 2 In the figure, the area limited by the purple dotted box represents the ideal design graphic, and the area of ​​the red dotted box represents the auxiliary graphic B on the mask. The blue line can represent the actual graphic on the corrected mask, that is, the main graphic A. The black line area within the purple dotted box is used to represent the graphic shape that actually appears on the wafer. By setting the auxiliary graphic balance lithography window, it can help improve the overlap between the black line and the purple dotted box. The green line area pointed by the red arrow represents a transfer error, and the auxiliary graphic B is transferred to the wafer. The correction method and correction system of the system of the present invention can avoid such problems. Figure 2 and Figure 3 As shown, a defect occurs when the auxiliary pattern B is transferred onto the wafer.

[0053] See also Figure 1 and Figure 4 As shown, the present invention provides a method for correcting patterns of a mask, and the correction method includes steps S100 to S400.

[0054] Step S100 , setting a plurality of main patterns A and a plurality of auxiliary patterns B, and collecting wafer data after photolithography of the main patterns A and the auxiliary patterns B.

[0055] Step S200 : Searching for a main pattern A having the same feature size as the auxiliary pattern B on the mask as a reference pattern.

[0056] Step S300 , obtaining a scaling ratio of the reference pattern transferred from the mask to the wafer, and obtaining virtual size data of the auxiliary pattern B transferred to the wafer according to the scaling ratio and size information of the auxiliary pattern B on the mask.

[0057] Step S400 : Using the virtual size data and the wafer data as modeling parameters of an optical proximity correction model, the optical proximity correction model is iteratively established until the correction error is less than a threshold.

[0058] See also Figure 1 、 Figure 4 and Figure 5 As shown, in one embodiment of the present invention, in step S100, a test mask and a test wafer are set, wherein the test data of the test mask and the test wafer will be used to collect data for establishing an optical proximity correction model. Specifically, a plurality of main graphics A and a plurality of auxiliary graphics B are set on the test mask. In this embodiment, the set main graphics A can call the graphic size data to be used in actual production, and the size range of the plurality of main graphics A traverses the graphic size data to be used in actual production to simulate the production situation as much as possible. It should be noted that the auxiliary graphics B added by the present invention is a rule-based auxiliary graphics B (ruleS-based SRAF). The present invention does not limit the rules for adding the auxiliary graphics B. The method for adding the auxiliary graphics B can be any rule in the prior art. In this embodiment, setting the widths of the plurality of auxiliary graphics B includes steps S110 to S130.

[0059] Step S110 : According to the optical signal intensity of the auxiliary pattern B during the photolithography process, the pattern width when the optical signal intensity is close to the exposure threshold is used as the visible width of the auxiliary pattern B.

[0060] Step S120: Set a test width range, where the test width range covers the visible width.

[0061] Step S130 : setting a plurality of auxiliary patterns B, wherein the widths of the plurality of auxiliary patterns B traverse a test width range.

[0062] See also Figure 1 、 Figures 4 to 6 As shown, in one embodiment of the present invention, in step S110, as Figure 6As shown in the figure, the curve represents the reflected light signal intensity of auxiliary pattern B during the exposure process. The horizontal line represents the exposure threshold of the exposed pattern, and the filled-in box area corresponds to the pattern type. In the area of ​​auxiliary pattern B, where the curve intersects or is about to intersect the horizontal line representing the exposure threshold, the light signal intensity at that point is set as the exposure intensity threshold of the light signal of auxiliary pattern B. When the light signal intensity reaches the exposure intensity threshold, auxiliary pattern B can be transferred to the wafer. Therefore, in this embodiment, the width of auxiliary pattern B where the light signal intensity is close to the exposure threshold is used as the visible width of auxiliary pattern B. Light signal intensity close to the exposure threshold means that the light signal intensity is equal to the exposure intensity threshold or the difference between the light signal intensity and the exposure intensity threshold is within the threshold range. In step S120, the test width range is not limited as long as it includes the visible width. In another embodiment of the present invention, the upper limit of the test width range is the modeling standard width plus, for example, 10 nm. The modeling standard width is, for example, a parameter standard of the optical proximity correction model and can be either an empirical parameter or a set parameter.

[0063] See also Figure 1 、 Figures 4 to 6 As shown, in one embodiment of the present invention, in step S130, multiple auxiliary graphics B are added to the test mask, where the widths of the multiple auxiliary graphics B traverse the test width range. In this embodiment, a minimum interval can be set, such as 1nm, requiring the width of the auxiliary graphics B to start from the minimum value of the test width range, and at least one corresponding auxiliary graphics B is set at each minimum interval. Specifically, for example, if the test width range is 20nm~30nm, then at least one auxiliary graphics B with a corresponding graphic width is set at each integer value of 20nm, 21nm, 22nm, and up to 30nm. In this embodiment, after setting and adding the auxiliary graphics B, the graphics on the test mask are transferred to the test wafer by photolithography, and then the graphic data transferred on the test wafer are collected as wafer data. The wafer data includes the graphic size data, graphic position data, and graphic shading data of the main graphic A after transfer, and the graphic position data and graphic shading data of the auxiliary graphics B.

[0064] See also Figure 1 、 Figures 4 to 10 As shown in the figure, the size data of the auxiliary pattern B after transfer cannot be directly measured. For clearly distinguishable patterns, the auxiliary pattern B can be distinguished by the naked eye. Figure 7 and Figure 10There is a clear distinction between these two types of transferred graphics. In this embodiment, the area calculation algorithm can be used to export the graphics and calculate the graphic area that appears in the transfer area corresponding to the auxiliary graphic B. The tool for calculating the area is, for example, the Image Contour extension tool. According to the calculated transfer area, the larger the transfer area, the higher the image exposure of the auxiliary graphic B. It should be noted that the transfer area is not equal to the graphic size data of the auxiliary graphic B after transfer. In theory, the auxiliary graphic B should not be transferred to the wafer, so theoretically the graphic size data of the auxiliary graphic B after transfer does not exist. In fact, the graphic size data of the auxiliary graphic B after transfer cannot be measured. In this embodiment, wafer data is obtained by monitoring and measurement. Specifically, wafer data can be obtained by a scanning electron microscope (SEM).

[0065] See also Figure 1 and Figure 4 As shown, in one embodiment of the present invention, in step S200, the feature size includes the same pattern spacing size. Figure 1 As shown, the distance between two main patterns A is D1, and the distance between a main pattern A and an auxiliary pattern B is also D1. Therefore, one of the two main patterns A can be used as a reference pattern. It should be noted that the number of patterns on a test reticle is extremely large, so the present invention does not explicitly require multiple reference patterns for a single auxiliary pattern B; it only requires one reference pattern that meets the requirements. In other embodiments of the present invention, multiple reference patterns can be found, and the average data of these reference patterns can be used as the reference data for the reference pattern.

[0066] See also Figure 1 and Figure 4 As shown, in one embodiment of the present invention, in step S300, the step of obtaining virtual size data is based on formula (1).

[0067] ADI CD2=(ADI CD1 / mask CD1)×mask CD2(1).

[0068] In formula (1), ADI CD2 is the virtual dimension data, ADI CD1 is the measured dimension of the main pattern A after it is transferred to the wafer, mask CD1 is the designed dimension of the main pattern A on the test mask, and mask CD2 is the designed dimension of the auxiliary pattern B on the test mask. The measured dimension of the main pattern A after it is transferred to the wafer can be obtained through measurement. The designed dimensions of the main pattern A on the test mask and the designed dimensions of the auxiliary pattern B on the test mask are both known dimensions. Therefore, according to formula (1), the virtual dimension data of the auxiliary pattern B can be obtained. The virtual dimension data represents the dimension data of the auxiliary pattern B if it is transferred to the test wafer.

[0069] See also Figure 1 、 Figure 4 、 Figures 7 to 10 As shown, in one embodiment of the present invention, after obtaining the virtual size data, the virtual size data of multiple auxiliary graphics B are adjusted until the virtual size data is positively correlated with the image exposure degree of the auxiliary graphics B, the width of the auxiliary graphics B, and the width of the main graphics A. In this embodiment, in the step of adjusting the virtual size data of the auxiliary graphics B, an electron microscope image of the auxiliary graphics B is obtained, and the occupied area data of the auxiliary graphics B on the electron microscope image is output, wherein the occupied area data is used to represent the image exposure degree, and the occupied area data is positively correlated with the image exposure degree. Figure 7 and Figure 10 As shown, Figure 7 The image of the lowest exposure, and Figure 10 The image exposure level is the highest. It should be noted that the image exposure level of each auxiliary graphic B can be sorted to achieve the highest modeling accuracy. In other embodiments of the present invention, sorting can also be performed by exposure level to improve data modeling efficiency and reduce the probability of data inconsistencies. For example, the virtual size data of auxiliary graphics B at level 1 is smaller than the virtual size data of auxiliary graphics B at level 2. Each level corresponds to a different exposed area of ​​auxiliary graphics B, and as the level increases, the occupied area increases and the image exposure level rises.

[0070] See also Figure 1 、 Figure 4 、 Figures 7 to 10As shown, in one embodiment of the present invention, in the step of adjusting the virtual size data of the auxiliary graphic B, the virtual size data is adjusted using the width of the auxiliary graphic B and the main graphic A on the test mask as a reference size. In this embodiment, the larger the width of the auxiliary graphic B, the larger the corresponding virtual size data. The larger the width of the reference graphic, the larger the virtual size data. Similar to image exposure accuracy, the image exposure level of each auxiliary graphic B can be ranked to achieve the highest modeling accuracy. In other embodiments of the present invention, the width range of the auxiliary graphic B and the width range of the reference graphic can also be graded according to the width range, thereby improving data modeling efficiency and reducing the probability of data inconsistencies. The width range of each level is different, and as the level increases, the graphic width increases, and the virtual size data of the auxiliary graphic B of the corresponding level increases.

[0071] See also Figure 1 、 Figure 4 and Figure 11 As shown, in one embodiment of the present invention, in step S400, the step of establishing an optical proximity correction model is based on optical correction performed on the wafer. During the photolithography process, photolithography is a multi-field coupling process involving optics, chemistry, and mechanics. A single model cannot cover all sources of distortion. Therefore, the initial model includes multiple sub-models, such as an optical interaction sub-model, a pattern density correlation sub-model, a size-spacing coupling sub-model, an auxiliary feature optimization sub-model, and a process fluctuation sub-model. The optical interaction sub-model is used to calculate the light intensity interference and diffraction effects between adjacent patterns. The pattern density correlation sub-model is used to address the impact of local pattern density on exposure uniformity. The size-spacing coupling sub-model is used to establish the joint influence of pattern size and spacing. The auxiliary feature optimization sub-model is used to automatically generate a layout plan for auxiliary pattern B. The process fluctuation sub-model is used to evaluate the impact of process parameter fluctuations on the proximity effect. It should be noted that the optical proximity correction model provided by the present invention is not limited to the sub-models mentioned above. In this embodiment, step S400 includes steps S410 to S450.

[0072] Step S410 : obtaining an initial generation model of an optical proximity correction model, wherein parameters involved in the initial generation model include virtual size data of the auxiliary pattern B.

[0073] Step S420: Set parameter value ranges for parameters involved in the initial generation model, and take values ​​from the parameter value ranges to obtain the initial generation model with determined parameters.

[0074] Step S430 , obtaining simulated lithography data of the mask pattern size data after being processed by the primary model.

[0075] Step S440 , obtaining the deviation value of the simulated lithography data and the wafer data, taking the first-generation model with the smallest deviation value as the second-generation model, and determining whether the deviation value of the second-generation model is less than a preset threshold.

[0076] Step S450: Adjust the parameter value range and the model framework combination of the first-generation model, and loop through steps S410 to S440 until the deviation value of the second-generation model is less than a preset threshold.

[0077] See also Figure 1 、 Figure 4 and Figure 11 As shown, in one embodiment of the present invention, based on error rates or empirical rules, highly important sub-models are selected and combined to obtain a primary model of the optical proximity correction model. The importance of the sub-models is determined by the developer. In step S410, the primary model is formed to include multiple parameters, including the virtual size data of the auxiliary figure B as one of the parameters of the primary model. In step S420, the process of selecting values ​​from the parameter value range can be performed in multiple rounds. For example, the parameter selection process of the primary model is illustrated using three parameters k1, k2, and k3 as an example. Parameters k1, k2, and k3 each have a parameter value range. In each round of selection, the values ​​of parameters k2 and k3 can be determined while parameter k1 varies, and so on. The value selection process is repeated until, after multiple rounds of selection, all available parameter combinations are traversed or the secondary model satisfies the deviation value less than a preset threshold. In step S430 , the pattern size data and pattern position data of the main pattern A and the auxiliary pattern B on the test mask are processed by the primary model, thereby simulating the transfer of the pattern on the test mask to simulated lithography data on the wafer.

[0078] See also Figure 1 、 Figure 4 and Figure 11As shown, in one embodiment of the present invention, in step S440, multiple first-generation models are generated in the same round. The deviation values ​​of the multiple first-generation models are compared, and the first-generation model with the smallest deviation value is selected as the second-generation model. A determination is then made as to whether the deviation value of the second-generation model is less than a preset threshold. If the deviation value of the second-generation model is less than the preset threshold, iteration is terminated, and the current second-generation model is used as the optical proximity correction model to be acquired for use in image correction of the reticle in production. If the deviation value of the second-generation model is greater than or equal to the preset threshold, the next round of parameter selection is performed, multiple first-generation models are generated again, and step S440 is continued to determine whether the deviation value of the second-generation model is less than the preset threshold. The process of selecting values, acquiring the second-generation model, and determining the deviation value is repeated until all available parameter value combinations are exhausted or the deviation value of the second-generation model is less than the preset threshold. Repeating steps S410 to S440 allows multiple second-generation models to be acquired. If, after exhausting all available parameter value combinations, the deviation value of the second-generation model is still greater than or equal to the preset threshold, step S450 is executed. In step S450, the parameter range can be adjusted multiple times, and steps S410 to S440 can be repeated. If a second-generation threshold that satisfies the conditions is still not obtained after multiple adjustments, the selected model framework combination can be adjusted. Repeat the adjustment until the deviation value of the second-generation model is less than the preset threshold. The second-generation model with a deviation value less than the preset threshold is used as the optical proximity correction model.

[0079] See also Figure 1 、 Figure 4 and Figure 12As shown, the present invention also provides a reticle pattern correction system 10, wherein the pattern correction system 10 includes a pattern correction system 10, a pattern database 11, a simulated lithography module 12, a pattern reference module 13, a data calculation module 14, a data adjustment module 15, and a modeling module 16. The pattern database 11 stores a plurality of main patterns A and a plurality of auxiliary patterns B. The pattern correction system 10 also includes a pattern retrieval module, which is connected to the pattern database 11 and can retrieve main patterns A and auxiliary patterns B with required dimensions from the pattern database 11 and input them into the simulated lithography module 12 through an input module. The simulated lithography module 12 is used to simulate the lithography process of transferring the pattern on the reticle to the wafer and obtain wafer data after the pattern transfer. The pattern reference module 13 is used to search for a main pattern A on the reticle with the same feature size as the auxiliary pattern B as a reference pattern. The data calculation module 14 is used to obtain the scale ratio of the reference pattern transferred from the reticle to the wafer. Based on the scale ratio and the size information of the auxiliary pattern B on the reticle, it obtains the virtual size data of the auxiliary pattern B transferred to the wafer. The data adjustment module 15 is used to adjust the virtual size data of each auxiliary pattern B based on the image exposure level of the auxiliary pattern B, the width of the auxiliary pattern B, and the width of the reference pattern. The modeling module 16 is used to use the virtual size data and wafer data as modeling parameters for the optical proximity correction model and iteratively establish the optical proximity correction model until the correction error is less than a threshold.

[0080] The unexpected technical effect of the graphic correction method and correction system based on the mask provided by the present invention is that: when the transfer of the auxiliary graphic cannot be quantitatively studied, the present invention can quantify the exposure degree of the auxiliary graphic through virtual size data, so that the auxiliary graphic is taken into account in the optical proximity correction to avoid the auxiliary graphic being transferred to the wafer. In addition, the present invention can dynamically adjust the virtual size data of the auxiliary graphic, so that the virtual size data can accurately reflect the degree to which the auxiliary graphic is exposed. The present invention can calibrate the auxiliary graphic by the visible width, thereby improving the parameter accuracy and parameter comprehensiveness of the wafer data and the virtual size data. The graphic correction method and correction system provided by the present invention are based on virtual data-driven modeling and multi-parameter collaborative optimization. While improving the photolithography process window through the auxiliary graphic, it can strictly avoid the auxiliary graphic from being developed on the wafer, thereby significantly improving the yield of the photolithography process.

[0081] The embodiments of the present invention disclosed above are intended only to illustrate the present invention. They do not describe all details in detail, nor do they limit the present invention to the specific embodiments described. Obviously, many modifications and variations are possible based on the content of this specification. These embodiments are selected and described in detail in this specification to better explain the principles and practical applications of the present invention, thereby enabling those skilled in the art to better understand and utilize the present invention. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. A method for correcting a pattern of a mask, characterized in that: The following steps are involved: Setting a plurality of main patterns and a plurality of auxiliary patterns, and collecting wafer data of the main patterns and the auxiliary patterns after photolithography; Searching for the main pattern having the same feature size as the auxiliary pattern on the mask as a reference pattern; Obtaining a scaling ratio of the reference pattern when it is transferred from the mask to the wafer, and obtaining virtual size data of the auxiliary pattern when it is transferred to the wafer according to the scaling ratio and size information of the auxiliary pattern on the mask; as well as The virtual size data and the wafer data are used as modeling parameters of an optical proximity correction model, and the optical proximity correction model is iteratively established until a correction error is less than a threshold.

2. The method for correcting a pattern of a mask according to claim 1, wherein: Setting the widths of the plurality of auxiliary graphics comprises the following steps: According to the optical signal intensity of the auxiliary pattern in the photolithography process, the pattern width when the optical signal intensity is close to the exposure threshold is used as the visible width of the auxiliary pattern; Setting a test width range, the test width range covering the display width; and A plurality of the auxiliary patterns are provided, wherein widths of the plurality of the auxiliary patterns traverse the test width range.

3. The method for correcting a pattern of a mask according to claim 1, wherein: The step of obtaining the virtual size data is based on the following formula: ADI CD2=(ADI CD1 / mask CD1)×mask CD2; Among them, ADI CD2 is the virtual size data, ADI CD1 is the measured size of the main pattern after being transferred to the wafer, mask CD1 is the designed size of the main pattern on the mask, and mask CD2 is the designed size of the auxiliary pattern on the mask.

4. The method for correcting a pattern of a mask according to claim 1, wherein: After obtaining the virtual size data, the virtual size data of the plurality of auxiliary patterns are adjusted until the virtual size data is positively correlated with the image exposure degree of the auxiliary pattern, the width of the auxiliary pattern, and the width of the reference pattern.

5. The method for correcting a pattern of a mask according to claim 4, wherein: In the step of adjusting the virtual size data of the auxiliary graphic, an electron microscope image of the auxiliary graphic is obtained, and the occupied area data of the auxiliary graphic on the electron microscope image is output, wherein the occupied area data is used to characterize the image exposure degree, and the occupied area data is positively correlated with the image exposure degree.

6. The method for correcting a pattern of a mask according to claim 4, wherein: In the step of adjusting the virtual size data of the auxiliary pattern, the virtual size data is adjusted with the widths of the auxiliary pattern and the reference pattern on the mask as reference sizes.

7. The method for correcting a pattern of a mask according to claim 1, wherein: The steps of establishing the optical proximity correction model include: Acquiring an initial model of the optical proximity correction model, wherein parameters involved in the initial model include the virtual size data of the auxiliary pattern; Setting parameter value ranges for parameters involved in the primary model, and taking values ​​from the parameter value ranges to obtain the primary model with determined parameters; Acquire simulated lithography data of the pattern size data of the mask after being processed by the primary model; Obtaining a deviation value between the simulated lithography data and the wafer data, taking the first-generation model with the smallest deviation value as the second-generation model, and determining whether the deviation value of the second-generation model is less than a preset threshold; and Adjust the parameter value range and the model framework combination of the first-generation model, and loop through the steps of obtaining the second-generation model until the deviation value of the second-generation model is less than the preset threshold.

8. The method for correcting a pattern of a mask according to claim 7, wherein: In the step of obtaining the second-generation model, values ​​are taken from the parameter value range in multiple rounds, and the model with the smallest deviation value is selected from the multiple first-generation models in each round as the second-generation model, until the deviation value of the second-generation model is less than the preset threshold.

9. The method for correcting a pattern of a mask according to claim 7, wherein: When the parameter value ranges of all parameters are traversed and the deviation value of the second-generation model is greater than or equal to the preset threshold, the model framework combination of the first-generation model is adjusted.

10. A mask pattern correction system, characterized in that: include: A graphic database storing a plurality of main graphics and a plurality of auxiliary graphics; A photolithography simulation module is used to simulate the photolithography process of transferring the pattern on the mask to the wafer and obtain wafer data after the pattern is transferred; a pattern reference module, configured to search on the mask for the main pattern having the same feature size as the auxiliary pattern as a reference pattern; a data calculation module, configured to obtain a scaling ratio of the reference pattern when it is transferred from the mask to the wafer, and obtain virtual size data of the auxiliary pattern when it is transferred to the wafer according to the scaling ratio and size information of the auxiliary pattern on the mask; as well as The modeling module is configured to use the virtual size data and the wafer data as modeling parameters of an optical proximity correction model, and iteratively establish the optical proximity correction model until a correction error is less than a threshold.

Images