Topography-aware optical proximity correction method, device, storage medium and electronic equipment

Abstract

The application discloses a topography-aware optical proximity correction method and device, a storage medium and an electronic device. The topography-aware optical proximity correction method comprises the following steps: obtaining an original design layout; performing topography-aware optical proximity correction on the original design layout by using a current TOPC model to generate an initial correction layout; performing in-situ measurement on a wafer manufactured based on the initial correction layout to obtain actual pattern data; comparing the actual pattern data with simulation pattern data generated based on the initial correction layout to generate model deviation data; analyzing the model deviation data by using a parameter correction model to generate a parameter correction amount; and updating the current TOPC model based on the parameter correction amount. The application can improve the accuracy of topography-aware optical proximity correction.

Description

Technical Field

[0001] This application relates to the field of photolithography technology, specifically to a method, apparatus, storage medium, and electronic device for correcting the optical proximity effect of terrain perception. Background Technology

[0002] In advanced semiconductor manufacturing, optical proximity is a key factor causing lithographic pattern distortion and affecting product yield. Optical Proximity Correction (OPC) technology compensates for this distortion by pre-correcting the mask pattern. As process nodes enter the nanoscale, the local influence of the wafer surface's three-dimensional topography on the lithography focus becomes significant, leading to the development of Topography-aware Optical Proximity Correction (TOPC) technology. TOPC predicts the topography through chemical-mechanical polishing (CMP) simulation and performs more accurate OPC corrections, becoming a core means of improving the process window for advanced processes.

[0003] However, due to the extreme complexity of semiconductor manufacturing processes, equipment conditions and process parameters change slowly over time (i.e., process drift). Current TOPC technology relies on static models and cannot perceive or adapt to these dynamic changes. This causes CMP topography prediction and OPC correction to gradually deviate from the actual process conditions, resulting in systematic pattern deviations and ultimately a continuous decline in product yield. Summary of the Invention

[0004] This application provides a method, apparatus, storage medium, and electronic device for correcting the optical proximity effect of terrain sensing, which can improve the accuracy of terrain sensing optical proximity effect correction.

[0005] In a first aspect, embodiments of this application provide a method for correcting the optical proximity effect in terrain sensing, including:

[0006] Obtain the original design layout;

[0007] The original design map is corrected for terrain-aware optical proximity effect using the current TOPC model to generate an initial corrected map.

[0008] In-situ measurements are performed on the wafer manufactured based on the initial modified layout to obtain actual pattern data;

[0009] The actual graphic data is compared with the simulation graphic data generated based on the initial corrected layout to generate model deviation data;

[0010] The model deviation data is analyzed using a parameter correction model to generate parameter correction values;

[0011] The current TOPC model is updated based on the parameter correction amount.

[0012] In the terrain-aware optical proximity effect correction method provided in this application embodiment, the current TOPC model includes a CMP module and an OPC module;

[0013] The step of using the current TOPC model to perform terrain-aware optical proximity effect correction on the original design map to generate an initial corrected map includes:

[0014] The original design layout was simulated using the CMP module to obtain the predicted morphology data of the wafer surface.

[0015] Based on the topography prediction data, the original design layout is divided into multiple out-of-focus zones;

[0016] For each defocused zone, the corresponding OPC module is invoked to perform optical proximity effect correction, generating an initial corrected layout.

[0017] In the terrain-aware optical proximity effect correction method provided in this application embodiment, the step of dividing the original design map into multiple defocus zones based on the topography prediction data includes:

[0018] Based on multiple preset discrete defocus threshold values, the defocus amount predicted for each local location in the shape prediction data is classified into the corresponding defocus interval.

[0019] All graphic units within the same out-of-focus area are assigned to the same out-of-focus partition, thereby dividing the original design layout into multiple out-of-focus partitions.

[0020] In the terrain-aware optical proximity effect correction method provided in this application embodiment, the step of dividing the original design map into multiple defocus zones based on the topography prediction data includes:

[0021] Acquire multiple discrete defocus states defined for the target process window, each of which is associated with a specific OPC module;

[0022] The defocus amount predicted for each local location in the topography prediction data is mapped to the corresponding discrete defocus state.

[0023] Based on the mapping results, the original design layout is divided into defocus partitions corresponding to each of the discrete defocus states.

[0024] In the terrain-aware optical proximity effect correction method provided in this application embodiment, the step of dividing the original design map into multiple defocus zones based on the topography prediction data includes:

[0025] Based on the morphology prediction data, surface depression areas and surface protrusion areas caused by the chemical mechanical polishing process are identified.

[0026] The surface depression region is divided into a first defocus zone associated with a first defocus amount;

[0027] The surface protrusion area is divided into a second defocus zone associated with the second defocus amount;

[0028] This divides the original design layout into multiple out-of-focus zones.

[0029] The terrain-aware optical proximity effect correction method provided in this application embodiment further includes, before obtaining the original design map:

[0030] Construct a parameter correction model.

[0031] In the terrain-sensing optical proximity effect correction method provided in this application embodiment, the construction of the parameter correction model includes:

[0032] Acquire historical production data, which includes historical design maps, corresponding historical topographic prediction data, historical measurement data, and historical model deviation data calculated based on the historical measurement data.

[0033] Training features are extracted from the historical production data to form a training sample set;

[0034] Based on the training sample set, a preset machine learning algorithm is trained to establish a mapping relationship from input features to model bias, thereby obtaining a parameter correction model.

[0035] Secondly, embodiments of this application provide a terrain-aware optical proximity effect correction device, comprising:

[0036] The layout acquisition unit is used to acquire the original design layout.

[0037] The map correction unit is used to perform terrain-aware optical proximity effect correction on the original design map using the current TOPC model, and generate an initial corrected map.

[0038] The in-situ measurement unit is used to perform in-situ measurements on the wafer manufactured based on the initial modified layout to obtain actual pattern data;

[0039] The data comparison unit is used to compare the actual graphic data with the simulation graphic data generated based on the initial corrected layout to generate model deviation data.

[0040] The data analysis unit is used to analyze the model deviation data using the parameter correction model and generate parameter correction values;

[0041] The model update unit is used to update the current TOPC model based on the parameter correction amount.

[0042] Thirdly, this application provides a storage medium storing a plurality of instructions adapted for loading by a processor to execute the terrain-aware optical proximity effect correction method described in any of the preceding claims.

[0043] Fourthly, this application provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the terrain-aware optical proximity effect correction method described in any of the preceding claims.

[0044] In summary, the terrain-aware optical proximity effect correction method provided in this application includes: acquiring an original design layout; applying a current TOPC model to the original design layout to correct for the terrain-aware optical proximity effect, generating an initial corrected layout; performing in-situ measurements on a wafer manufactured based on the initial corrected layout to obtain actual graphic data; comparing the actual graphic data with simulation graphic data generated based on the initial corrected layout to generate model deviation data; analyzing the model deviation data using a parameter correction model to generate parameter correction values; and updating the current TOPC model based on the parameter correction values. This application embodiment can improve the accuracy of terrain-aware optical proximity effect correction. Attached Figure Description

[0045] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0046] Figure 1 This is a schematic diagram illustrating an application scenario of the terrain-sensing optical proximity effect correction method provided in this application embodiment.

[0047] Figure 2 This is a schematic flowchart of the terrain-sensing optical proximity effect correction method provided in the embodiments of this application.

[0048] Figure 3This is a schematic diagram of the structure of the terrain-sensing optical proximity effect correction device provided in the embodiments of this application.

[0049] Figure 4 This is a schematic diagram of the structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0050] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0051] It should be noted that, in this document, 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 that element. Furthermore, components, features, and elements with the same names in different embodiments of this application may have the same meaning or different meanings, the specific meaning of which must be determined by its interpretation in that specific embodiment or further in conjunction with the context of that specific embodiment.

[0052] It should be understood that the specific embodiments described herein are merely illustrative of this application and are not intended to limit this application.

[0053] In the following description, the use of suffixes such as "module," "part," or "unit" to denote elements is solely for the purpose of illustrative purposes and has no specific meaning in itself. Therefore, "module," "part," or "unit" may be used interchangeably.

[0054] In the description of this application, it should be noted that the terms "upper," "lower," "left," "right," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this application. In addition, terms such as "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0055] Because semiconductor manufacturing processes are extremely complex, equipment conditions and process parameters change slowly over time (i.e., process drift). Current TOPC technology relies on static models and cannot perceive or adapt to these dynamic changes. This causes CMP topography predictions and OPC corrections to gradually deviate from the actual process conditions, resulting in systematic pattern deviations and ultimately a continuous decline in product yield.

[0056] Based on this, embodiments of this application provide a method, apparatus, storage medium, and electronic device for correcting the optical proximity effect of terrain perception. Specifically, the terrain perception optical proximity effect correction apparatus can be integrated into an electronic device, which can be a server or a terminal, etc. The terminal can include mobile phones, wearable smart devices, tablet computers, laptops, and personal computers (PCs), etc. The server can be a single server or a server cluster composed of multiple servers, and can be a physical server or a virtual server.

[0057] For example, such as Figure 1 As shown, the electronic device can acquire the original design layout; use the current TOPC model to correct the terrain-aware optical proximity effect of the original design layout to generate an initial corrected layout; perform in-situ measurements on the wafer manufactured based on the initial corrected layout to obtain actual graphic data; compare the actual graphic data with the simulation graphic data generated based on the initial corrected layout to generate model deviation data; analyze the model deviation data using the parameter correction model to generate parameter correction values; and update the current TOPC model based on the parameter correction values.

[0058] The technical solutions shown in this application will be described in detail below through specific embodiments. It should be noted that the order of description of the following embodiments is not intended to limit the priority of the embodiments.

[0059] Please see Figure 2 , Figure 2 This is a schematic flowchart of the terrain-sensing optical proximity effect correction method provided in an embodiment of this application. The specific flow of the terrain-sensing optical proximity effect correction method can be as follows:

[0060] 101. Obtain the original design layout.

[0061] The original design layout is a two-dimensional planar graphic data file that defines the physical structure of one or more layers of an integrated circuit. This original design layout fully describes the geometry, dimensions, and relative positions of all functional units in the chip, including transistors, interconnects, and vias. In the semiconductor industry, original design layouts typically use industry-standard data formats such as GDSII or OASIS.

[0062] In some embodiments, the original design layout can be read from local or network storage devices, such as directly connected hard drives, network-attached storage, or storage area networks, from layout files pre-generated and stored by the design department. In some embodiments, the original design layout can be retrieved from a design database, for example, through an application programming interface or database query command, from an integrated circuit design database.

[0063] 102. Use the current TOPC model to correct the terrain-aware optical proximity effect on the original design map and generate an initial corrected map.

[0064] The current TOPC model is a dynamically updated, comprehensive computational model for performing terrain-aware optical proximity effect correction. The current TOPC model mainly includes a CMP module and a model library consisting of several OPC modules. The CMP module is a mathematical model based on physical mechanisms or empirical data. Its input is the graphic layout of the design pattern (e.g., pattern density), and its output is predicted three-dimensional topography data of the wafer surface (e.g., surface height or thickness distribution). Each OPC module is calibrated for a specific process condition (e.g., different defocus amounts) to perform accurate correction calculations for the pattern under specific defocus conditions.

[0065] In this embodiment, when the system begins processing the first production batch, the "current TOPC model" refers to the initial static model obtained during the process development phase using traditional offline calibration methods. Starting from the second batch, the "current TOPC model" refers to the optimized model that has undergone steps 105 and 106, and has been incrementally updated using parameter corrections, making it more compatible with the latest process state.

[0066] In this embodiment of the application, the step "using the current TOPC model to correct the terrain-aware optical proximity effect of the original design map and generate an initial corrected map" may include:

[0067] 1021. Use the CMP module to perform chemical mechanical polishing simulation on the original design layout to obtain the morphology prediction data of the wafer surface.

[0068] Understandably, the CMP module is a mathematical model built upon physical mechanisms or empirical data. It receives the original design layout obtained in step 101 as input and predicts the morphology of the wafer surface after manufacturing by simulating the dynamic process of chemical mechanical polishing.

[0069] Specifically, this CMP module can calculate the amount of material removed or the remaining thickness at each location on the original design layout after the CMP process by solving a series of empirical formulas or physical equations (such as the Preston equation). Finally, it outputs topography prediction data covering the entire chip, typically represented as a two-dimensional height or thickness distribution map. This topography prediction data can quantify surface depressions (such as erosion effects) and surface bumps (such as dish-shaped depression effects) caused by the CMP process.

[0070] 1022. Based on the topography prediction data, the original design layout is divided into multiple out-of-focus zones.

[0071] In this embodiment, continuous topography prediction data can be transformed into discrete regions that can be processed by different OPC modules. The core is to convert the predicted topography prediction data into local defocusing amounts in the photolithography process and partition it, thereby dividing the original design layout into multiple defocusing zones.

[0072] In the specific implementation process, any one of the three division methods can be used to divide the original design map: division based on a preset threshold, division based on a process window mapping method, or division based on morphological features.

[0073] When the division method is based on a preset threshold, specifically, according to multiple preset discrete defocus value thresholds, the predicted defocus amount of each local position in the topography prediction data is classified into the corresponding defocus interval; all layout graphic units within the same defocus interval are divided into the same defocus partition, thereby dividing the original design layout into multiple defocus partitions.

[0074] For example, a set of discrete defocus thresholds can be pre-stored (e.g., -15nm, -5nm, +5nm, +15nm). This allows each local location in the topography prediction data to be categorized into a specific defocus interval defined by these thresholds (e.g., interval A where the defocus is between -15nm and -5nm) based on its predicted defocus amount. Subsequently, all layout graphic units falling within the same defocus interval are merged into the same defocus partition.

[0075] When the partitioning method is a mapping partitioning method based on the process window, it can specifically be to obtain multiple discrete defocus states defined for the target process window, with each discrete defocus state associated with a specific OPC module; map the defocus amount predicted at each local position in the topography prediction data to the corresponding discrete defocus state; and divide the original design layout into defocus partitions corresponding to each discrete defocus state based on the mapping results.

[0076] For example, multiple discrete defocus states defined for the lithography process window can be obtained first (e.g., optimal focus, positive defocus +10nm, negative defocus -10nm), where each discrete defocus state is associated with a specific, specially calibrated OPC module. Then, the predicted defocus amount at each local location is mapped to the discrete defocus state closest to it. Finally, the original design layout is divided into defocus partitions corresponding one-to-one with these discrete defocus states.

[0077] When the division method is based on morphological features, it can specifically be based on morphological prediction data to identify surface depression areas and surface protrusion areas caused by chemical mechanical polishing; divide the surface depression areas into a first defocus zone associated with a first defocus amount; divide the surface protrusion areas into a second defocus zone associated with a second defocus amount; thereby dividing the original design layout into multiple defocus zones.

[0078] The core of this partitioning method lies in directly dividing the area based on the morphological features with clear physical significance resulting from the CMP process, rather than simply relying on the numerical threshold of defocusing. For example, suppose the original design layout contains two typical graphic layout areas: Region A (dense graphic area) and Region B (sparse graphic area). Region A consists of a large number of densely arranged metal lines, with a local graphic density exceeding 80%. Region B contains an isolated metal line with no other graphics around it, and a local graphic density below 20%. After performing the CMP simulation in step 1021, Region A, due to its large material area, will experience stronger mechanical abrasion and chemical corrosion during the CMP process, resulting in a faster material removal rate and ultimately forming a surface depression area below the average plane (i.e., the CMP erosion effect). In Region B, the vast empty area between the isolated lines makes the material removal rate slower, and compared to the depression area, the isolated lines themselves appear more convex, forming a surface convex area (i.e., the CMP dish-shaped depression effect).

[0079] By analyzing the height / thickness values ​​in the topography prediction data, all continuous regions below the preset average height can be automatically identified and marked as "surface depression regions"; simultaneously, all continuous regions above the preset average height are identified and marked as "surface bulge regions". The identified surface depression regions can then be divided into a first defocus zone. In the photolithography imaging model, depression means an increased distance between the bottom of the photoresist and the ideal focal plane; therefore, this first defocus zone is associated with a negative defocus amount (e.g., -15nm). The identified surface bulge regions are divided into a second defocus zone. Bulge means the bottom of the photoresist is closer to the exposure lens; therefore, this second defocus zone is associated with a positive defocus amount (e.g., +10nm).

[0080] 1023. Call the corresponding OPC module for each out-of-focus zone to perform optical proximity effect correction and generate an initial correction pattern.

[0081] Specifically, from the current TOPC model library, an OPC module corresponding to the defocus amount can be matched and invoked for each defocus zone. Then, using the matched OPC module, optical proximity correction is performed on all graphics within that defocus zone. After the optical proximity correction for all defocus zones is completed, the independently corrected graphic data for each defocus zone can be merged to synthesize a complete initial corrected layout.

[0082] 103. Perform in-situ measurements on the wafers manufactured based on the initial modified layout to obtain actual graphic data.

[0083] First, the initial revised layout can be transferred to photolithography equipment. Through a series of semiconductor manufacturing processes such as photolithography, development, and etching, the original design layout is transferred onto a physical wafer, forming a wafer with actual physical characteristics. Then, using in-situ metrology tools tightly integrated with the production line (online), key pattern structures on the wafer can be measured non-destructively in real time, thereby obtaining actual pattern data.

[0084] The actual graphic data includes, but is not limited to, actual key dimension values ​​at multiple locations, three-dimensional contour parameters of specific graphics, overlay error values ​​between different layers, and layout position identifiers corresponding to each data point.

[0085] 104. Compare the actual graphic data with the simulation graphic data generated based on the initial revised layout to generate model deviation data.

[0086] First, the actual graphic data and the simulated graphic data can be aligned. For example, using the layout coordinates of each measured data point in the actual graphic data, the simulated predicted value at the same coordinate position can be found in the simulated graphic data. For instance, find the actual linewidth measured by CD-SEM at coordinates (X1, Y1), and the simulated linewidth predicted at that position.

[0087] After data alignment is completed, a series of calculation functions can be executed to compare the measured and predicted values ​​and generate quantified model deviation data. This model deviation data includes, but is not limited to, critical dimension deviations, edge placement errors, and overlay errors.

[0088] 105. Analyze the model deviation data using the parameter correction model to generate parameter correction values.

[0089] The parameter correction model is a pre-trained machine learning model. That is, the parameter correction model needs to be built before step 101.

[0090] Specifically, the parameter calibration model can be constructed as follows:

[0091] 1051. Obtain historical production data, which includes historical design maps, corresponding historical topographic prediction data, historical measurement data, and historical model deviation data calculated based on historical measurement data.

[0092] Historical design layouts refer to the original design layout files used in past production. Historical topography prediction data refers to the topography prediction results obtained by simulating these historical design layouts using the CMP module used at the time. Historical measurement data refers to the actual graphic data obtained by in-situ measurement based on the wafers produced at that time. Historical model deviation data refers to the model deviation data calculated by comparing the historical measurement data with the model prediction values ​​at that time.

[0093] 1052. Extract training features from historical production data to form a training sample set.

[0094] The training features can include layout geometric features (such as local pattern density, linewidth, spacing, pattern orientation, perimeter to area ratio, etc.), topographic features (such as CMP-predicted thickness, height gradient, surface curvature, etc.), spatial location features (such as X and Y coordinates on the wafer, used to capture spatially related process changes), and contextual features (such as the defocusing partition, measurement tool identifier, production timestamp, etc.). Each training feature is paired with its corresponding historical model bias data to form a training sample. All training samples form a training sample set for machine learning training.

[0095] 1053. Based on the training sample set, the preset machine learning algorithm is trained to establish a mapping relationship from input features to model bias, and a parameter correction model is obtained.

[0096] Using a pre-constructed training sample set, a pre-defined machine learning algorithm can be trained in a supervised manner. The training process allows the machine learning algorithm to learn and establish a complex nonlinear mapping relationship between input features and output model bias data. After training, a parameter-corrected model capable of predicting model bias data from input features is obtained. This machine learning algorithm includes, but is not limited to, neural networks, random forests, support vector machines, or Gaussian process regression.

[0097] For model bias data, input features consistent with those from the training phase can be extracted from the corresponding layout coordinates to form the current feature vector. This current feature vector is then input into the parameter calibration model. Based on its knowledge learned from historical data, the model analyzes the correlation between the current bias and various layout and morphological features, thereby inferring the root cause of the bias. For example, the parameter calibration model might identify: "In areas with low pattern density and large CMP predicted thickness, a systematic positive CD bias (measured values ​​are too large) has occurred." Based on the inference of the root cause, the parameter calibration model, through internal calculations (such as gradient backpropagation, sensitivity analysis, or optimization algorithms), can deduce which parameters in the current TOPC model need adjustment in what direction, most effectively eliminating the identified bias patterns. Finally, it outputs specific, quantified parameter correction amounts.

[0098] 106. Update the current TOPC model based on the parameter correction amount.

[0099] The parameter correction amount is a series of structured instructions that clearly indicate the module object to be adjusted, the specific parameter name, and the value and direction of the adjustment.

[0100] Specifically, based on the parameter correction amount, the CMP module and / or a specific OPC module in the current TOPC model can be accessed, and the specified parameter can be incrementally updated. Incremental update refers to adding or subtracting a small correction amount from the current value of the parameter, rather than directly replacing it with a new value. This ensures the smoothness and stability of the model update, avoiding the risk of oscillations introduced by excessively large single update magnitudes. After the update is complete, a new version identifier or timestamp can be added to the updated model, and it can be set as the current TOPC model.

[0101] In this embodiment, for each subsequent production batch or at preset time intervals, the complete process from step 101 to step 106 is automatically repeated, forming a continuous "correction-measurement-learning-update" closed loop. This cyclical operation mode ensures that the TOPC model is no longer a static offline tool, but a dynamic adaptive model that can continuously operate online. This cyclical operation mode can capture process drift during production in real time, and continuously optimize the parameters of its internal CMP and OPC modules through machine learning analysis of real-time measurement data. Thus, without manual intervention, the accuracy of the TOPC model can remain synchronized with the current actual process state, thereby continuously ensuring optimal graphic fidelity and production yield throughout the entire product lifecycle.

[0102] In summary, the terrain-aware optical proximity effect correction method provided in this application includes: acquiring the original design layout; using the current TOPC model to correct the terrain-aware optical proximity effect of the original design layout, generating an initial corrected layout; performing in-situ measurements on the wafer manufactured based on the initial corrected layout to obtain actual graphic data; comparing the actual graphic data with the simulation graphic data generated based on the initial corrected layout to generate model deviation data; analyzing the model deviation data using a parameter correction model to generate parameter correction values; and updating the current TOPC model based on the parameter correction values. This application, by constructing a closed-loop adaptive system of "correction-measurement-learning-update," can automatically diagnose and correct model deviations using real-time measurement data from the production line. This enables the process model (including the CMP module and OPC module) upon which the terrain-aware optical proximity effect correction depends to continuously track and adapt to the drift of the actual process state, thereby dynamically maintaining and continuously improving correction accuracy throughout the entire product lifecycle. This effectively overcomes the accuracy degradation problem caused by process drift in traditional static models, ultimately ensuring graphic fidelity and product yield. In other words, the embodiments of this application can improve the accuracy of terrain-aware optical proximity effect correction.

[0103] To facilitate better implementation of the terrain-sensing optical proximity effect correction method provided in this application, this application also provides a terrain-sensing optical proximity effect correction device. The meanings of the terms used are the same as in the terrain-sensing optical proximity effect correction method described above, and specific implementation details can be found in the descriptions within the method embodiments.

[0104] Please see Figure 3 , Figure 3 This is a schematic diagram of the structure of the terrain-sensing optical proximity effect correction device provided in an embodiment of this application. The terrain-sensing optical proximity effect correction device may include a map acquisition unit 201, a map correction unit 202, an in-situ measurement unit 203, a data comparison unit 204, a data analysis unit 205, and a model update unit 206.

[0105] The layout acquisition unit 201 is used to acquire the original design layout;

[0106] The map correction unit 202 is used to perform terrain-aware optical proximity effect correction on the original design map using the current TOPC model, and generate an initial corrected map.

[0107] The in-situ measurement unit 203 is used to perform in-situ measurements on the wafer manufactured based on the initial modified layout to obtain actual pattern data;

[0108] The data comparison unit 204 is used to compare the actual graphic data with the simulation graphic data generated based on the initial corrected layout to generate model deviation data.

[0109] Data analysis unit 205 is used to analyze model deviation data using a parameter correction model and generate parameter correction values;

[0110] Model update unit 206 is used to update the current TOPC model based on parameter correction.

[0111] For specific implementation methods of each of the above units, please refer to the embodiments of the terrain-aware optical proximity effect correction method described above, which will not be repeated here.

[0112] In summary, the terrain-aware optical proximity effect correction device provided in this application embodiment can acquire the original design layout through the layout acquisition unit 201; the layout correction unit 202 uses the current TOPC model to perform terrain-aware optical proximity effect correction on the original design layout to generate an initial corrected layout; the in-situ measurement unit 203 performs in-situ measurement on the wafer manufactured based on the initial corrected layout to obtain actual graphic data; the data comparison unit 204 compares the actual graphic data with the simulation graphic data generated based on the initial corrected layout to generate model deviation data; the data analysis unit 205 analyzes the model deviation data using the parameter correction model to generate parameter correction values; and the model update unit 206 updates the current TOPC model based on the parameter correction values. This application embodiment constructs a closed-loop adaptive system of "correction-measurement-learning-update," which can automatically diagnose and correct model deviations using real-time measurement data from the production line. This enables the process model (including the CMP and OPC modules) upon which the terrain-aware optical proximity effect correction relies to continuously track and adapt to the drift of the actual process state, thereby dynamically maintaining and continuously improving correction accuracy throughout the entire product lifecycle. This effectively overcomes the accuracy degradation problem caused by process drift in traditional static models, ultimately ensuring image fidelity and product yield. In other words, this application embodiment can improve the accuracy of terrain-aware optical proximity effect correction.

[0113] This application also provides an electronic device that may integrate the terrain-aware optical proximity effect correction device of this application, such as... Figure 4 As shown, it illustrates a structural schematic diagram of the electronic device involved in the embodiments of this application, specifically:

[0114] The electronic device may include components such as a processor 301 with one or more processing cores and a memory 302 with one or more computer-readable storage media. Those skilled in the art will understand that... Figure 4 The electronic device structure shown does not constitute a limitation on the electronic device and may include more or fewer components than shown, or combine certain components, or have different component arrangements. Wherein:

[0115] The processor 301 is the control center of the electronic device. It connects various parts of the electronic device via various interfaces and lines. By running or executing software programs stored in the memory 302 and / or this application, and by calling data stored in the memory 302, it performs various functions and processes data, thereby providing overall monitoring of the electronic device. Optionally, the processor 301 may include one or more processing cores; preferably, the processor 301 may integrate an application processor and a modem processor, wherein the application processor mainly handles the operation of the storage medium, user interface, and application programs, while the modem processor mainly handles wireless communication. It is understood that the modem processor may not be integrated into the processor 301.

[0116] The memory 302 can be used to store software programs and this application. The processor 301 executes various functional applications and data processing by running the software programs and this application stored in the memory 302. The memory 302 may mainly include a program storage area and a data storage area. The program storage area may store applications required for operating the storage medium and at least one function; the data storage area may store data created based on the use of the electronic device. In addition, the memory 302 may include high-speed random access memory and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other volatile solid-state storage device. Accordingly, the memory 302 may also include a memory controller to provide the processor 301 with access to the memory 302.

[0117] Although not shown, the electronic device may also include a display unit, an input unit, and a power supply, etc., which will not be described in detail here. Specifically, in this embodiment, the processor 301 in the electronic device loads the executable files corresponding to the processes of one or more application programs into the memory 302 according to the following instructions, and the processor 301 runs the application programs stored in the memory 302 to realize various functions, as follows:

[0118] Obtain the original design layout;

[0119] The original design map is corrected for terrain-aware optical proximity effect using the current TOPC model to generate an initial corrected map.

[0120] In-situ measurements were performed on the wafers manufactured based on the initial modified layout to obtain the actual pattern data;

[0121] The actual graphic data is compared with the simulation graphic data generated based on the initial revised layout to generate model deviation data;

[0122] The parameter correction model is used to analyze the model bias data and generate parameter correction values.

[0123] The current TOPC model is updated based on the parameter correction.

[0124] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be performed by instructions, or by instructions controlling related hardware. These instructions can be stored in a computer-readable storage medium and loaded and executed by a processor.

[0125] Therefore, embodiments of this application provide a storage medium storing a plurality of instructions that can be loaded by a processor to execute steps in any of the methods provided in embodiments of this application. For example, the instructions can execute the following steps:

[0126] Obtain the original design layout;

[0127] The original design map is corrected for terrain-aware optical proximity effect using the current TOPC model to generate an initial corrected map.

[0128] In-situ measurements were performed on the wafers manufactured based on the initial modified layout to obtain the actual pattern data;

[0129] The actual graphic data is compared with the simulation graphic data generated based on the initial revised layout to generate model deviation data;

[0130] The parameter correction model is used to analyze the model bias data and generate parameter correction values.

[0131] The current TOPC model is updated based on the parameter correction.

[0132] For details on the implementation of each of the above operations, please refer to the previous examples, which will not be repeated here.

[0133] The storage medium may include: read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0134] Since the instructions stored in the storage medium can execute the steps of any method provided in the embodiments of this application, the beneficial effects that any method provided in the embodiments of this application can achieve can be realized. For details, please refer to the previous embodiments, which will not be repeated here.

[0135] The above provides a detailed description of the terrain-aware optical proximity effect correction method, device, storage medium, and electronic device provided in this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the core ideas of this application. At the same time, those skilled in the art will recognize that there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for correcting optical proximity effect in terrain sensing, characterized in that, include: S101. Obtain the original design layout; S102. Use the current TOPC model to perform terrain-aware optical proximity effect correction on the original design map to generate an initial corrected map; wherein, the current TOPC model includes a CMP module and an OPC module; S103. Perform in-situ measurements on the wafer manufactured based on the initial modified layout to obtain actual pattern data; S104. Compare the actual graphic data with the simulation graphic data generated based on the initial corrected layout to generate model deviation data; S105. Analyze the model deviation data using a parameter correction model to generate parameter correction values; the analysis includes: associating the model deviation data with layout features and topographic features to infer the root cause of the deviation, and generating parameter correction values ​​for the CMP module and / or for the OPC module based on the inference results. S106. Update the current TOPC model online based on the parameter correction amount, including: performing online incremental updates on the CMP module and / or OPC module in the current TOPC model, and using the updated TOPC model as the new current TOPC model; Repeat steps S101-S106 above to continuously update the current TOPC model during the production process based on the in-situ measurement data of subsequent batches. This is used to correct the terrain-aware optical proximity effect of the original design map of subsequent batches, thereby forming a closed-loop adaptive correction process from correction, manufacturing, measurement, learning to re-correction.

2. The terrain-sensing optical proximity effect correction method as described in claim 1, characterized in that, The step of using the current TOPC model to perform terrain-aware optical proximity effect correction on the original design map to generate an initial corrected map includes: The original design layout was simulated using the CMP module to obtain the predicted morphology data of the wafer surface. Based on the topography prediction data, the original design layout is divided into multiple out-of-focus zones; For each defocused zone, the corresponding OPC module is invoked to perform optical proximity effect correction, generating an initial corrected layout.

3. The terrain-sensing optical proximity effect correction method as described in claim 2, characterized in that, Based on the topography prediction data, the original design layout is divided into multiple out-of-focus zones, including: Based on multiple preset discrete defocus threshold values, the defocus amount predicted for each local location in the shape prediction data is classified into the corresponding defocus interval. All graphic units within the same out-of-focus area are assigned to the same out-of-focus partition, thereby dividing the original design layout into multiple out-of-focus partitions.

4. The terrain-aware optical proximity effect correction method as described in claim 2, characterized in that, Based on the topography prediction data, the original design layout is divided into multiple out-of-focus zones, including: Acquire multiple discrete defocus states defined for the target process window, each of which is associated with a specific OPC module; The defocus amount predicted for each local location in the topography prediction data is mapped to the corresponding discrete defocus state. Based on the mapping results, the original design layout is divided into defocus partitions corresponding to each of the discrete defocus states.

5. The terrain-aware optical proximity effect correction method as described in claim 2, characterized in that, Based on the topography prediction data, the original design layout is divided into multiple out-of-focus zones, including: Based on the morphology prediction data, surface depression areas and surface protrusion areas caused by the chemical mechanical polishing process are identified. The surface depression region is divided into a first defocus zone associated with a first defocus amount; The surface protrusion area is divided into a second defocus zone associated with the second defocus amount; This divides the original design layout into multiple out-of-focus zones.

6. The terrain-aware optical proximity effect correction method as described in claim 1, characterized in that, Before obtaining the original design layout, the following is also included: Construct a parameter correction model.

7. The terrain-aware optical proximity effect correction method as described in claim 6, characterized in that, The construction of the parameter correction model includes: Acquire historical production data, which includes historical design maps, corresponding historical topographic prediction data, historical measurement data, and historical model deviation data calculated based on the historical measurement data. Training features are extracted from the historical production data to form a training sample set; Based on the training sample set, a preset machine learning algorithm is trained to establish a mapping relationship from input features to model bias, thereby obtaining a parameter correction model.

8. A terrain-sensing optical proximity effect correction device, used to perform the terrain-sensing optical proximity effect correction method as described in claim 1, characterized in that, The terrain-aware optical proximity effect correction device includes: The layout acquisition unit is used to acquire the original design layout. The map correction unit is used to perform terrain-aware optical proximity effect correction on the original design map using the current TOPC model, and generate an initial corrected map; wherein, the current TOPC model includes a CMP module and an OPC module; The in-situ measurement unit is used to perform in-situ measurements on the wafer manufactured based on the initial modified layout to obtain actual pattern data; The data comparison unit is used to compare the actual graphic data with the simulation graphic data generated based on the initial corrected layout to generate model deviation data. The data analysis unit is used to analyze the model deviation data using a parameter correction model and generate parameter correction values. The analysis includes: associating the model deviation data with layout features and topographic features to infer the root cause of the deviation, and generating parameter correction values ​​for the CMP module and / or for the OPC module based on the inference results. The model update unit is used to update the current TOPC model online based on the parameter correction amount, including: performing online incremental updates of the CMP module and / or OPC module in the current TOPC model, and using the updated TOPC model as the new current TOPC model; repeating the above steps S101-S106, so that the current TOPC model is continuously iteratively updated based on the in-situ measurement data of subsequent batches during the production process, which is used to correct the terrain-aware optical proximity effect of the original design map of subsequent batches, thereby forming a closed-loop adaptive correction process from correction, manufacturing, measurement, learning to re-correction.

9. A storage medium, characterized in that, The storage medium stores multiple instructions adapted for loading by a processor to execute the terrain-aware optical proximity effect correction method according to any one of claims 1-7.

10. An electronic device, characterized in that, The system includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the terrain-aware optical proximity effect correction method as described in any one of claims 1-7.

Images