A method, device and electronic device for establishing a near-field calculation model of extreme ultraviolet light mask diffraction

Abstract

The present invention discloses a method, device and electronic device for establishing an extreme ultraviolet mask diffraction near-field calculation model, and relates to the field of microelectronic technology. The method for establishing the extreme ultraviolet mask diffraction near-field calculation model includes: obtaining a diffraction matrix database corresponding to a preset number of mask layouts; cutting and classifying each mask layout based on the diffraction matrix database corresponding to the mask layout to obtain multiple types of feature graphics; based on each type of feature graphics, determining a convolution kernel with mask diffraction near-field characteristics; determining the near-field data corresponding to the test layout according to the convolution kernel, and completing the establishment of the extreme ultraviolet mask diffraction near-field calculation model. This model trains a set of effective convolution kernels for the extreme ultraviolet lithography mask diffraction near-field, and can be applied to any incident light angle, while maintaining the same order of magnitude of calculation accuracy as the machine learning method, but with a greatly reduced time cost.

Description

Technical Field

[0001] The present invention relates to the field of microelectronics technology, and in particular to a method, a device and an electronic device for establishing an extreme ultraviolet light mask diffraction near-field calculation model. Background Art

[0002] Extreme ultraviolet lithography (EUVL) is a microelectronic lithography technology that uses EUV rays with a wavelength of 11 to 14 nm as the exposure light source. It is suitable for mass production of integrated circuits with a feature size of 32 nm and finer line width. EUV lithography principle: Plasma is generated by laser excitation or discharge. The plasma emits EUV radiation. The EUV radiation is incident on the reflective mask through a focusing system composed of periodic multilayer thin film reflectors. The reflected EUV light waves are then projected through a projection system composed of reflectors to image the geometric pattern of the integrated circuit on the reflective mask into the photoresist on the silicon wafer, thereby forming the lithographic pattern required for the integrated circuit. When the mask reflects the incident light, a diffraction electric field is generated near its surface, which is called the mask near field. The diffraction near field data consists of XX, XY, YX, and YY diffraction data.

[0003] Computational lithography techniques, such as optical proximity correction (OPC) and source mask optimization (SMO), are important tools for improving lithography imaging performance. During the OPC and SMO iterations, EUV imaging models need to be called repeatedly, and the speed and accuracy of the imaging model have a great impact on the speed and performance of computational lithography. At the same time, researchers of new EUV machines and lithography processes also need high-precision imaging models as auxiliary analysis tools. Quickly and accurately calculating the EUV mask diffraction spectrum is the key to establishing the EUV imaging model. Unlike DUV masks, EUV masks are designed with reflections and can be hundreds of nanometers thick, much larger than the wavelength of the incident light. Due to the oblique incidence morphology and thick mask effects, the traditional Kirchhoff thin mask model is no longer accurate. The most accurate way to solve this problem is to use rigorous electromagnetic field (EMF) simulation methods such as the finite-difference time-domain (FDTD) method and the rigorous coupled wave analysis (RCWA) method to model mask diffraction at multiple incident angles. Although the EMF method is accurate, it takes too long and has too high a computational cost, so it is not suitable for large-scale calculations.

[0004] In order to improve the speed of simulation, a series of fast mask models have been proposed, mainly including models based on regional decomposition method, models based on structural decomposition method and models based on machine learning. The regional decomposition method (DDM) is a typical fast model based on layout decomposition. The DDM method can quickly calculate the diffraction spectrum and near field of the EUV mask. This method converts the complex two-dimensional (2D) pattern into a superposition of a series of one-dimensional (1D) patterns. Liu.P et al. proposed a fast mask diffraction near-field calculation model suitable for 193 immersion lithography. The model decomposes the mask pattern into regions, edges and corners. The corresponding filter is obtained through rigorous simulation, which can be used to calculate the spatial image of any mask pattern. The structural decomposition method (SDM) decomposes the mask structure into absorption layers and multilayer structures, and models them separately. The diffraction spectrum of the mask is calculated by simulating the propagation of the incident light. SDM has a clear physical meaning and has been successfully applied to the theoretical analysis of the three-dimensional effect of the mask. The machine learning method decomposes the mask layout into graphics with concave corners, convex corners and edge features. By establishing a layout database with these three types of features, the decomposed layout graphics are matched with the graphics in the database, and then the corresponding near-field data are extracted for splicing. This model has made progress in the accuracy of mask diffraction near-field calculations.

[0005] Unlike 193nm immersion lithography, EUV lithography uses a reflective mask. The characteristics of the diffraction near field of the reflective mask and the projection mask are completely different. Therefore, the filter used in the model based on the DDM method proposed by Liu.P et al. is no longer suitable for EUV lithography. SDM uses a point pulse to describe the boundary diffraction at the edge of the absorption stripe. The point pulse is a constant and is only related to the absorption material and thickness. It cannot describe the effect of the incident angle on the diffraction spectrum of the absorber. At the same time, in the model, the diffraction area under oblique incidence is simply regarded as the Hopkins shift of the diffraction under normal incidence. Therefore, with the increase of the incident angle and the decrease of the mask half pitch, the accuracy of SDM gradually decreases. Although the calculation accuracy of the model based on the machine learning method has been improved, the time cost is also increasing. Summary of the invention

[0006] The purpose of the present invention is to provide a method, device and electronic equipment for establishing an extreme ultraviolet mask diffraction near-field calculation model, so as to solve the problems of low accuracy and time-consuming of existing extreme ultraviolet lithography calculation models.

[0007] In a first aspect, the present invention provides a method for establishing an extreme ultraviolet mask diffraction near-field calculation model, comprising:

[0008] Acquire a preset number of diffraction matrix databases corresponding to mask patterns, wherein each of the mask patterns corresponds to at least one diffraction matrix database, wherein each of the mask patterns corresponds to at least one diffraction matrix;

[0009] Cutting and classifying each of the mask patterns based on a diffraction matrix database corresponding to the mask patterns to obtain multiple types of feature patterns;

[0010] Based on the characteristic pattern of each type, determining a convolution kernel having a mask diffraction near-field characteristic;

[0011] The near-field data corresponding to the test layout is determined according to the convolution kernel, and the establishment of the extreme ultraviolet light mask diffraction near-field calculation model is completed.

[0012] Compared with the prior art, the method for establishing the extreme ultraviolet light mask diffraction near-field calculation model provided in the embodiment of the present application can obtain a diffraction matrix database corresponding to a preset number of mask patterns, wherein each of the mask patterns corresponds to at least one diffraction matrix database; based on the diffraction matrix database corresponding to the mask pattern, each of the mask patterns is cut and classified to obtain multiple types of feature patterns; based on the feature patterns of each type, a convolution kernel having the corresponding mask diffraction near-field characteristics is determined; according to the convolution kernel, the near-field data corresponding to the test pattern is determined to complete the establishment of the extreme ultraviolet light mask diffraction near-field calculation model. This model trains a set of effective convolution kernels for the extreme ultraviolet lithography mask diffraction near-field and can be applied to any incident light angle. At the same time, it can maintain the same order of magnitude of calculation accuracy as the machine learning method, but the time cost is greatly reduced.

[0013] In a possible implementation, each mask pattern is cut and classified based on the diffraction matrix database corresponding to the mask pattern to obtain multiple types of feature patterns, including:

[0014] Based on the diffraction matrix corresponding to each of the mask plates, the layout matrix of the mask plate is extended outward by a first preset number of pixels to obtain a corresponding enlarged layout matrix;

[0015] Based on the diffraction matrix corresponding to each of the mask patterns, the layout matrix of the mask pattern is reduced inward by a second preset number of pixels to obtain a corresponding reduced layout matrix;

[0016] Determine a layout edge pattern matrix based on the enlarged layout matrix and the reduced layout matrix;

[0017] The layout edge pattern matrix is ​​segmented according to a plurality of outer edge features to obtain the plurality of types of feature patterns.

[0018] In a possible implementation, segmenting the layout edge pattern matrix according to a plurality of outer edge features to obtain the plurality of types of feature patterns includes:

[0019] Segmenting the layout edge pattern matrix according to a plurality of outer edge features to obtain a plurality of feature two-dimensional matrices having corresponding edge features;

[0020] The plurality of feature two-dimensional matrices are secondary classified according to different feature points of the plurality of outer edge features to obtain the plurality of types of feature graphics.

[0021] In a possible implementation, the diffraction matrix includes an XX diffraction matrix, a YY diffraction matrix, an XY diffraction matrix, and a YX diffraction matrix; the convolution kernel includes a diffraction near-field real convolution kernel and a diffraction near-field imaginary convolution kernel; and the determining of the convolution kernel having the mask diffraction near-field characteristic based on the characteristic pattern of each type includes:

[0022] Based on the characteristic pattern of each category, by extracting pixel points in the characteristic pattern of each category, determining a pixel point coordinate set of the characteristic pattern of each category;

[0023] Determine, based on the pixel point coordinate set, the pixel point coordinates of the characteristic pattern of each type corresponding to the near field data in the XX diffraction matrix, the YY diffraction matrix, the XY diffraction matrix and the YX diffraction matrix;

[0024] Based on the diffraction characteristics of the XX diffraction matrix, the diffraction characteristics of the YY diffraction matrix and the diffraction characteristics of the intermediate reduced layout, the near-field data of the XX diffraction matrix and the near-field data of the YY diffraction matrix are fitted to determine a first numerical number of the first convolution kernels and a second numerical number of the second convolution kernels having the outer edge feature;

[0025] Based on the diffraction characteristics of the XY diffraction matrix, the diffraction characteristics of the YX diffraction matrix and the diffraction characteristics of the intermediate reduced layout, the near-field data of the XY diffraction matrix and the near-field data of the YX diffraction matrix are fitted to determine a third numerical value of the third convolution kernels and a fourth numerical value of the fourth convolution kernels having the outer edge feature;

[0026] A diffraction near-field real convolution kernel and a diffraction near-field imaginary convolution kernel are determined based on a first numerical value of the first convolution kernel, a second numerical value of the second convolution kernel, a third numerical value of the third convolution kernel, and a fourth numerical value of the fourth convolution kernel.

[0027] In a possible implementation, the convolution kernel includes a diffraction near-field real convolution kernel and a diffraction near-field imaginary convolution kernel; determining the near-field data corresponding to the test layout according to the convolution kernel to complete the establishment of the extreme ultraviolet mask diffraction near-field calculation model includes:

[0028] Obtaining a test diffraction matrix corresponding to the test pattern;

[0029] Cutting and classifying the test layout based on the test diffraction matrix corresponding to the test layout to obtain various types of characteristic test patterns;

[0030] Determine the diffraction near-field real part data corresponding to each type of characteristic test pattern based on the diffraction near-field real part convolution kernel;

[0031] Determine the diffraction near-field imaginary part data corresponding to each type of characteristic test pattern based on the diffraction near-field imaginary part convolution kernel;

[0032] The near-field data is determined based on the diffraction near-field real data and the diffraction near-field imaginary data, and the establishment of the extreme ultraviolet light mask diffraction near-field calculation model is completed.

[0033] In a possible implementation, the step of obtaining a preset number of diffraction matrix databases corresponding to mask layouts, wherein each mask layout corresponds to at least one diffraction matrix, includes:

[0034] A diffraction matrix database corresponding to a preset number of mask patterns is determined based on a time-domain finite difference electromagnetic calculation method, wherein each mask pattern corresponds to at least one diffraction matrix.

[0035] In a second aspect, the present invention further provides a device for establishing an extreme ultraviolet mask diffraction near-field calculation model, the device comprising:

[0036] An acquisition module, used for acquiring a preset number of diffraction matrix databases corresponding to mask patterns, wherein each of the mask patterns corresponds to at least one diffraction matrix database, wherein each of the mask patterns corresponds to at least one diffraction matrix;

[0037] A cutting module, used for cutting and classifying each of the mask patterns based on a diffraction matrix database corresponding to the mask patterns to obtain multiple types of feature patterns;

[0038] A determination module, used for determining a convolution kernel having a mask diffraction near-field characteristic based on the characteristic pattern of each type;

[0039] The model building module is used to determine the near-field data corresponding to the test layout according to the convolution kernel, and complete the establishment of the extreme ultraviolet light mask diffraction near-field calculation model.

[0040] In a possible implementation, the cutting module includes:

[0041] An outward extension submodule, used for extending the layout matrix of the mask layout outward by a first preset number of pixels based on the diffraction matrix corresponding to each of the mask templates, to obtain a corresponding enlarged layout matrix;

[0042] An inward reduction submodule, used to reduce the layout matrix of the mask pattern inwardly by a second preset number of pixels based on the diffraction matrix corresponding to each mask pattern, to obtain a corresponding reduced layout matrix;

[0043] A first determination submodule is used to determine a layout edge pattern matrix based on the enlarged layout matrix and the reduced layout matrix;

[0044] The segmentation submodule is used to segment the layout edge pattern matrix according to multiple outer edge features to obtain the multiple types of feature patterns.

[0045] In a possible implementation, the segmentation submodule includes:

[0046] A segmentation unit, used for segmenting the layout edge pattern matrix according to a plurality of outer edge features to obtain a plurality of feature two-dimensional matrices with corresponding edge features;

[0047] The classification unit is used to perform secondary classification on the plurality of feature two-dimensional matrices according to different feature points of the plurality of outer edge features to obtain the plurality of types of feature graphics.

[0048] In a possible implementation, the diffraction matrix includes an XX diffraction matrix, a YY diffraction matrix, an XY diffraction matrix, and a YX diffraction matrix; the convolution kernel includes a diffraction near-field real part convolution kernel and a diffraction near-field imaginary part convolution kernel; and the determination module includes:

[0049] A second determination submodule is used to determine a pixel point coordinate set of each type of the characteristic pattern by extracting pixel points in the characteristic pattern of each type based on the characteristic pattern of each type;

[0050] A third determination submodule, for determining, based on the pixel point coordinate set, the near-field data corresponding to the pixel point coordinates of the characteristic pattern of each type in the XX diffraction matrix, the YY diffraction matrix, the XY diffraction matrix and the YX diffraction matrix;

[0051] a fourth determination submodule, configured to fit the near-field data of the XX diffraction matrix and the near-field data of the YY diffraction matrix based on the diffraction characteristics of the XX diffraction matrix, the diffraction characteristics of the YY diffraction matrix, and the diffraction characteristics of the intermediate reduced layout, and determine a first numerical number of the first convolution kernels and a second numerical number of the second convolution kernels having the outer edge feature;

[0052] a fifth determination submodule, configured to fit the near-field data of the XY diffraction matrix and the near-field data of the YX diffraction matrix based on the diffraction characteristics of the XY diffraction matrix, the diffraction characteristics of the YX diffraction matrix, and the diffraction characteristics of the intermediate reduced layout, and determine a third numerical value of the third convolution kernels and a fourth numerical value of the fourth convolution kernels having the outer edge feature;

[0053] The sixth determination submodule is used to determine the diffraction near-field real convolution kernel and the diffraction near-field imaginary convolution kernel based on the first numerical value of the first convolution kernel, the second numerical value of the second convolution kernel, the third numerical value of the third convolution kernel and the fourth numerical value of the fourth convolution kernel.

[0054] In a possible implementation, the convolution kernel includes a diffraction near-field real part convolution kernel and a diffraction near-field imaginary part convolution kernel; and the model building module includes:

[0055] An acquisition submodule, used for acquiring a test diffraction matrix corresponding to the test layout;

[0056] A cutting and classification submodule, used for performing cutting and classification processing on the test layout based on a test diffraction matrix corresponding to the test layout to obtain various types of characteristic test patterns;

[0057] A seventh determination submodule, configured to determine the diffraction near-field real part data corresponding to each type of characteristic test pattern based on the diffraction near-field real part convolution kernel;

[0058] An eighth determination submodule, configured to determine the diffraction near-field imaginary part data corresponding to each type of characteristic test pattern based on the diffraction near-field imaginary part convolution kernel;

[0059] The ninth determination submodule is used to determine the near-field data based on the diffraction near-field real data and the diffraction near-field imaginary data, and complete the establishment of the extreme ultraviolet light mask diffraction near-field calculation model.

[0060] In a possible implementation, the acquisition module includes:

[0061] The tenth determination submodule is used to determine the diffraction matrix database corresponding to a preset number of the mask patterns based on a time-domain finite difference electromagnetic calculation method, wherein each of the mask patterns corresponds to at least one diffraction matrix.

[0062] In a third aspect, the present invention further provides an electronic device comprising: one or more processors; and one or more machine-readable media having instructions stored thereon, which, when executed by the one or more processors, enables the method for establishing the extreme ultraviolet light mask diffraction near-field computational model described in any one of the first aspects to be executed.

[0063] Compared with the prior art, the beneficial effects of the device for establishing the extreme ultraviolet light mask diffraction near-field calculation model and the electronic device provided by the present invention are the same as the beneficial effects of the method for establishing the extreme ultraviolet light mask diffraction near-field calculation model described in the above technical scheme, and will not be repeated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0065] Figure 1 A schematic flow chart of a method for establishing an extreme ultraviolet mask diffraction near-field calculation model provided in an embodiment of the present application is shown;

[0066] Figure 2 A schematic flow chart of another method for establishing an extreme ultraviolet mask diffraction near-field calculation model provided in an embodiment of the present application is shown;

[0067] Figure 3 A schematic diagram of a test mask provided in an embodiment of the present application is shown;

[0068] Figure 4 A schematic diagram showing four diffraction matrices obtained by calculating a test mask 1 provided in an embodiment of the present application is shown;

[0069] Figure 5 A schematic diagram showing four diffraction matrices obtained by calculating a test mask 2 provided in an embodiment of the present application is shown;

[0070] Figure 6 A schematic diagram showing four diffraction matrices obtained by another calculation test mask 1 provided in an embodiment of the present application is shown;

[0071] Figure 7 A schematic diagram showing four diffraction matrices obtained by calculating another test mask 2 provided in an embodiment of the present application is shown;

[0072] Figure 8 A schematic diagram showing four diffraction matrices obtained by calculating a test mask 1 provided in an embodiment of the present application is shown;

[0073] Fig. 9 A schematic diagram showing four diffraction matrices obtained by calculating a test mask 2 provided in an embodiment of the present application is shown;

[0074] Fig.10 A schematic diagram showing four diffraction matrices obtained by calculating a test mask 1 provided in an embodiment of the present application is shown;

[0075] Fig.11A schematic diagram showing four diffraction matrices obtained by calculating a test mask 2 provided in an embodiment of the present application is shown;

[0076] Fig.12 A schematic diagram showing four diffraction matrices obtained by calculating a test mask 1 provided in an embodiment of the present application is shown;

[0077] Fig.13 A schematic diagram showing four diffraction matrices obtained by calculating a test mask 2 provided in an embodiment of the present application is shown;

[0078] Fig.14 A structural block diagram of a device for establishing an extreme ultraviolet mask diffraction near-field calculation model provided by an embodiment of the present invention is shown;

[0079] Fig.15 A schematic diagram of the hardware structure of a technology level determination device in an embodiment of the present invention;

[0080] Fig.16 Schematic diagram of the structure of the chip in the embodiment of the present invention.

[0081] Reference numerals:

[0082] 40-electronic device; 401-processor; 401-1-first processor; 401-2-second processor; 402-communication interface; 403-communication line; 404-memory; 50-chip; 405-bus system. DETAILED DESCRIPTION

[0083] In order to clearly describe the technical solutions of the embodiments of the present invention, in the embodiments of the present invention, words such as "first" and "second" are used to distinguish the same items or similar items with basically the same functions and effects. For example, the first threshold and the second threshold are only used to distinguish different thresholds, and their order is not limited. Those skilled in the art can understand that words such as "first" and "second" do not limit the quantity and execution order, and words such as "first" and "second" do not necessarily limit them to be different.

[0084] It should be noted that, in the present invention, words such as "exemplary" or "for example" are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary" or "for example" in the present invention should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary" or "for example" is intended to present related concepts in a specific way.

[0085] In the present invention, "at least one" means one or more, and "plurality" means two or more. "And / or" describes the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone, where A and B can be singular or plural. The character " / " generally indicates that the associated objects before and after are in an "or" relationship. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, at least one of a, b or c can mean: a, b, c, the combination of a and b, the combination of a and c, the combination of b and c, or the combination of a, b and c, where a, b, c can be single or multiple.

[0086] Figure 1 A schematic diagram of a process for establishing a near-field calculation model of extreme ultraviolet mask diffraction provided in an embodiment of the present application is shown, such as Figure 1 As shown, the method for establishing the extreme ultraviolet mask diffraction near-field calculation model includes:

[0087] Step 101: Obtain a diffraction matrix database corresponding to a preset number of mask patterns.

[0088] Wherein, each of the mask patterns corresponds to at least one diffraction matrix.

[0089] In the present application, a diffraction matrix database corresponding to a preset number of mask patterns can be determined based on a time-domain finite difference electromagnetic calculation method, wherein each of the mask patterns corresponds to at least one diffraction matrix.

[0090] Among them, the embodiment of the present application does not limit the number of mask patterns, which can be 30, 35, or 40, and can be specifically limited according to the actual application scenario.

[0091] Step 102: Cut and classify each mask pattern based on the diffraction matrix database corresponding to the mask pattern to obtain multiple types of feature patterns.

[0092] In the present application, based on the diffraction matrix corresponding to each of the mask plates, the layout matrix of the mask plate can be extended outward by a first preset number of pixels to obtain a corresponding enlarged layout matrix; based on the diffraction matrix corresponding to each of the mask plates, the layout matrix of the mask plate can be reduced inward by a second preset number of pixels to obtain a corresponding reduced layout matrix; based on the enlarged layout matrix and the reduced layout matrix, a layout edge graphic matrix is ​​determined; and the layout edge graphic matrix is ​​segmented according to multiple outer edge features to obtain the multiple types of feature graphics.

[0093] Step 103: Based on the characteristic pattern of each type, determine a corresponding convolution kernel having mask diffraction near-field characteristics.

[0094] In the present application, the diffraction matrix includes an XX diffraction matrix, a YY diffraction matrix, an XY diffraction matrix and a YX diffraction matrix; the convolution kernel includes a diffraction near-field real part convolution kernel and a diffraction near-field imaginary part convolution kernel; the specific implementation process of step 103 may include:

[0095] Based on the characteristic pattern of each category, by extracting pixel points in the characteristic pattern of each category, determining a pixel point coordinate set of the characteristic pattern of each category;

[0096] Determine, based on the pixel point coordinate set, the pixel point coordinates of the characteristic pattern of each type corresponding to the near field data in the XX diffraction matrix, the YY diffraction matrix, the XY diffraction matrix and the YX diffraction matrix;

[0097] Based on the diffraction characteristics of the XX diffraction matrix, the diffraction characteristics of the YY diffraction matrix and the diffraction characteristics of the intermediate reduced layout, the near-field data of the XX diffraction matrix and the near-field data of the YY diffraction matrix are fitted to determine a first numerical number of the first convolution kernels and a second numerical number of the second convolution kernels having the outer edge feature;

[0098] Based on the diffraction characteristics of the XY diffraction matrix, the diffraction characteristics of the YX diffraction matrix and the diffraction characteristics of the intermediate reduced layout, the near-field data of the XY diffraction matrix and the near-field data of the YX diffraction matrix are fitted to determine a third numerical value of the third convolution kernels and a fourth numerical value of the fourth convolution kernels having the outer edge feature;

[0099] A diffraction near-field real convolution kernel and a diffraction near-field imaginary convolution kernel are determined based on a first numerical value of the first convolution kernel, a second numerical value of the second convolution kernel, a third numerical value of the third convolution kernel, and a fourth numerical value of the fourth convolution kernel.

[0100] Step 104: Determine the near-field data corresponding to the test layout according to the convolution kernel, and complete the establishment of the extreme ultraviolet light mask diffraction near-field calculation model.

[0101] In the present application, the convolution kernel includes a diffraction near-field real convolution kernel and a diffraction near-field imaginary convolution kernel; specifically, a test diffraction matrix corresponding to the test layout can be obtained; based on the test diffraction matrix corresponding to the test layout, the test layout is cut and classified to obtain various types of characteristic test patterns; based on the diffraction near-field real convolution kernel, the diffraction near-field real data corresponding to the various types of characteristic test patterns are determined; based on the diffraction near-field imaginary convolution kernel, the diffraction near-field imaginary data corresponding to the various types of characteristic test patterns are determined; based on the diffraction near-field real data and the diffraction near-field imaginary data, the near-field data is determined to complete the establishment of the diffraction near-field calculation model of the extreme ultraviolet light mask.

[0102] In summary, the method for establishing the near-field calculation model of the extreme ultraviolet mask diffraction provided in the embodiment of the present application can obtain a preset number of diffraction matrix databases corresponding to mask layouts, wherein each of the mask layouts corresponds to at least one diffraction matrix database; based on the diffraction matrix database corresponding to the mask layout, each of the mask layouts is cut and classified to obtain multiple types of feature patterns;

[0103] Based on the characteristic patterns of each type, a convolution kernel with corresponding mask diffraction near-field characteristics is determined; the near-field data corresponding to the test layout is determined according to the convolution kernel, and the establishment of the extreme ultraviolet mask diffraction near-field calculation model is completed. This model trains a set of effective convolution kernels for the extreme ultraviolet lithography mask diffraction near-field and can be applied to any incident light angle. At the same time, it can maintain the same order of magnitude of calculation accuracy as the machine learning method, but the time cost is greatly reduced.

[0104] Figure 2 A schematic flow chart of another method for establishing an extreme ultraviolet mask diffraction near-field calculation model provided in an embodiment of the present application is shown, such as Figure 2 As shown, the method for establishing the extreme ultraviolet light mask diffraction near-field calculation model includes:

[0105] Step 201: Obtain a diffraction matrix database corresponding to a preset number of mask patterns.

[0106] In the present application, a diffraction matrix database corresponding to a preset number of mask patterns is determined based on a time-domain finite difference electromagnetic calculation method, wherein each of the mask patterns corresponds to at least one diffraction matrix.

[0107] Among them, the embodiment of the present application does not limit the number of mask patterns, which can be 30, 35, or 40, and can be specifically limited according to the actual application scenario.

[0108] For example, taking the number of mask patterns as 30, a database with 30 mask patterns and their corresponding XX, XY, YX, and YY diffraction matrices is established. The specific process may include: the mask pattern is represented by a two-dimensional matrix M with pixels of 400*400, each element of the matrix represents a pixel of the mask pattern, and the value of the element is 0 or 1. The element 0 represents that the pixel is a background area, and the element 1 represents that the pixel is a graphic area.

[0109] The XX, XY, YX, and YY diffraction matrices corresponding to each layout are calculated using a rigorous electromagnetic calculation method based on FDTD. The simulation environment is set as follows: point light source illumination, with point light source coordinates of x=0.3 and y=0.3. The polarization is TE polarization. The angle of incidence is 6°. The material of the mask absorption layer is tantalum boron nitride (TaBN) with a thickness of 60nm. The mask reflection layer consists of 40 layers of Mo / Si double-layer film, with a single Mo layer thickness of 1nm and a single Si layer thickness of 0.725nm. Multilayer defects and aberrations are not considered in this study. The spacing and size of the mask pattern are both wafer sizes.

[0110] The calculated XX, XY, YX, and YY diffraction matrices are represented by two-dimensional matrices Representation, n represents the number of the layout, the value is 1 to 30.

[0111] Step 202: Based on the diffraction matrix corresponding to each of the mask plates, the layout matrix of the mask plate is extended outward by a first preset number of pixels to obtain a corresponding enlarged layout matrix.

[0112] In the present application, the number range of the first preset number of pixel points can be greater than or equal to 10 and less than or equal to 20. The embodiment of the present application does not make any specific limitation on this number and can be adjusted according to the actual application scenario.

[0113] For example, taking a single layout as an example, the following operations are performed on all layouts. When the first preset number of pixels is 15, the edge of the layout matrix can be extended outward by 15 pixels to obtain an enlarged layout matrix M B .

[0114] Step 203: based on the diffraction matrix corresponding to each of the mask images, the layout matrix of the mask image is reduced inward by a second preset number of pixels to obtain a corresponding reduced layout matrix.

[0115] In the present application, the number range of the second preset number of pixel points can be greater than or equal to 10 and less than or equal to 20. The embodiment of the present application does not make any specific limitation on this number and can be adjusted according to the actual application scenario.

[0116] For example, taking a single layout as an example, the following operations are performed on all layouts. When the second preset number of pixels is 10, the edge of the layout matrix can be reduced by 10 pixels inward to obtain a reduced layout matrix M c

[0117] Step 204: Determine a layout edge pattern matrix based on the enlarged layout matrix and the reduced layout matrix.

[0118] In this application, M B -M c Then we get the layout edge pattern matrix M e .

[0119] Step 205: segmenting the layout edge pattern matrix according to a plurality of outer edge features to obtain the plurality of types of feature patterns.

[0120] In the present application, the layout edge graphic matrix can be segmented according to multiple outer edge features to obtain multiple feature two-dimensional matrices with corresponding edge features; the multiple feature two-dimensional matrices can be secondary classified according to different feature points of the multiple outer edge features to obtain the multiple types of feature graphics.

[0121] The outer edge features may include concave corners, convex corners and edges. Different feature points may be different orientations of the outer edge features.

[0122] Specifically, the edge graph can be segmented according to the features of the outer edge line, based on the features of concave angle, convex angle, and edge. The graph with concave angle features is segmented using the two-dimensional matrix M concave Characterize, the graphics with convex corner features are represented by a two-dimensional matrix M convex Characterize, the graph with edge features is represented by a two-bit matrix M edge Representation.

[0123] Furthermore, the obtained figures can be classified into two categories according to the different directions of concave corners, convex corners and edges. The figures with concave corners facing the upper left are represented by the two-dimensional matrix M concave1 Characterize the figure with a concave corner facing downward to the left using the two-dimensional matrix M concave2 Characterize the graph with a concave corner facing up to the right using the two-dimensional matrix M concave3 Characterize the graph with a concave corner facing downward to the right using the two-dimensional matrix M concave4 The figure with the upper left convex corner is represented by the two-dimensional matrix M convex1 Characterize, the figure with the lower left convex corner is represented by the two-dimensional matrix M convex2 Characterize the graph with the upper right convex corner with the two-dimensional matrix M convex3 Characterize the figure with the lower right convex corner with the two-dimensional matrix M convex4A graph with right-facing edges is represented by a two-dimensional matrix M edge1 Representation, a graph with left-facing edges is represented by a two-dimensional matrix M edge2 Representation, a graph with upward-facing edges is represented by a two-dimensional matrix M edge3 Representation, a graph with downward facing edges is represented by a two-dimensional matrix M edge4 After segmenting all the layouts, a collection of 13 types of graphics can be obtained (including four orientations of concave corner feature graphics, four orientations of convex corner feature graphics, four orientations of edge feature graphics, and reduced layouts).

[0124] Step 206: Based on the characteristic pattern of each type, determine a corresponding convolution kernel having mask diffraction near-field characteristics.

[0125] Wherein, the diffraction matrix includes an XX diffraction matrix, a YY diffraction matrix, an XY diffraction matrix and a YX diffraction matrix; the convolution kernel includes a diffraction near-field real part convolution kernel and a diffraction near-field imaginary part convolution kernel;

[0126] The specific implementation process of the above step 206 may include:

[0127] Sub-step A1: Based on the characteristic pattern of each category, by extracting the pixel points in the characteristic pattern of each category, a pixel point coordinate set of the characteristic pattern of each category is determined.

[0128] In this application, the coordinates of the pixel points of various characteristic patterns can be extracted. Taking a pattern with a concave corner facing the upper left as an example, the coordinate set is represented as m is the total number of pixels of this type of feature graphics. Finally, various convex corner graphics, various concave corner graphics, various edge graphics, and reduced layout M can be obtained. c A set of pixel coordinates.

[0129] Sub-step A2: Determine, based on the pixel point coordinate set, the pixel point coordinates of the characteristic pattern of each type corresponding to the near-field data in the XX diffraction matrix, the YY diffraction matrix, the XY diffraction matrix and the YX diffraction matrix.

[0130] Sub-step A3: Based on the diffraction characteristics of the XX diffraction matrix, the diffraction characteristics of the YY diffraction matrix and the diffraction characteristics of the intermediate reduced layout, fit the near-field data of the XX diffraction matrix and the near-field data of the YY diffraction matrix to determine a first numerical number of the first convolution kernels and a second numerical number of the second convolution kernels having the outer edge characteristics.

[0131] Among them, the first value can be 48, and the second value can be 4. Specifically, for the XX and YY diffraction characteristics, 48 ​​convolution kernels with a dimension of 21*21 (24 convolution kernels with XX diffraction characteristic information, including 12 real convolution kernels and 12 imaginary convolution kernels. This is because the diffraction near-field data is in complex form, so convolution kernels are fitted for both the real and imaginary parts. 24 convolution kernels with YY diffraction characteristic information, including 12 real convolution kernels and 12 imaginary convolution kernels) and 4 convolution kernels with a dimension of 51*51 with diffraction characteristic information of the intermediate reduced pattern can be fitted by the least squares method.

[0132] Since the process of solving the convolution kernel for the XX and YY diffraction characteristics is the same, the specific steps are described using the XX diffraction characteristic:

[0133] 1) Take a pixel point with a concave corner facing the upper left as an example, allocate a square area S with the same dimension as the required convolution kernel with the pixel point coordinate as the center, and then obtain the matrix of the S area in the original map M matrix where the point is located as M S . S The matrix data is arranged in rows to form a row vector a=[k1,k2...,k n ], n = 31*31 represents the number of pixels in the S area.

[0134] 2) Obtain the value L corresponding to the pixel point coordinate in the XX diffraction matrix, where L is a complex number and extract the real part L i and the imaginary part L j , forming two column vectors b = [L i ,1],c=[L j ,1].

[0135] 3) Taking a figure with a concave corner facing the upper left as an example, the row vectors a corresponding to all the pixel points of this type of figure are arranged into a column to obtain a matrix A = [a1; a2; ...; a m ], m is the total number of pixels, and the column vectors b and c are arranged in columns to obtain B = [b1; b2; ...; b m ], C = [c1; c2; ...; c m ].

[0136] 4) Through the following formula, we can find the convolution kernel with the real part characteristics of the diffraction near field of the pattern mask XX facing the upper left and concave corner:

[0137]

[0138] XX diffraction near-field imaginary part characteristics convolution kernel

[0139]

[0140] Similarly, the convolution kernel having the diffraction near-field characteristic information of other characteristic patterns can also be obtained through the above process, which will not be elaborated in the embodiments of the present application.

[0141] Sub-step A4: Based on the diffraction characteristics of the XY diffraction matrix, the diffraction characteristics of the YX diffraction matrix and the diffraction characteristics of the intermediate reduced layout, the near-field data of the XY diffraction matrix and the near-field data of the YX diffraction matrix are fitted to determine a third numerical value of the third convolution kernels and a fourth numerical value of the fourth convolution kernels having the outer edge features.

[0142] Among them, the third value can be 32, and the fourth value can be 4. Specifically, the near-field data of the coordinates of various graphic pixel points corresponding to the XY and YX diffraction matrices can be extracted. Because the diffraction characteristics of XY and YX are different from those of XX and YY, it is necessary to set another set of convolution kernel dimensions. For the XY and YX diffraction characteristics, the least squares method is used to fit 32 dimensions of 31*31 convolution kernels with four types of concave angle features and four types of convex angle features, a total of 8 types of graphic diffraction characteristic information, 8 dimensions of 101*31 convolution kernels with two types of left and right edge features, a total of 2 types of graphic diffraction information, 8 dimensions of 31*101 convolution kernels with 2 types of upward and downward edge features, and 4 dimensions of 61*61 convolution kernels with diffraction characteristic information of the middle reduced layout. The specific steps are similar to sub-step A3, and it is only necessary to change the S area according to the different dimensions of the required convolution kernel.

[0143] Sub-step A5: Determine the diffraction near-field real convolution kernel and the diffraction near-field imaginary convolution kernel based on the first numerical value of the first convolution kernel, the second numerical value of the second convolution kernel, the third numerical value of the third convolution kernel and the fourth numerical value of the fourth convolution kernel.

[0144] Step 207: Determine the near-field data corresponding to the test layout according to the convolution kernel, and complete the establishment of the extreme ultraviolet light mask diffraction near-field calculation model.

[0145] In the present application, the convolution kernel includes a diffraction near-field real part convolution kernel and a diffraction near-field imaginary part convolution kernel. The specific implementation process of the above step 207 may include:

[0146] Sub-step B1: Obtain a test diffraction matrix corresponding to the test pattern.

[0147] Sub-step B2: cutting and classifying the test layout based on the test diffraction matrix corresponding to the test layout to obtain various types of characteristic test patterns.

[0148] Sub-step B3: determining the diffraction near-field real part data corresponding to each type of characteristic test pattern based on the diffraction near-field real part convolution kernel.

[0149] Sub-step B4: determining the diffraction near-field imaginary part data corresponding to each type of characteristic test pattern based on the diffraction near-field imaginary part convolution kernel.

[0150] Sub-step B5: Determine the near-field data based on the diffraction near-field real part data and the diffraction near-field imaginary part data, and complete the establishment of the extreme ultraviolet light mask diffraction near-field calculation model.

[0151] Calculate the XX and YY diffraction matrices of the test pattern M. The methods for calculating the XX and YY diffraction matrices are similar. The specific steps for calculating the XX diffraction matrix are as follows: 1) Perform the same cutting and classification operations on the test pattern as in step 2. The near-field calculation method for the diffraction of patterns with various concave angles, convex angles, and edge features is similar. Take the calculation of the real part of the near-field diffraction of the pattern with the concave angle feature facing the upper left as an example. 2) Perform the following operations on this type of pattern matrix Represents the convolution operation.

[0152] Through the convolution operation, the pixel values ​​of all non-zero elements in the new matrix are changed to 1, and a new matrix H is obtained. concave1 , and then perform a dot multiplication operation with the M matrix to obtain G concave1 ;

[0153] F concave1 =V concave1 .*G concave1 , M represents the test layout matrix;

[0154] G concave1 and Perform convolution operation to get V concave1 ;

[0155]

[0156] V concave1 With G concave1 Perform a dot multiplication to get F concave1 ;

[0157] F concave1 =V concave1 .*G concave1 ;

[0158] F concave1 The matrix is ​​the real part data of the XX diffraction near field of this type of graphics. Similarly, the real part data of the XX diffraction near field of other types of graphics can also be calculated through this step. Finally, the real part data matrix of the XX diffraction near field of various types of graphics is added together to obtain the real part data matrix F of the XX diffraction near field of the test pattern M.M .

[0159]

[0160] The calculation process of the imaginary part data is similar to that of the real part. It only needs to be transformed into the convolution kernel corresponding to the imaginary part. The XX diffraction near-field imaginary part data matrix R of the test pattern M can be obtained. M ;

[0161] The XX diffraction near-field data matrix T of the test layout M =V M +jR M ;

[0162] In this application, the XY and YX diffraction matrices of the test pattern M are obtained. Although the XY, YX and XX, YY diffraction near-field characteristics are different, the solution steps are similar to the above process, and only the convolution kernel needs to be replaced with the convolution kernel corresponding to the XY, YX diffraction near-field solved in step 3.

[0163] The method for establishing the near-field calculation model of the extreme ultraviolet mask diffraction in the embodiment of the present application can establish an EUV mask diffraction near-field calculation model based on linear fitting. This model trains a set of effective convolution kernels for the EUV mask diffraction near-field and can be applied to any incident light angle, while maintaining the same order of magnitude of calculation accuracy as the machine learning method, but with significantly reduced time costs.

[0164] In this application, the extreme ultraviolet mask diffraction near-field calculation model provided in this application is verified through rigorous experimental simulation. Specifically, it can be verified by setting five positions of point light source illumination conditions, namely the origin of the light source coordinate system, the point on the positive x-axis, the point on the negative x-axis, the point on the y-axis, and the point outside the axis. The time and accuracy of obtaining the mask diffraction near-field by the extreme ultraviolet mask diffraction near-field calculation model, the strict electromagnetic field method and the machine learning method are compared, thereby demonstrating the advantages of the proposed model. The error calculation of this experiment is the root mean square error between the diffraction matrix obtained by fitting the proposed model and the result of the strict electromagnetic calculation.

[0165] Among them, the waveguide method is used to calculate the near field of the mask by the strict electromagnetic calculation method. By changing the coordinates of the point light source, the calculation results of the machine learning method under the corresponding lighting conditions can be obtained. The correction function used in the near field calculation model of the extreme ultraviolet mask diffraction is pre-calculated.

[0166] The following two masks are used as test masks, and their near-field results are calculated using three different methods under different point light source illumination. The test mask size is 600 pixels * 600 pixels, the spacing of each pixel on the wafer is 1nm, and the NA value of the lithography system is 0.33. The following shows the experimental results under different point light source illumination.

[0167] Figure 3 A schematic diagram of a test mask provided in an embodiment of the present application is shown. Figure 4 A schematic diagram of four diffraction matrices obtained by calculating a test mask 1 provided in an embodiment of the present application is shown. Under the illumination condition of a point light source of x=0, y=0, the four diffraction matrices obtained by calculating the test mask 1 are obtained by using an EMF method, a method for establishing a near-field calculation model of extreme ultraviolet light mask diffraction (Proposed), a machine learning method, and a Kirchhoff model. Each diffraction matrix is ​​drawn in the form of a modulus.

[0168] Figure 5 A schematic diagram of four diffraction matrices obtained by calculating a test mask 2 provided by an embodiment of the present application is shown. Under the illumination condition of a point light source of x=0, y=0, the four diffraction matrices obtained by calculating the test mask 2 are obtained by using an EMF method, a proposed method for establishing a near-field calculation model of extreme ultraviolet light mask diffraction, a machine learning method (ML), and a Kirchhoff model (Kirchhoff). Each diffraction matrix is ​​drawn in the form of a modulus.

[0169] Table 1 shows the RMS and runtime of the test mask 1 calculated by various methods under the illumination condition of x=0, y=0 point light source:

[0170]

[0171] Table 1

[0172] Table 2: RMS and running time of test mask 2 calculated by various methods under x=0, y=0 point light source illumination conditions:

[0173]

[0174] Table 2

[0175] Figure 6 A schematic diagram of four diffraction matrices obtained by calculating another test mask 1 provided in an embodiment of the present application is shown. Under the illumination conditions of a point light source of x=0.3, y=0.3, four diffraction matrices obtained by calculating the test mask 1 are obtained by using an EMF method, a method for establishing a near-field calculation model of extreme ultraviolet light mask diffraction, a machine learning method, and a Kirchhoff model. Each diffraction matrix is ​​drawn in the form of a modulus.

[0176] Figure 7 A schematic diagram of four diffraction matrices obtained by calculating another test mask 2 provided in an embodiment of the present application is shown. Under the illumination condition of a point light source of x=0.3, y=0.3, the four diffraction matrices obtained by calculating the test mask 2 are obtained by using the EMF method, the method for establishing the near-field calculation model of the extreme ultraviolet light mask diffraction, the machine learning method, and the Kirchhoff model. Each diffraction matrix is ​​drawn in the form of a modulus.

[0177] Table 3 shows the RMS and runtime of the test mask 1 calculated by various methods under the illumination condition of x=0.3, y=0.3 point light source:

[0178]

[0179] Table 3

[0180] Table 4 shows the RMS and runtime of the test mask 2 calculated by various methods under the illumination condition of x=0.3, y=0.3 point light source:

[0181]

[0182] Table 4

[0183] Figure 8 A schematic diagram of four diffraction matrices obtained by calculating a test mask 1 provided in an embodiment of the present application is shown. Under the illumination conditions of a point light source of x=0.3, y=0, the four diffraction matrices obtained by calculating the test mask 1 are obtained by using an EMF method, a method for establishing a near-field calculation model of extreme ultraviolet light mask diffraction, a machine learning method, and a Kirchhoff model. Each diffraction matrix is ​​drawn in the form of a modulus.

[0184] Fig. 9 A schematic diagram of four diffraction matrices obtained by calculating a test mask 2 provided in an embodiment of the present application is shown. Under the illumination conditions of a point light source of x=0.3, y=0.3, the four diffraction matrices obtained by calculating the test mask 2 are obtained by using an EMF method, a method for establishing a near-field calculation model of extreme ultraviolet light mask diffraction, a machine learning method, and a Kirchhoff model. Each diffraction matrix is ​​drawn in the form of a modulus.

[0185] Table 5 shows the RMS and runtime of the test mask 1 calculated by various methods under the illumination conditions of x=0.3, y=0 point light source:

[0186]

[0187] Table 5

[0188] Table 6 shows the RMS and runtime of the test mask 2 calculated by various methods under the illumination condition of x=0.3, y=0 point light source:

[0189]

[0190] Table 6

[0191] Fig.10 A schematic diagram of four diffraction matrices obtained by calculating a test mask 1 provided in an embodiment of the present application is shown. Under the illumination conditions of a point light source of x=0, y=0.3, the four diffraction matrices obtained by calculating the test mask 1 are obtained by using an EMF method, a method for establishing a near-field calculation model of extreme ultraviolet light mask diffraction, a machine learning method, and a Kirchhoff model. Each diffraction matrix is ​​drawn in the form of a modulus.

[0192] Fig.11 A schematic diagram of four diffraction matrices obtained by calculating a test mask 2 provided in an embodiment of the present application is shown. Under the illumination conditions of a point light source of x=0, y=0.3, the four diffraction matrices obtained by calculating the test mask 2 are obtained by using an EMF method, a method for establishing a near-field calculation model of extreme ultraviolet light mask diffraction, a machine learning method, and a Kirchhoff model. Each diffraction matrix is ​​drawn in the form of a modulus.

[0193] Table 7 shows the RMS and runtime (Runtime) of the test mask 1 calculated by various methods under the conditions of x=0, y=0.3 point light source illumination:

[0194]

[0195]

[0196] Table 7

[0197] Table 8 shows the RMS and runtime (Runtime) of the test mask 2 calculated by various methods under the conditions of x=0.3, y=0 point light source illumination:

[0198]

[0199] Table 8

[0200] Fig.12 A schematic diagram of four diffraction matrices obtained by calculating a test mask 1 provided in an embodiment of the present application is shown. Under the illumination conditions of a point light source of x=-0.3, y=0, the four diffraction matrices obtained by calculating the test mask 1 are obtained by using an EMF method, a method for establishing a near-field calculation model of extreme ultraviolet light mask diffraction, a machine learning method, and a Kirchhoff model. Each diffraction matrix is ​​drawn in the form of a modulus.

[0201] Fig.13A schematic diagram of four diffraction matrices obtained by calculating the test mask 2 provided in an embodiment of the present application is shown. Under the illumination conditions of a point light source of x=-0.3, y=0, the four diffraction matrices obtained by calculating the test mask 2 are obtained by using the EMF method, the method for establishing the near-field calculation model of the extreme ultraviolet light mask diffraction, the machine learning method, and the Kirchhoff model. Each diffraction matrix is ​​drawn in the form of a modulus.

[0202] Table 9 shows the RMS and runtime (Runtime) of the test mask 1 calculated by various methods under the illumination conditions of x=-0.3, y=0 point light source:

[0203]

[0204] Table 9

[0205] Table 10 shows the RMS and runtime (Runtime) of the test mask 2 calculated by various methods under the illumination conditions of x=-0.3, y=0 point light source:

[0206]

[0207] Table 10

[0208] By comparing the time and error of calculating the diffraction matrix of test mask 1 and test mask 2 using the EMF method, the method for establishing the near-field calculation model of extreme ultraviolet mask diffraction, the method based on machine learning, and the model based on Kirchhoff approximation under illumination of point light sources at five different positions, it can be concluded that the near-field calculation model of extreme ultraviolet mask diffraction provided by the embodiment of the present invention is basically at the same order of magnitude as the machine learning method in terms of calculation error, and even the machine learning method will be inaccurate under the illumination conditions of some light sources, while the model provided by the embodiment of the present invention has been stable under all light source illumination conditions and maintained within a relatively low error range. In terms of calculation time, the model provided by the embodiment of the present invention is more than 30 times faster than the EMF method, and the calculation time is shortened by nearly 70% compared with machine learning.

[0209] In summary, the method for establishing the extreme ultraviolet light mask diffraction near-field calculation model provided in the embodiment of the present application can obtain a diffraction matrix database corresponding to a preset number of mask patterns, wherein each of the mask patterns corresponds to at least one diffraction matrix database; based on the diffraction matrix database corresponding to the mask pattern, each of the mask patterns is cut and classified to obtain multiple types of feature patterns; based on the feature patterns of each type, a convolution kernel having the corresponding mask diffraction near-field characteristics is determined; according to the convolution kernel, the near-field data corresponding to the test pattern is determined to complete the establishment of the extreme ultraviolet light mask diffraction near-field calculation model. This model trains a set of effective convolution kernels for the extreme ultraviolet lithography mask diffraction near-field and can be applied to any incident light angle. At the same time, it can maintain the same order of magnitude of calculation accuracy as the machine learning method, but the time cost is greatly reduced.

[0210] Fig.14 FIG. 1 is a block diagram showing a device for establishing a near-field calculation model of extreme ultraviolet mask diffraction provided by an embodiment of the present invention. Fig.14 As shown, the device for establishing the extreme ultraviolet mask diffraction near-field calculation model includes:

[0211] An acquisition module 301 is used to acquire a preset number of diffraction matrix databases corresponding to mask layouts, wherein each of the mask layouts corresponds to at least one diffraction matrix database, wherein each of the mask layouts corresponds to at least one diffraction matrix;

[0212] A cutting module 302 is used to perform cutting and classification processing on each of the mask patterns based on a diffraction matrix database corresponding to the mask patterns to obtain multiple types of feature patterns;

[0213] A determination module 303, configured to determine a convolution kernel having a mask diffraction near-field characteristic based on the characteristic pattern of each type;

[0214] The model building module 304 is used to determine the near-field data corresponding to the test layout according to the convolution kernel, and complete the establishment of the extreme ultraviolet light mask diffraction near-field calculation model.

[0215] In summary, the device for establishing the extreme ultraviolet light mask diffraction near-field calculation model provided in the embodiment of the present application can obtain the diffraction matrix databases corresponding to a preset number of mask patterns, wherein each of the mask patterns corresponds to at least one diffraction matrix database; based on the diffraction matrix database corresponding to the mask pattern, each of the mask patterns is cut and classified to obtain multiple types of feature patterns, and based on the feature patterns of each type, a convolution kernel having the corresponding mask diffraction near-field characteristics is determined; according to the convolution kernel, the near-field data corresponding to the test pattern is determined to complete the establishment of the extreme ultraviolet light mask diffraction near-field calculation model. This model trains a set of effective convolution kernels for the extreme ultraviolet lithography mask diffraction near-field, and can be applied to any incident light angle, while maintaining the same order of magnitude of calculation accuracy as the machine learning method, but with a greatly reduced time cost.

[0216] In a possible implementation, the cutting module includes:

[0217] An outward extension submodule, used for extending the layout matrix of the mask layout outward by a first preset number of pixels based on the diffraction matrix corresponding to each of the mask templates, to obtain a corresponding enlarged layout matrix;

[0218] An inward reduction submodule, used to reduce the layout matrix of the mask pattern inwardly by a second preset number of pixels based on the diffraction matrix corresponding to each mask pattern, to obtain a corresponding reduced layout matrix;

[0219] A first determination submodule is used to determine a layout edge pattern matrix based on the enlarged layout matrix and the reduced layout matrix;

[0220] The segmentation submodule is used to segment the layout edge pattern matrix according to multiple outer edge features to obtain the multiple types of feature patterns.

[0221] In a possible implementation, the segmentation submodule includes:

[0222] A segmentation unit, used for segmenting the layout edge pattern matrix according to a plurality of outer edge features to obtain a plurality of feature two-dimensional matrices with corresponding edge features;

[0223] The classification unit is used to perform secondary classification on the plurality of feature two-dimensional matrices according to different feature points of the plurality of outer edge features to obtain the plurality of types of feature graphics.

[0224] In a possible implementation, the diffraction matrix includes an XX diffraction matrix, a YY diffraction matrix, an XY diffraction matrix, and a YX diffraction matrix; the convolution kernel includes a diffraction near-field real part convolution kernel and a diffraction near-field imaginary part convolution kernel; and the determination module includes:

[0225] A second determination submodule is used to determine a pixel point coordinate set of each type of the characteristic pattern by extracting pixel points in the characteristic pattern of each type based on the characteristic pattern of each type;

[0226] A third determination submodule, for determining, based on the pixel point coordinate set, the near-field data corresponding to the pixel point coordinates of the characteristic pattern of each type in the XX diffraction matrix, the YY diffraction matrix, the XY diffraction matrix and the YX diffraction matrix;

[0227] a fourth determination submodule, configured to fit the near-field data of the XX diffraction matrix and the near-field data of the YY diffraction matrix based on the diffraction characteristics of the XX diffraction matrix, the diffraction characteristics of the YY diffraction matrix, and the diffraction characteristics of the intermediate reduced layout, and determine a first numerical number of the first convolution kernels and a second numerical number of the second convolution kernels having the outer edge feature;

[0228] a fifth determination submodule, configured to fit the near-field data of the XY diffraction matrix and the near-field data of the YX diffraction matrix based on the diffraction characteristics of the XY diffraction matrix, the diffraction characteristics of the YX diffraction matrix, and the diffraction characteristics of the intermediate reduced layout, and determine a third numerical value of the third convolution kernels and a fourth numerical value of the fourth convolution kernels having the outer edge feature;

[0229] The sixth determination submodule is used to determine the diffraction near-field real convolution kernel and the diffraction near-field imaginary convolution kernel based on the first numerical value of the first convolution kernel, the second numerical value of the second convolution kernel, the third numerical value of the third convolution kernel and the fourth numerical value of the fourth convolution kernel.

[0230] In a possible implementation, the convolution kernel includes a diffraction near-field real part convolution kernel and a diffraction near-field imaginary part convolution kernel; and the model building module includes:

[0231] An acquisition submodule, used for acquiring a test diffraction matrix corresponding to the test layout;

[0232] A cutting and classification submodule, used for performing cutting and classification processing on the test layout based on a test diffraction matrix corresponding to the test layout to obtain various types of characteristic test patterns;

[0233] A seventh determination submodule, configured to determine the diffraction near-field real part data corresponding to each type of characteristic test pattern based on the diffraction near-field real part convolution kernel;

[0234] An eighth determination submodule, configured to determine the diffraction near-field imaginary part data corresponding to each type of characteristic test pattern based on the diffraction near-field imaginary part convolution kernel;

[0235] The ninth determination submodule is used to determine the near-field data based on the diffraction near-field real data and the diffraction near-field imaginary data, and complete the establishment of the extreme ultraviolet light mask diffraction near-field calculation model.

[0236] In a possible implementation, the acquisition module includes:

[0237] The tenth determination submodule is used to determine the diffraction matrix database corresponding to a preset number of the mask patterns based on a time-domain finite difference electromagnetic calculation method, wherein each of the mask patterns corresponds to at least one diffraction matrix.

[0238] In summary, the device for establishing the extreme ultraviolet light mask diffraction near-field calculation model provided in the embodiment of the present application can obtain the diffraction matrix databases corresponding to a preset number of mask patterns, wherein each of the mask patterns corresponds to at least one diffraction matrix database; based on the diffraction matrix database corresponding to the mask pattern, each of the mask patterns is cut and classified to obtain multiple types of feature patterns; based on the feature patterns of each type, a convolution kernel having the corresponding mask diffraction near-field characteristics is determined; according to the convolution kernel, the near-field data corresponding to the test pattern is determined to complete the establishment of the extreme ultraviolet light mask diffraction near-field calculation model. This model trains a set of effective convolution kernels for the extreme ultraviolet lithography mask diffraction near-field and can be applied to any incident light angle. At the same time, it can maintain the same order of magnitude of calculation accuracy as the machine learning method, but the time cost is greatly reduced.

[0239] All relevant contents of each step involved in the above method embodiment can be referred to the functional description of the corresponding functional module, and will not be repeated here.

[0240] In some possible implementations, the apparatus for establishing the extreme ultraviolet light mask diffraction near-field calculation model may further include a storage module for storing program codes and data of the base station.

[0241] Among them, the processing module can be a processor or a controller, for example, a central processing unit (CPU), a general processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, transistor logic devices, hardware components or any combination thereof. It can implement or execute various exemplary logic boxes, modules and circuits described in conjunction with the disclosure of the present invention. The processor can also be a combination that implements a computing function, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, and the like. The communication module can be a transceiver, a transceiver circuit or a communication interface, and the like. The storage module can be a memory.

[0242] When the processing module is a processor, the communication module is a communication interface, and the storage module is a memory, the device for establishing the near-field calculation model of the extreme ultraviolet mask diffraction involved in the embodiment of the present invention can be Figure 4 Electronic equipment shown.

[0243] Optionally, an embodiment of the present application also provides an electronic device comprising: one or more processors; and one or more machine-readable media having instructions stored thereon, which, when executed by the one or more processors, enables the method for establishing the extreme ultraviolet light mask diffraction near-field computational model described in any one of the first aspects to be executed.

[0244] Fig.15 FIG. 1 is a schematic diagram showing the hardware structure of an electronic device provided by an embodiment of the present invention. Fig.15 As shown, the electronic device 40 includes a processor 401 and a communication interface 402 .

[0245] like Fig.15 As shown, the processor may be a general-purpose central processing unit (CPU), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits for controlling the execution of the program of the present invention. The communication interface may be one or more. The communication interface may use any transceiver-like device for communicating with other devices or communication networks.

[0246] like Fig.15As shown, the terminal device may further include a communication line 403. The communication line may include a path to transmit information between the components.

[0247] Optional, such as Fig.15 As shown, the terminal device may further include a memory 404. The memory is used to store computer-executable instructions for executing the solution of the present invention, and the execution is controlled by the processor. The processor is used to execute the computer-executable instructions stored in the memory, thereby implementing the method provided by the embodiment of the present invention.

[0248] like Fig.15 As shown, the above-mentioned memory may be a read-only memory (ROM) or other types of static storage devices that can store static information and instructions, a random access memory (RAM) or other types of dynamic storage devices that can store information and instructions, or an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical disc, laser disc, optical disc, digital versatile disc, Blu-ray disc, etc.), a magnetic disk storage medium or other magnetic storage device, or any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto. The memory may exist independently and be connected to the processor through a communication line. The memory may also be integrated with the processor.

[0249] Optionally, the computer-executable instructions in the embodiment of the present invention may also be referred to as application program codes, which is not specifically limited in the embodiment of the present invention.

[0250] In a specific implementation, as an example, Fig.15 As shown, the processor 401 may include one or more CPUs, such as Fig.15 CPU0 and CPU1 in.

[0251] In a specific implementation, as an example, Fig.15 As shown, the terminal device may include multiple processors, such as Fig.15 The first processor 401-1 and the second processor 401-2 in the embodiment of the present invention are shown in FIG. Each of these processors may be a single-core processor or a multi-core processor.

[0252] Fig.16 Schematic diagram of the structure of the chip provided by the embodiment of the present invention. Fig.16As shown, the chip 50 includes one or more (including two) processors 401 and a communication interface 402 .

[0253] Optional, such as Fig.16 As shown, the chip also includes a memory 404, which may include a read-only memory and a random access memory, and provides operation instructions and data to the processor. A portion of the memory may also include a non-volatile random access memory (NVRAM).

[0254] In some embodiments, Fig.16 As shown, the memory stores the following elements, execution modules or data structures, or their subsets, or their extended sets.

[0255] In the embodiment of the present invention, Fig.16 As shown, the corresponding operation is performed by calling the operation instruction stored in the memory (the operation instruction may be stored in the operating system).

[0256] like Fig.16 As shown, the processor controls the processing operations of any one of the terminal devices, and the processor can also be called a central processing unit (CPU).

[0257] like Fig.16 As shown, the memory may include a read-only memory and a random access memory, and provide instructions and data to the processor. A portion of the memory may also include NVRAM. For example, in an application, the memory, the communication interface, and the memory are coupled together through a bus system, wherein the bus system may include a power bus, a control bus, and a status signal bus in addition to a data bus. However, for the sake of clarity, in Fig.16 Various buses are labeled as bus system 405 .

[0258] like Fig.16As shown, the method disclosed in the above embodiment of the present invention can be applied to a processor or implemented by a processor. The processor may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by an integrated logic circuit of hardware in the processor or an instruction in the form of software. The above processor may be a general-purpose processor, a digital signal processor (digital signal processing, DSP), an ASIC, a field-programmable gate array (field-programmable gate array, FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The methods, steps and logic block diagrams disclosed in the embodiments of the present invention can be implemented or executed. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The steps of the method disclosed in the embodiment of the present invention can be directly embodied as being executed by a hardware decoding processor, or executed by a combination of hardware and software modules in the decoding processor. The software module may be located in a mature storage medium in the field such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register, etc. The storage medium is located in a memory, and the processor reads the information in the memory and completes the steps of the above method in combination with its hardware.

[0259] In one possible implementation, Fig.16 As shown, the communication interface is used to perform Figure 1 Step 101 of the method for establishing the near-field calculation model of extreme ultraviolet mask diffraction in the embodiment shown. The processor is used to execute Figure 1 The method for establishing the EUV mask diffraction near-field calculation model in the illustrated embodiment includes steps 102 and 103.

[0260] On the one hand, a computer-readable storage medium is provided, in which instructions are stored. When the instructions are executed, the functions performed by the technology level determination device in the above embodiment are implemented.

[0261] On the one hand, a chip is provided, which is applied to a terminal device, the chip includes at least one processor and a communication interface, the communication interface is coupled to at least one processor, and the processor is used to run instructions to implement the functions performed by the technology level determination device in the above embodiment.

[0262] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented by software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer programs or instructions. When the computer program or instruction is loaded and executed on a computer, the process or function described in the embodiment of the present invention is executed in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, a terminal, a user device or other programmable device. The computer program or instruction can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer program or instruction can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center by wired or wireless means. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server, data center, etc. that integrates one or more available media. The available medium can be a magnetic medium, such as a floppy disk, a hard disk, or a tape; it can also be an optical medium, such as a digital video disc (DVD); it can also be a semiconductor medium, such as a solid state drive (SSD).

[0263] Although the present invention is described herein in conjunction with various embodiments, in the process of implementing the claimed invention, those skilled in the art may understand and implement other variations of the disclosed embodiments by viewing the drawings, the disclosure, and the appended claims. In the claims, the word "comprising" does not exclude other components or steps, and "one" or "an" does not exclude multiple situations. A single processor or other unit may implement several functions listed in the claims. Certain measures are recorded in different dependent claims, but this does not mean that these measures cannot be combined to produce good results.

[0264] Although the present invention has been described in conjunction with specific features and embodiments thereof, it is apparent that various modifications and combinations may be made thereto without departing from the spirit and scope of the present invention. Accordingly, this specification and the accompanying drawings are merely exemplary illustrations of the present invention as defined by the appended claims and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present invention. Obviously, those skilled in the art may make various modifications and variations to the present invention without departing from the spirit and scope of the present invention. Thus, the present invention is intended to include such modifications and variations if they fall within the scope of the claims of the present invention and their equivalents.

Claims

1. A method for establishing a near-field calculation model of extreme ultraviolet mask diffraction, characterized in that: The method comprises: Obtaining a diffraction matrix database corresponding to a preset number of mask layouts, wherein each of the mask layouts corresponds to at least one diffraction matrix; Cutting and classifying each of the mask patterns based on a diffraction matrix database corresponding to the mask patterns to obtain multiple types of feature patterns; Based on the characteristic pattern of each type, determining a convolution kernel having a mask diffraction near-field characteristic; The near-field data corresponding to the test layout is determined according to the convolution kernel, and the establishment of the extreme ultraviolet light mask diffraction near-field calculation model is completed.

2. The method according to claim 1, characterized in that The method further comprises cutting and classifying each mask pattern based on the diffraction matrix database corresponding to the mask pattern to obtain multiple types of feature patterns, including: Based on the diffraction matrix corresponding to each of the mask layouts, the layout matrix of the mask layout is extended outward by a first preset number of pixels to obtain a corresponding enlarged layout matrix; Based on the diffraction matrix corresponding to each of the mask layouts, the layout matrix of the mask layout is reduced inward by a second preset number of pixels to obtain a corresponding reduced layout matrix; Determine a layout edge pattern matrix based on the enlarged layout matrix and the reduced layout matrix; The layout edge pattern matrix is ​​segmented according to a plurality of outer edge features to obtain the plurality of types of feature patterns.

3. The method according to claim 2, characterized in that The step of segmenting the layout edge pattern matrix according to a plurality of outer edge features to obtain the plurality of types of feature patterns includes: Segmenting the layout edge pattern matrix according to a plurality of outer edge features to obtain a plurality of feature two-dimensional matrices having corresponding edge features; The plurality of feature two-dimensional matrices are secondary classified according to different feature points of the plurality of outer edge features to obtain the plurality of types of feature graphics.

4. The method according to claim 3, characterized in that: The diffraction matrix includes an XX diffraction matrix, a YY diffraction matrix, an XY diffraction matrix and a YX diffraction matrix; the convolution kernel includes a diffraction near-field real part convolution kernel and a diffraction near-field imaginary part convolution kernel; the convolution kernel having the mask diffraction near-field characteristic is determined based on the characteristic pattern of each type, including: Based on the characteristic pattern of each category, by extracting pixel points in the characteristic pattern of each category, determining a pixel point coordinate set of the characteristic pattern of each category; Determine, based on the pixel point coordinate set, the pixel point coordinates of the characteristic pattern of each type corresponding to the near field data in the XX diffraction matrix, the YY diffraction matrix, the XY diffraction matrix and the YX diffraction matrix; Based on the diffraction characteristics of the XX diffraction matrix, the diffraction characteristics of the YY diffraction matrix and the diffraction characteristics of the intermediate reduced layout, the near-field data of the XX diffraction matrix and the near-field data of the YY diffraction matrix are fitted to determine a first numerical value of first convolution kernels and a second numerical value of second convolution kernels having the outer edge feature; Based on the diffraction characteristics of the XY diffraction matrix, the diffraction characteristics of the YX diffraction matrix, and the diffraction characteristics of the intermediate reduced layout, the near-field data of the XY diffraction matrix and the near-field data of the YX diffraction matrix are fitted to determine a third numerical value of third convolution kernels and a fourth numerical value of fourth convolution kernels having the outer edge feature; A diffraction near-field real convolution kernel and a diffraction near-field imaginary convolution kernel are determined based on a first numerical value of the first convolution kernel, a second numerical value of the second convolution kernel, a third numerical value of the third convolution kernel, and a fourth numerical value of the fourth convolution kernel.

5. The method according to any one of claims 1 to 4, characterized in that: The convolution kernel includes a diffraction near-field real convolution kernel and a diffraction near-field imaginary convolution kernel; the near-field data corresponding to the test layout is determined according to the convolution kernel to complete the establishment of the extreme ultraviolet mask diffraction near-field calculation model, including: Obtaining a test diffraction matrix corresponding to the test pattern; Cutting and classifying the test layout based on the test diffraction matrix corresponding to the test layout to obtain various types of characteristic test patterns; Determine the diffraction near-field real part data corresponding to each type of characteristic test pattern based on the diffraction near-field real part convolution kernel; Determine the diffraction near-field imaginary part data corresponding to each type of characteristic test pattern based on the diffraction near-field imaginary part convolution kernel; The near-field data is determined based on the diffraction near-field real data and the diffraction near-field imaginary data, and the establishment of the extreme ultraviolet light mask diffraction near-field calculation model is completed.

6. The method according to any one of claims 1 to 4, characterized in that: The step of obtaining a diffraction matrix database corresponding to a preset number of mask layouts, wherein each of the mask layouts corresponds to at least one diffraction matrix, comprises: A diffraction matrix database corresponding to a preset number of mask patterns is determined based on a time-domain finite difference electromagnetic calculation method, wherein each mask pattern corresponds to at least one diffraction matrix.

7. A device for establishing a near-field calculation model of extreme ultraviolet mask diffraction, characterized in that: The device comprises: An acquisition module, used for acquiring a preset number of diffraction matrix databases corresponding to mask patterns, wherein each of the mask patterns corresponds to at least one diffraction matrix database, wherein each of the mask patterns corresponds to at least one diffraction matrix; A cutting module, used for cutting and classifying each of the mask patterns based on a diffraction matrix database corresponding to the mask patterns to obtain multiple types of feature patterns; A determination module, used for determining a convolution kernel having a mask diffraction near-field characteristic based on the characteristic pattern of each type; The model building module is used to determine the near-field data corresponding to the test layout according to the convolution kernel, and complete the establishment of the extreme ultraviolet light mask diffraction near-field calculation model.

8. The device according to claim 7, characterized in that The cutting module comprises: An outward extension submodule, used for extending the layout matrix of the mask layout outward by a first preset number of pixels based on the diffraction matrix corresponding to each of the mask layouts, to obtain a corresponding enlarged layout matrix; An inward reduction submodule, used for reducing the layout matrix of the mask layout inward by a second preset number of pixels based on the diffraction matrix corresponding to each of the mask layouts, to obtain a corresponding reduced layout matrix; A first determination submodule is used to determine a layout edge pattern matrix based on the enlarged layout matrix and the reduced layout matrix; The segmentation submodule is used to segment the layout edge pattern matrix according to multiple outer edge features to obtain the multiple types of feature patterns.

9. The device according to claim 8, characterized in that The segmentation submodule comprises: A segmentation unit, used for segmenting the layout edge pattern matrix according to a plurality of outer edge features to obtain a plurality of feature two-dimensional matrices with corresponding edge features; The classification unit is used to perform secondary classification on the plurality of feature two-dimensional matrices according to different feature points of the plurality of outer edge features to obtain the plurality of types of feature graphics.

10. An electronic device, characterized in that: include: one or more processors; and one or more machine-readable media having instructions stored thereon, which, when executed by the one or more processors, enable the method for establishing the extreme ultraviolet light mask diffraction near-field calculation model as described in any one of claims 1 to 6 to be executed.

Images