A method and apparatus for optical proximity correction

Abstract

The application discloses a kind of related to integrated circuit production field, especially a kind of optical proximity correction method and device, by receiving wafer surface topography pattern, target wafer pattern and illumination light intensity;The wafer surface topography pattern is input into the pre-trained photoresist thickness distribution model with the nominal photoresist thickness of pre-set, and the corresponding photoresist thickness distribution is obtained;According to the photoresist thickness distribution, the reflectivity distribution is calculated;By the illumination light intensity and current mask pattern, the incident light intensity distribution is calculated;According to the reflectivity distribution and the incident light intensity distribution, the optical image distribution is calculated;Photoresist pattern correction is carried out to the optical image distribution, and the photoresist image is obtained;The edge placement error of the photoresist image corresponding to the target wafer pattern is calculated;According to the edge placement error, the edge section of the current mask pattern is corrected.The application improves the accuracy and precision of the final photoresist image.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of integrated circuit production, and in particular to an optical proximity correction method and device. BACKGROUND

[0002] In the photolithography process of large scale integrated circuit manufacturing, the chip pattern on the mask is imaged in the photosensitive photoresist material through the illumination and imaging system, and the pattern is formed after development. The photoresist is a thin film which is substantially transparent to the wavelength of the incident light source. The incident light entering the photoresist thin film interferes with the reflected light at the bottom of the thin film, forming a standing wave in the photoresist thin film layer. According to the thin film interference theory, when the thickness of the photoresist is different, the equivalent reflectivity of the incident light at the air-photoresist interface also changes, causing the change of the light intensity (i.e. exposure energy) coupled into the photoresist, and further causing the difference between the photoresist region where the photochemical change occurs and the preset target region, that is, it will cause the change of the wafer critical dimension after lithography. As shown in Figure 1 , the relationship curve of the equivalent reflectivity reff of the incident light on the upper surface of the photoresist (i.e. the ratio of all reflected light energy to incident light energy) and the thickness t of the photoresist generally has a sinusoidal curve characteristic. Correspondingly, the relationship between the exposure energy and the wafer critical dimension after lithography and the thickness of the photoresist is also similar to Figure 1 . Therefore, in order to reduce the change of the wafer critical dimension after lithography, in the photoresist spin coating process, the process parameters need to be optimized to make the photoresist thickness uniform on the entire wafer. However, in the actual chip manufacturing process, the "pattern" left by the previous process makes the surface of the wafer not completely flat, and the wafer topography will cause the non-uniformity of the photoresist thickness. As shown in Figure 2 , the photoresist thickness will fluctuate when it crosses the "step" on the wafer surface, and this change in photoresist thickness will cause changes in the wafer critical dimension and three-dimensional topography after lithography, resulting in loss of process window and potential defects.

[0003] In order to eliminate the influence of wafer topography on the lithography process, a planarization layer is generally introduced on the wafer to reduce the surface fluctuation, so as to obtain uniform photoresist thickness. At the same time, a bottom anti-reflective coating (BARC) is introduced below the photoresist layer to further reduce the change of exposure energy caused by the change of photoresist thickness. However, for the lithography process of some layers (such as ion implantation layer), due to the limitation of process integration and cost control considerations, it is not possible to introduce a planarization layer or a bottom anti-reflective coating. For the lithography process of these layers, the fluctuating wafer topography is the most important challenge for the control of the lithography process.

[0004] Therefore, how to further compensate the photoresist pattern change caused by the different photoresist thickness of the wafer surface topography in optical simulation, improve the precision and accuracy of optical proximity correction, and reduce the cost, have good universality, the problem of the person skilled in the art is urgent to solve. SUMMARY

[0005] The purpose of the present application is to provide an optical proximity correction method and device to solve the problem that new structures must be added in the prior art to compensate for the photoresist pattern change caused by the different photoresist thickness of the wafer surface topography, which increases production cost, has poor universality and poor compensation effect.

[0006] To solve the above technical problems, the present application provides an optical proximity correction method, comprising:

[0007] Receiving a wafer surface topography pattern, a target wafer pattern and an illumination light intensity;

[0008] Inputting the wafer surface topography pattern and a preset nominal photoresist thickness into a pre-trained photoresist thickness distribution model to obtain a corresponding photoresist thickness distribution;

[0009] Calculating a reflectivity distribution according to the photoresist thickness distribution;

[0010] Calculating an incident light intensity distribution by the illumination light intensity and a current mask pattern; the initial current mask pattern is the same as the target wafer pattern; the current mask pattern comprises a plurality of edge segments;

[0011] Calculating an optical image distribution according to the reflectivity distribution and the incident light intensity distribution;

[0012] Correcting the optical image distribution to obtain a photoresist image;

[0013] Calculating an edge placement error of the photoresist image corresponding to the target wafer pattern;

[0014] According to the edge placement error, correcting the edge segments of the current mask pattern.

[0015] Optionally, in the optical proximity correction method, calculating an optical image distribution according to the reflectivity distribution and the incident light intensity distribution comprises:

[0016] Calculating the optical image distribution by the following formula:

[0017] I f(x,y) =I i(x,y) ×(1-r eff(x,y) );

[0018] Wherein, I f(x,y)is the optical image distribution; r eff(x,y) is the reflectivity distribution; I i(x,y) is the incident light intensity.

[0019] Optionally, in the optical proximity correction method, the calculating the optical image distribution according to the reflectivity distribution and the incident light intensity distribution comprises:

[0020] calculating the nominal reflectivity and the nominal optical image distribution corresponding to the current mask pattern under the condition corresponding to the nominal photoresist thickness and the incident light intensity distribution;

[0021] calculating the optical image distribution according to the nominal optical image distribution, the nominal reflectivity and the reflectivity distribution.

[0022] Optionally, in the optical proximity correction method, the calculating the nominal optical image distribution corresponding to the current mask pattern under the condition corresponding to the nominal photoresist thickness and the incident light intensity distribution comprises:

[0023] the nominal optical image distribution is calculated by the following formula:

[0024] I f 0 (x,y) =I i(x,y) ×(1-r eff 0 (x,y) );

[0025] wherein, I f 0 (x,y) is the nominal optical image distribution; r eff 0 (x,y) is the nominal reflectivity; I i(x,y) is the incident light intensity.

[0026] Optionally, in the optical proximity correction method, the calculating the optical image distribution according to the nominal optical image distribution, the nominal reflectivity and the reflectivity distribution comprises:

[0027] the optical image distribution is calculated by the following formula:

[0028] I f(x,y) =I f 0 (x,y) ×[(1-r eff(x,y) ) / (1-r eff 0 (x,y) )];

[0029] wherein, I f(x,y)is the optical image distribution; I f 0 (x,y) is the nominal optical image distribution; r eff(x,y) is the reflectivity distribution; r eff 0 (x,y) is the nominal reflectivity.

[0030] Optionally, in the optical proximity correction method, the optical image distribution is corrected by a photoresist pattern to obtain a photoresist image, including:

[0031] The photoresist image is obtained by the following formula:

[0032] ;

[0033] wherein R(x, y) is the photoresist image; F j is the jth photoresist model term based on the photoresist pattern correction; c j is the F j corresponding linear coefficient; I f(x,y) is the optical image distribution.

[0034] Optionally, in the optical proximity correction method, before the edge placement error is used to correct the edge segment of the current mask pattern, the method further includes:

[0035] determining whether the edge placement error of the photoresist image corresponding to the target wafer pattern exceeds a preset error tolerance threshold;

[0036] Correspondingly, the correction of the edge segment of the current mask pattern according to the edge placement error includes:

[0037] When the edge placement error exceeds the preset error tolerance threshold, the edge segment of the current mask pattern is adjusted according to the edge placement error to obtain an updated current mask pattern, and the edge placement error corresponding to the updated current mask pattern is calculated until the edge placement error does not exceed the error tolerance threshold.

[0038] Optionally, in the optical proximity correction method, the training method of the photoresist thickness distribution model includes:

[0039] establishing an uncorrected photoresist thickness model; the photoresist thickness model can first perform a polygon expansion operation on an input wafer surface topography pattern to obtain a polygon expansion pattern, then calculate a preliminary upper surface height of the polygon expansion pattern respectively through a preset thickness increase coefficient, then perform convolution processing on the preliminary upper surface height through a preset Gaussian smoothing coefficient to obtain a smoothed upper surface height, and determine a photoresist thickness distribution through the smoothed upper surface height;

[0040] According to a plurality of sample surface topography patterns, a plurality of photoresist laying samples are prepared, and sample photoresist thickness distributions corresponding to the photoresist laying samples are obtained;

[0041] Using the sample surface topography patterns and the corresponding sample photoresist thickness distributions, the thickness increase coefficient, the Gaussian smoothing coefficient and the polygon expansion operation are calibrated, so that the difference between the simulated photoresist thickness distribution obtained by inputting the sample surface topography pattern and the corresponding sample photoresist thickness distribution is less than a preset calibration tolerance threshold.

[0042] Optionally, in the optical proximity correction method, when the loss function value of the following formula is less than the preset calibration tolerance threshold, it is considered that the difference between the simulated photoresist thickness distribution obtained by inputting the sample surface topography pattern and the corresponding sample photoresist thickness distribution is less than the preset calibration tolerance threshold:

[0043] ;

[0044] Wherein, cost is the loss function value; W is the convex pattern size in the sample surface topography pattern; S is the convex pattern pitch in the sample surface topography pattern; is the difference between the photoresist thickness in the simulated photoresist thickness distribution and the photoresist thickness in the sample photoresist thickness distribution at the corresponding position of the convex pattern in the sample surface topography pattern under the corresponding convex pattern size and convex pattern pitch; is the difference between the photoresist thickness in the simulated photoresist thickness distribution and the photoresist thickness in the sample photoresist thickness distribution at the corresponding position between two adjacent expanded convex patterns in the polygon expansion pattern corresponding to the sample surface topography pattern under the corresponding convex pattern size and convex pattern pitch.

[0045] An optical proximity correction device comprises:

[0046] A receiving module for receiving a wafer surface topography pattern, a target wafer pattern and an illumination light intensity;

[0047] A simulation module for inputting the wafer surface topography pattern and a preset nominal photoresist thickness into a pre-trained photoresist thickness distribution model to obtain a corresponding photoresist thickness distribution;

[0048] a reflectivity module configured to calculate a reflectivity distribution according to the photoresist thickness distribution;

[0049] an incident light intensity module configured to calculate an incident light intensity distribution by the illumination light intensity and a current mask pattern, the initial current mask pattern being the same as the target wafer pattern, the current mask pattern including a plurality of edge segments;

[0050] an optical image module configured to calculate an optical image distribution according to the reflectivity distribution and the incident light intensity distribution;

[0051] a photoresist module configured to correct a photoresist pattern of the optical image distribution to obtain a photoresist image;

[0052] an EPE module configured to calculate an edge placement error of the photoresist image corresponding to each of the edge segments of the target wafer pattern;

[0053] a correction module configured to correct the edge segments of the current mask pattern according to the edge placement error.

[0054] The optical proximity correction method provided by the present application receives a wafer surface topography pattern, a target wafer pattern and an illumination light intensity, inputs the wafer surface topography pattern and a preset nominal photoresist thickness into a pre-trained photoresist thickness distribution model to obtain a corresponding photoresist thickness distribution, calculates a reflectivity distribution according to the photoresist thickness distribution, calculates an incident light intensity distribution by the illumination light intensity and a current mask pattern, the initial current mask pattern being the same as the target wafer pattern, the current mask pattern including a plurality of edge segments, calculates an optical image distribution according to the reflectivity distribution and the incident light intensity distribution, corrects a photoresist pattern of the optical image distribution to obtain a photoresist image, calculates an edge placement error of the photoresist image corresponding to the target wafer pattern, and corrects the edge segments of the current mask pattern according to the edge placement error.

[0055] In the present application, after the incident light intensity distribution of the irradiation light intensity transmitted through the current mask pattern is calculated, the photoresist image is not directly calculated by using the incident light intensity distribution, but the thickness of each position of the photoresist is further calculated by a photoresist thickness distribution model (i.e. photoresist thickness distribution), and then the reflectivity of each position is calculated by the thickness (i.e. reflectivity distribution), and then the photochemical reaction of the photoresist at each position is calculated according to the reflectivity distribution and the incident light intensity distribution, so that the photochemical reaction of the photoresist at which positions can be more accurately judged, thereby greatly improving the accuracy and precision of the finally obtained photoresist image, and further improving the calculation accuracy of the edge placement error, i.e. improving the accuracy and precision of the optical proximity correction. The present application also provides an optical proximity correction device with the above beneficial effects. BRIEF DESCRIPTION OF DRAWINGS

[0056] In order to more clearly illustrate the technical solutions of the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or the prior art description will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present application, and those skilled in the art can also obtain other drawings according to these drawings without creative labor.

[0057] Figure 1 A schematic diagram of the relationship between the equivalent reflectivity of incident light on the surface of the photoresist and the thickness of the photoresist;

[0058] Figure 2 A schematic diagram of the structure of the photoresist laid on the surface of the wafer;

[0059] Figure 3 A flowchart of one specific embodiment of the optical proximity correction method provided by the present application;

[0060] Figure 4 A process structure schematic diagram of one specific embodiment of the optical proximity correction method provided by the present application;

[0061] Figure 5 A process structure schematic diagram of another specific embodiment of the optical proximity correction method provided by the present application;

[0062] Figure 6 A flowchart of another specific embodiment of the optical proximity correction method provided by the present application;

[0063] Figure 7 A structure schematic diagram of one specific embodiment of the optical proximity correction device provided by the present application.

[0064] Reference signs:

[0065] 100 - receiving module; 200 - simulation module; 300 - reflectivity module; 400 - incident light intensity module; 500 - optical image module; 600 - photoresist module; 700 - EPE module; 800 - correction module. DETAILED DESCRIPTION

[0066] Although the wafer surface topography has a significant impact on the lithography process as described above, the impact of wafer surface topography is usually ignored in the model-based optical proximity correction (OPC). In the optical proximity correction, the optical image of the current mask pattern is obtained by model simulation and further calculation of the photoresist image, and then the edge placement error (EPE) is calculated based on the intensity of the target pattern edge in the photoresist image, and finally the mask pattern is corrected according to the correction amount calculated based on the edge placement error. In the optical proximity correction model, the photoresist thickness is usually defined as a fixed value, i.e. the thickness of the photoresist on a flat wafer. Therefore, when using the prior art to perform optical proximity correction, the effect of wafer surface topography is not compensated for, and when the wafer surface is not flat, a large correction accuracy error and lithography process window loss will be caused.

[0067] In order for those skilled in the art to better understand the present application, the present application will be further described in detail below in combination with the drawings and specific embodiments. Obviously, the described embodiments are only part of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0068] The core of the present application is to provide an optical proximity correction method, a flowchart of one specific embodiment of which is shown in Figure 3 As shown in the figure, it is called embodiment one, which comprises:

[0069] S101: receiving a wafer surface topography pattern, a target wafer pattern and an illumination light intensity.

[0070] The lens surface topography pattern is the pattern information recording the undulating condition of the wafer surface on which the photoresist layer is to be set.

[0071] The illumination light intensity is the illumination intensity of the incident light on the mask plate.

[0072] S102: inputting the wafer surface topography pattern and the preset nominal photoresist thickness into the pre-trained photoresist thickness distribution model to obtain the corresponding photoresist thickness distribution.

[0073] Of course, the material properties of the photoresist itself are also pre-input into the photoresist thickness distribution model, and after this step, the thickness distribution information of the photoresist in the corresponding area of the wafer surface topography pattern is obtained, that is, the photoresist thickness distribution.

[0074] S103: Calculate the reflectivity distribution according to the photoresist thickness distribution.

[0075] As described above, different photoresist thicknesses will cause the photoresist to absorb different energies of incident light, which will cause the photoresist area that undergoes photochemical reaction to change, resulting in a decrease in photoetching pattern transfer accuracy. Therefore, in this step, the reflectivity distribution of the photoresist at each position is calculated, which can facilitate the subsequent separate calculation of how much energy of the incident light is absorbed at each position and whether the photochemical reaction standard is reached.

[0076] The equivalent reflectivity of the photoresist at each position can be calculated based on the refractive index, absorption coefficient and thickness of each layer of film on the photoresist according to thin film theory, which will not be described here.

[0077] S104: Calculate the incident light intensity distribution by the irradiation light intensity and the current mask pattern; the initial current mask pattern is the same as the target wafer pattern; the current mask pattern includes multiple edge segments.

[0078] The edge segment can move in the direction of the corresponding normal line, which is used to adjust the edge position of the current mask pattern in the subsequent step.

[0079] The incident light intensity distribution, that is, the light intensity distribution left on the surface on which the photoresist is laid under the shielding of the current mask pattern.

[0080] S105: Calculate the optical image distribution according to the reflectivity distribution and the incident light intensity distribution.

[0081] The optical image distribution in this step refers to the distribution of energy coupled into the photoresist film.

[0082] Specifically, this step includes:

[0083] The optical image distribution is calculated by the following formula (1):

[0084] I f(x,y) =I i(x,y) ×(1-r eff(x,y) ); (1)

[0085] Where I f(x,y) is the optical image distribution; r eff(x,y) is the reflectivity distribution; and I i(x,y) is the incident light intensity.

[0086] It can be seen that, for any position (x, y), the method for calculating the value of the corresponding optical image distribution is to find the value of the corresponding incident light intensity (i.e. I i(x,y) ) through the incident optical distribution in the previous step, and then subtract the part reflected away from the value of the incident light intensity, and the remaining part is the energy absorbed by the photoresist in the incident light intensity.

[0087] S106: Perform photoresist pattern correction on the optical image distribution to obtain a photoresist image.

[0088] As a specific embodiment, performing photoresist pattern correction on the optical image distribution to obtain a photoresist image comprises:

[0089] The photoresist image is obtained by the following formula (2):

[0090] ; (2)

[0091] Wherein, R(x, y) is the photoresist image; F j is the jth photoresist model term based on the photoresist pattern correction; c j is the corresponding linear coefficient of F j ; I f(x,y) is the optical image distribution.

[0092] This step can refer to existing related technologies. Each of the photoresist model terms corresponds to a corresponding operation on the optical image distribution, such as a first photoresist model term for sharpening the optical image distribution, a second photoresist model term for blunting the optical image distribution, and a third photoresist model term for light acid diffusion operation on the optical image distribution. The present application will not be described in more detail. After the calculation of each photoresist model term is completed, the linear superposition of all photoresist model terms is performed, and the photoresist image is obtained. Through the preset edge threshold, the simulated mask pattern shape can be outlined on the photoresist image.

[0093] S107: Calculate the edge placement error of the photoresist image corresponding to the target wafer pattern.

[0094] The edge placement error of the photoresist image corresponding to the target wafer pattern in this step refers to the edge placement error between the simulated mask pattern shape outlined on the photoresist image according to the edge threshold and the target wafer pattern.

[0095] S108: According to the edge placement error, correct the edge segment of the current mask pattern.

[0096] The modification of the edge segment of the current mask pattern can refer to the existing related art. For example, the edge placement error of a certain edge segment is -1.5 nm, that is, the edge of the simulated mask pattern shape is inwardly shrunk by 1.5 nm relative to the edge segment at the corresponding position of the target wafer pattern. In the adjustment of the current mask pattern, the corresponding edge segment position can be outwardly adjusted by 1.5 nm. Of course, the adjustment distance does not necessarily have to be consistent with the edge placement error of the edge segment. The edge placement error of the edge segment can be adjusted by a corresponding coefficient or by other methods according to the edge placement error. The present application does not make too much repetition here.

[0097] As a preferred embodiment, before the modification of the edge segment of the current mask pattern according to the edge placement error, the method further comprises:

[0098] A1: judging whether the edge placement error of the photoresist image corresponding to the target wafer pattern exceeds a preset error tolerance threshold.

[0099] Specifically, the error tolerance threshold can correspond to the sum of the edge placement errors of all edge segments, that is, the present step can be case one:

[0100] judging whether the sum of the edge placement errors of all edge segments of the photoresist image corresponding to the target wafer pattern exceeds a preset error tolerance threshold.

[0101] Of course, the present step can also be case two:

[0102] judging whether the average value of the edge placement errors of all edge segments of the photoresist image corresponding to the target wafer pattern exceeds a preset error tolerance threshold.

[0103] Alternatively, the present step can also be case three:

[0104] judging whether the edge placement error of any edge segment of the photoresist image corresponding to the target wafer pattern does not exceed a preset error tolerance threshold.

[0105] Of course, the value of the error tolerance threshold can be different for the above different cases.

[0106] Correspondingly, the modification of the edge segment of the current mask pattern according to the edge placement error comprises:

[0107] A2: when the edge placement error exceeds a preset error tolerance threshold, adjusting the edge segment of the current mask pattern according to the edge placement error to obtain an updated current mask pattern, and calculating the edge placement error corresponding to the updated current mask pattern until the edge placement error does not exceed the error tolerance threshold.

[0108] Corresponding to case three in the last step, the execution condition in this step should be changed to "when there is an edge placement error corresponding to the edge segment of the target wafer pattern of the photoresist image exceeding the preset error tolerance threshold……".

[0109] The update cycle of the current mask pattern is set in this step, and the update of the current mask pattern will be stopped only after the edge placement error does not exceed the error tolerance threshold, thereby greatly improving the similarity between the pattern left on the wafer after lithography of the finally obtained current mask pattern and the target wafer pattern.

[0110] Further, a cycle threshold can also be set, that is, when the number of update adjustment times of the current mask pattern exceeds the cycle threshold, the adjustment and update are no longer continued, and the current mask pattern at this time is output as the final mask pattern, thereby improving the calculation efficiency.

[0111] The training method of the photoresist thickness distribution model is described below, and the training method of the photoresist thickness distribution model comprises:

[0112] B1: Establish an uncorrected photoresist thickness model; the photoresist thickness model can first perform a polygon expansion operation on the input wafer surface topography pattern to obtain a polygon expansion pattern, then calculate the preliminary upper surface height of the polygon expansion pattern through a preset thickness increase coefficient, then perform convolution processing on the preliminary upper surface height through a preset Gaussian smoothing coefficient to obtain a smoothed upper surface height, and determine the photoresist thickness distribution through the smoothed upper surface height.

[0113] Reference can be made to Figure 4 After the photoresist film is set, the area where the photoresist film protrudes does not strictly coincide with the edge of the protruding pattern of the wafer surface topography pattern, but will expand to the outside of the pattern (the width of the expansion is denoted by L in Figure 4 Corresponding to this, it is referred to as the "polygon expansion operation" mentioned in this step, that is, based on the edge of the protruding pattern, the pattern is expanded outward, and the expanded edge is taken as the edge of the pattern where the photoresist film protrudes.

[0114] Specifically, the input sample wafer surface topography pattern is polygonally expanded by the following formula (3):

[0115] ; (3)

[0116] wherein, is the expanded topography pattern; is the sample wafer surface topography pattern (i.e., the protruding pattern in the foregoing); S LFor the preset polygon expansion operation, the edge of all the patterns in the polygon can be expanded outward by a distance L according to a corresponding normal method.

[0117] Further, the preliminary upper surface height of the polygon expansion pattern is calculated by a preset thickness increasing coefficient, and the preliminary upper surface height is convoluted by a preset Gaussian smoothing coefficient to obtain a smooth upper surface height, including:

[0118] The preliminary upper surface height of the convex pattern corresponding region of the wafer surface topography pattern is calculated by the following formula (4):

[0119] ; (4)

[0120] wherein, is the preliminary upper surface height, and T is a preset nominal photoresist thickness.

[0121] The preliminary upper surface height of the non-convex pattern corresponding region of the wafer surface topography pattern is calculated by the following formula (5):

[0122] ; (5)

[0123] wherein, is the preliminary upper surface height, and T is a preset nominal photoresist thickness, is a thickness increasing coefficient; and D is the thickness of the convex pattern of the wafer surface topography pattern.

[0124] Under the premise of the above formula (4) and formula (5), the preliminary upper surface is convoluted by the following formula (6) to obtain a smooth upper surface height, and the upper surface height of the expansion topography pattern but not belonging to the convex pattern region is filled:

[0125] ; (6)

[0126] wherein, is the smooth upper surface height; is the Gaussian smoothing coefficient; is a convolution operation; is the preliminary upper surface height.

[0127] The photoresist thickness distribution is determined by the smooth upper surface height, including:

[0128] The photoresist thickness of the convex pattern corresponding region of the wafer surface topography pattern is calculated by the following formula (7):

[0129] ​​(7)

[0130] wherein, is the photoresist thickness; is the height of the smooth upper surface.

[0131] The photoresist thickness of the convex pattern corresponding area of the wafer surface topography pattern is calculated by using the following formula (8):

[0132] (8)

[0133] wherein, is the photoresist thickness; is the height of the smooth upper surface; and D is the thickness of the convex pattern of the wafer surface topography pattern.

[0134] B2: A plurality of photoresist laying samples are made according to a plurality of preset sample surface topography patterns, and a sample photoresist thickness distribution corresponding to each photoresist laying sample is obtained.

[0135] The sample photoresist thickness distribution obtained will be used as a correction basis for the corresponding sample surface topography pattern in the simulation below.

[0136] B3: The thickness increase coefficient, the Gaussian smoothing coefficient and the polygon expansion operation are calibrated using the sample surface topography pattern and the corresponding sample photoresist thickness distribution, so that the difference between the simulated photoresist thickness distribution obtained by inputting the sample surface topography pattern and the corresponding sample photoresist thickness distribution is less than a preset calibration tolerance threshold.

[0137] The calibration of the above parameters can refer to the existing related technology, which will not be described here in detail. Specifically, when the loss function value of the following formula (9) is less than the preset calibration tolerance threshold, it is considered that the difference between the simulated photoresist thickness distribution obtained by inputting the sample surface topography pattern and the corresponding sample photoresist thickness distribution is less than the preset calibration tolerance threshold:

[0138] (9)

[0139] wherein, cost is the loss function value; W is the convex pattern size in the sample surface topography pattern; S is the convex pattern spacing in the sample surface topography pattern; is the difference between the photoresist thickness in the simulated photoresist thickness distribution and the photoresist thickness in the sample photoresist thickness distribution at the corresponding position of the convex pattern in the sample surface topography pattern under the corresponding convex pattern size and convex pattern spacing; The difference between the photoresist thickness in the simulation photoresist thickness distribution and the photoresist thickness in the sample photoresist thickness distribution at the corresponding position between two adjacent extended convex patterns in the polygonal extended pattern corresponding to the sample surface topography pattern under the corresponding convex pattern size and convex pattern pitch.

[0140] In the specific embodiment, the loss function for calibrating the above parameters is further defined, specifically, the thickness difference at the corresponding position of the extended convex pattern and the thickness difference at the corresponding position between the extended convex patterns are calculated respectively, which can be referred to as Figure 5 , Figure 5 The corresponding position t1 in the specific embodiment is identified in , and The corresponding position t2 is obtained, and in order to ensure that the obtained difference only shows the difference size, the obtained thickness difference is squared and then summed, so that the obtained loss function has higher representativeness, and the simulation accuracy of the photoresist thickness distribution model obtained by training is higher and closer to the actual situation. Of course, other ways can also be used to correct the above parameters, which are not limited in the present application.

[0141] The optical proximity correction method provided by the present application comprises the following steps: receiving a wafer surface topography pattern, a target wafer pattern and illumination light intensity; inputting the wafer surface topography pattern and a preset nominal photoresist thickness into a pre-trained photoresist thickness distribution model to obtain a corresponding photoresist thickness distribution; calculating a reflectivity distribution according to the photoresist thickness distribution; calculating an incident light intensity distribution by the illumination light intensity and a current mask pattern; the initial current mask pattern is the same as the target wafer pattern; the current mask pattern comprises a plurality of edge segments; calculating an optical image distribution according to the reflectivity distribution and the incident light intensity distribution; performing photoresist pattern correction on the optical image distribution to obtain a photoresist image; calculating an edge placement error of the photoresist image corresponding to the target wafer pattern; and correcting the edge segments of the current mask pattern according to the edge placement error. In the present application, after the incident light intensity distribution of the illumination light intensity transmitted through the current mask pattern is calculated, the incident light intensity distribution is not directly used to calculate the photoresist image, but the thickness of each position of the photoresist is further calculated by the photoresist thickness distribution model (i.e. the photoresist thickness distribution), and then the reflectivity of each position is calculated by the thickness (i.e. the reflectivity distribution), and then the light absorption of the photoresist at each position is calculated according to the reflectivity distribution and the incident light intensity distribution. Therefore, the photochemical reaction of the photoresist at which positions can be more accurately judged, so that the accuracy and precision of the finally obtained photoresist image are greatly improved, and the calculation accuracy of the edge placement error is also improved, that is, the accuracy and precision of the optical proximity correction are improved.

[0142] On the basis of the first embodiment, the method for obtaining the optical image distribution is further limited to obtain the second embodiment, and the corresponding flowchart is as shown in Figure 6 , which comprises:

[0143] S201: receiving a wafer surface topography pattern, a target wafer pattern and an illumination light intensity.

[0144] S202: inputting the wafer surface topography pattern and a preset nominal photoresist thickness into a pre-trained photoresist thickness distribution model to obtain a corresponding photoresist thickness distribution.

[0145] S203: calculating a reflectivity distribution according to the photoresist thickness distribution.

[0146] S204: calculating an incident light intensity distribution by the illumination light intensity and a current mask pattern; the initial current mask pattern is the same as the target wafer pattern; the current mask pattern comprises a plurality of edge segments.

[0147] S205: calculating a nominal reflectivity and a nominal optical image distribution corresponding to the current mask pattern under the conditions of the nominal photoresist thickness and the incident light intensity distribution.

[0148] In this step, the nominal optical image distribution corresponding to the current mask pattern under the conditions of the nominal photoresist thickness and the incident light intensity distribution is calculated, comprising:

[0149] The nominal optical image distribution is calculated by the following formula (10):

[0150] I f 0 (x,y) =I i(x,y) ×(1-r eff 0 (x,y) ); (10)

[0151] Wherein, I f 0 (x,y) is the nominal optical image distribution; r eff 0 (x,y) is the nominal reflectivity; I i(x,y) is the incident light intensity.

[0152] The formula (10) can greatly improve the calculation efficiency by referring to the formula (1) in the foregoing, which will not be described herein.

[0153] Since the thickness of the photoresist film is assumed to be uniform and consistent with the nominal photoresist thickness in this step, the nominal reflectivity at each position is also the same, and the specific calculation method will not be described herein.

[0154] S206: Calculate the optical image distribution according to the nominal optical image distribution, the nominal reflectivity, and the reflectivity distribution.

[0155] As a preferred embodiment, this step includes:

[0156] The optical image distribution is calculated by the following formula (11):

[0157] I f(x,y) =I f 0 (x,y) ×[(1-r eff(x,y) ) / (1-r eff 0 (x,y) )]; (11)

[0158] Wherein, I f(x,y) is the optical image distribution; I f 0 (x,y) is the nominal optical image distribution; r eff(x,y) is the reflectivity distribution; r eff 0 (x,y) is the nominal reflectivity.

[0159] This preferred embodiment can further improve the calculation efficiency.

[0160] S207: Perform photoresist pattern correction on the optical image distribution to obtain a photoresist image.

[0161] S208: Calculate the edge placement error of the photoresist image corresponding to the target wafer pattern.

[0162] S209: According to the edge placement error, correct the edge segment of the current mask pattern.

[0163] The difference between this specific embodiment and the above specific embodiment is that the calculation method of the optical image distribution is limited in this specific embodiment, and the remaining steps are the same as the above specific embodiment, which will not be described herein.

[0164] In the embodiment, the nominal optical image distribution is calculated first, and then the final optical image distribution is obtained according to the reflectivity distribution of the photoresist film of different thicknesses at each position obtained by simulation. Because the thickness of the photoresist film is consistent in the existing simulation process, the nominal optical image distribution can be automatically obtained in the prior art, and the reflectivity distribution of the photoresist film of different thicknesses at each position can be calculated by the staff on this basis. After the reflectivity distribution is obtained, the optical image distribution can be quickly obtained by formula (11). In other words, the optical proximity correction method provided by the preferred embodiment does not need to change the existing simulation calculation process, can greatly reduce the modification cost of the existing system, and improves the universality of the application.

[0165] The optical proximity correction device provided by the embodiment of the application is described below. The optical proximity correction device described below can be referred to in correspondence with the optical proximity correction method described above.

[0166] Figure 7 The structure block diagram of the optical proximity correction device provided by the embodiment of the application is described with reference to Figure 7 The optical proximity correction device can include:

[0167] The receiving module 100 is configured to receive a wafer surface topography pattern, a target wafer pattern, and illumination light intensity.

[0168] The simulation module 200 is configured to input the wafer surface topography pattern and a preset nominal photoresist thickness into a pre-trained photoresist thickness distribution model to obtain a corresponding photoresist thickness distribution.

[0169] The reflectivity module 300 is configured to calculate a reflectivity distribution according to the photoresist thickness distribution.

[0170] The incident light intensity module 400 is configured to calculate an incident light intensity distribution by using the illumination light intensity and a current mask pattern. The initial current mask pattern is the same as the target wafer pattern. The current mask pattern includes a plurality of edge segments.

[0171] The optical image module 500 is configured to calculate an optical image distribution according to the reflectivity distribution and the incident light intensity distribution.

[0172] The photoresist module 600 is configured to perform photoresist pattern correction on the optical image distribution to obtain a photoresist image.

[0173] The EPE module 700 is configured to calculate an edge placement error of each edge segment of the target wafer pattern corresponding to the photoresist image.

[0174] The correction module 800 is configured to correct the edge segment of the current mask pattern according to the edge placement error.

[0175] As a preferred embodiment, the optical image module 500 comprises:

[0176] The first formula unit is configured to calculate the optical image distribution by the following formula:

[0177] I f(x,y) =I i(x,y) ×(1-r eff(x,y) ); (1)

[0178] wherein I f(x,y) is the optical image distribution; r eff(x,y) is the reflectivity distribution; and I i(x,y) is the incident light intensity.

[0179] As a preferred embodiment, the optical image module 500 comprises:

[0180] The nominal thickness calculation unit is configured to calculate the nominal reflectivity and the nominal optical image distribution corresponding to the current mask pattern under the condition that the nominal photoresist thickness and the incident light intensity distribution are corresponding.

[0181] The reflectivity adjustment unit is configured to calculate the optical image distribution according to the nominal optical image distribution, the nominal reflectivity and the reflectivity distribution.

[0182] As a preferred embodiment, the optical image module 500 comprises:

[0183] The second formula unit is configured to calculate the nominal optical image distribution by the following formula:

[0184] I f 0 (x,y) =I i(x,y) ×(1-r eff 0 (x,y) ); (10)

[0185] wherein I f 0 (x,y) is the nominal optical image distribution; r eff 0 (x,y) is the nominal reflectivity; and I i(x,y) is the incident light intensity.

[0186] As a preferred embodiment, the optical image module 500 comprises:

[0187] A third formula unit is configured to calculate the optical image distribution according to the following formula:

[0188] I f(x,y) =I f 0 (x,y) ×[(1-r eff(x,y) ) / (1-r eff 0 (x,y) )]; (11)

[0189] wherein I f(x,y) is the optical image distribution; I f 0 (x,y) is the nominal optical image distribution; r eff(x,y) is the reflectivity distribution; and r eff 0 (x,y) is the nominal reflectivity.

[0190] As a preferred embodiment, the photoresist module 600 comprises:

[0191] A fourth formula unit is configured to obtain the photoresist image according to the following formula:

[0192] ; (2)

[0193] wherein R(x, y) is the photoresist image; F j is the jth photoresist model term based on the photoresist pattern correction; c j is the linear coefficient corresponding to F j ; and I f(x,y) is the optical image distribution.

[0194] As a preferred embodiment, the correction module 800 further comprises:

[0195] A judging unit is configured to judge whether the edge placement error of the photoresist image corresponding to the target wafer pattern exceeds a preset error tolerance threshold.

[0196] A conditional updating unit is configured to, when the edge placement error exceeds the preset error tolerance threshold, adjust the edge segment of the current mask pattern according to the edge placement error to obtain an updated current mask pattern, and calculate the edge placement error corresponding to the updated current mask pattern until the edge placement error does not exceed the error tolerance threshold.

[0197] As a preferred embodiment, the training device of the photoresist thickness distribution model comprises:

[0198] a model establishing module, configured to establish an uncorrected photoresist thickness model; the photoresist thickness model is capable of performing a polygon expansion operation on an input wafer surface topography pattern to obtain a polygon expansion pattern, calculating a preliminary upper surface height of the polygon expansion pattern respectively through a preset thickness increase coefficient, performing a convolution processing on the preliminary upper surface height through a preset Gaussian smoothing coefficient to obtain a smoothed upper surface height, and determining a photoresist thickness distribution through the smoothed upper surface height;

[0199] a sample module, configured to make a plurality of photoresist laying samples according to a plurality of preset sample surface topography patterns, and obtain sample photoresist thickness distributions corresponding to the photoresist laying samples;

[0200] a calibration module, configured to calibrate the thickness increase coefficient, the Gaussian smoothing coefficient and the polygon expansion operation by using the sample surface topography patterns and the corresponding sample photoresist thickness distributions, so that a difference between a simulated photoresist thickness distribution obtained by inputting the sample surface topography patterns and the corresponding sample photoresist thickness distribution is less than a preset calibration tolerance threshold.

[0201] As a preferred implementation, the calibration module comprises:

[0202] a loss function unit, configured to consider that a difference between a simulated photoresist thickness distribution obtained by inputting the sample surface topography patterns and the corresponding sample photoresist thickness distribution is less than a preset calibration tolerance threshold when a loss function value of the following formula is less than the preset calibration tolerance threshold:

[0203] ; (9)

[0204] wherein, cost is the loss function value; W is a convex pattern size in the sample surface topography pattern; S is a convex pattern pitch in the sample surface topography pattern; is a difference between a photoresist thickness in the simulated photoresist thickness distribution and a photoresist thickness in the sample photoresist thickness distribution at a corresponding position of a convex pattern in the sample surface topography pattern under a corresponding convex pattern size and convex pattern pitch; is a difference between a photoresist thickness in the simulated photoresist thickness distribution and a photoresist thickness in the sample photoresist thickness distribution at a corresponding position between two adjacent expanded convex patterns in the polygon expansion pattern corresponding to the sample surface topography pattern under a corresponding convex pattern size and convex pattern pitch.

[0205] The optical proximity correction device provided by the application comprises a receiving module 100, which is used for receiving a wafer surface topography pattern, a target wafer pattern and illumination light intensity; an analog module 200, which is used for inputting the wafer surface topography pattern and a preset nominal photoresist thickness into a pre-trained photoresist thickness distribution model to obtain a corresponding photoresist thickness distribution; a reflectivity module 300, which is used for calculating a reflectivity distribution according to the photoresist thickness distribution; an incident light intensity module 400, which is used for calculating an incident light intensity distribution by using the illumination light intensity and a current mask pattern; the current mask pattern is the same as the target wafer pattern; the current mask pattern comprises a plurality of edge segments; an optical image module 500, which is used for calculating an optical image distribution according to the reflectivity distribution and the incident light intensity distribution; a photoresist module 600, which is used for performing photoresist pattern correction on the optical image distribution to obtain a photoresist image; an EPE module 700, which is used for calculating edge placement errors of each edge segment of the target wafer pattern corresponding to the photoresist image; and a correction module 800, which is used for correcting the edge segments of the current mask pattern according to the edge placement errors. In the application, after the incident light intensity distribution of the illumination light intensity penetrating through the current mask pattern is calculated, the incident light intensity distribution is not directly used to calculate the photoresist image, but the thickness of the photoresist at each position (i.e., the photoresist thickness distribution) is further calculated by using the photoresist thickness distribution model, and then the reflectivity at each position (i.e., the reflectivity distribution) is calculated by using the thickness, and then the light absorption of the photoresist at each position is calculated according to the reflectivity distribution and the incident light intensity distribution, so that it can be more accurately judged which positions of the photoresist have photochemical reactions, thereby greatly improving the accuracy and precision of the obtained photoresist image, and further improving the calculation accuracy of the edge placement errors, i.e., improving the accuracy and precision of the optical proximity correction.

[0206] The optical proximity correction device of the embodiment is used for implementing the optical proximity correction method described above, and the specific embodiments of the optical proximity correction device can be seen from the embodiment part of the optical proximity correction method in the foregoing description, for example, the receiving module 100, the analog module 200, the reflectivity module 300, the incident light intensity module 400, the optical image module 500, the photoresist module 600, the EPE module 700 and the correction module 800 are respectively used for implementing steps S101, S102, S103, S104, S105, S106, S107 and S108 in the optical proximity correction method described above, and the specific embodiments can be referred to the description of the corresponding embodiment part, which will not be described herein.

[0207] The application further provides an optical proximity correction device, comprising:

[0208] a memory, used for storing a computer program;

[0209] The processor is used for executing the computer program to realize the steps of the optical proximity correction method. The optical proximity correction method provided by the application comprises the following steps: receiving a wafer surface topography pattern, a target wafer pattern and illumination light intensity; inputting the wafer surface topography pattern and a preset nominal photoresist thickness into a pre-trained photoresist thickness distribution model to obtain a corresponding photoresist thickness distribution; calculating a reflectivity distribution according to the photoresist thickness distribution; calculating an incident light intensity distribution according to the illumination light intensity and a current mask pattern; the initial current mask pattern is the same as the target wafer pattern; the current mask pattern comprises a plurality of edge segments; calculating an optical image distribution according to the reflectivity distribution and the incident light intensity distribution; performing photoresist pattern correction on the optical image distribution to obtain a photoresist image; calculating an edge placement error of the photoresist image corresponding to the target wafer pattern; and correcting the edge segments of the current mask pattern according to the edge placement error. In the application, after the incident light intensity distribution of the illumination light intensity penetrating the current mask pattern is calculated, the incident light intensity distribution is not directly used to calculate the photoresist image, but the thickness of the photoresist at each position (i.e., the photoresist thickness distribution) is further calculated through the photoresist thickness distribution model, and then the reflectivity at each position (i.e., the reflectivity distribution) is calculated through the thickness, and then the light absorption of the photoresist at each position is calculated according to the reflectivity distribution and the incident light intensity distribution, so that it can be more accurately judged which positions of the photoresist have photochemical reaction, thereby greatly improving the accuracy and precision of the finally obtained photoresist image, and further improving the calculation accuracy of the edge placement error, that is, improving the accuracy and precision of the optical proximity correction.

[0210] The application further provides a computer readable storage medium, wherein the computer readable storage medium stores a computer program, and the computer program is executed by a processor to implement the steps of the optical proximity correction method according to any one of the above embodiments. The optical proximity correction method provided by the application comprises the following steps: receiving a wafer surface topography pattern, a target wafer pattern and illumination light intensity; inputting the wafer surface topography pattern and a preset nominal photoresist thickness into a pre-trained photoresist thickness distribution model to obtain a corresponding photoresist thickness distribution; calculating a reflectivity distribution according to the photoresist thickness distribution; calculating an incident light intensity distribution according to the illumination light intensity and a current mask pattern; the initial current mask pattern is the same as the target wafer pattern; the current mask pattern comprises a plurality of edge segments; calculating an optical image distribution according to the reflectivity distribution and the incident light intensity distribution; performing photoresist pattern correction on the optical image distribution to obtain a photoresist image; calculating an edge placement error of the photoresist image corresponding to the target wafer pattern; and correcting the edge segments of the current mask pattern according to the edge placement error. In the application, after the incident light intensity distribution of the illumination light intensity penetrating the current mask pattern is calculated, the incident light intensity distribution is not directly used to calculate the photoresist image, but the thickness of the photoresist at each position (i.e., the photoresist thickness distribution) is further calculated through the photoresist thickness distribution model, and then the reflectivity at each position (i.e., the reflectivity distribution) is calculated through the thickness, and then the light absorption of the photoresist at each position is calculated according to the reflectivity distribution and the incident light intensity distribution, so that the positions of the photoresist which have undergone photochemical reaction can be more accurately judged, thereby greatly improving the accuracy and precision of the final photoresist image, and further improving the calculation accuracy of the edge placement error, i.e., improving the accuracy and precision of the optical proximity correction.

[0211] The embodiments in the specification are described in a progressive manner, and each embodiment focuses on the difference from other embodiments. The same or similar parts of each embodiment can be referred to each other. For the device disclosed by the embodiments, the description is relatively simple because it corresponds to the method disclosed by the embodiments. The relevant parts can be referred to the description of the method.

[0212] It should be noted that, in the specification, the terms such as first and second, etc. are merely intended to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "contain" or any other variants thereof are intended to cover non-exclusive inclusion, so that the process, method, article or device including a series of elements not only includes those elements, but also includes other elements not explicitly listed or inherent to such process, method, article or device. Without more limitations, the element defined by the statement "including a" does not exclude the presence of additional identical elements in the process, method, article or device including the element.

[0213] The skilled person will further appreciate that the elements of an example described in connection with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or combinations of both. The disclosures have been made in the context of the examples described in general terms to clearly illustrate the interchangeability of hardware and software. The decision to implement a function in either hardware or software depends on the specific application and design constraints imposed on the technical solution. The skilled person can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present invention.

[0214] The steps of the methods or algorithms described in connection with the embodiments disclosed herein can be implemented directly in hardware, by a software module executed by a processor, or a combination of both. The software module can be stored in a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

[0215] The optical proximity correction method and device provided by the present application are described in detail above. The principles and implementation modes of the present application are described by applying specific examples in this paper, and the above description of the examples is only used to help understand the method of the present application and its core idea. It should be pointed out that, for ordinary skilled persons in the technical field, some improvements and modifications can be made to the present application without departing from the principles of the present application, and these improvements and modifications also fall within the protection scope of the present application.

Claims

1. An optical proximity correction method, characterized in that, include: Receive the surface topography pattern of the wafer, the target wafer pattern, and the intensity of the irradiated light; The wafer surface topography pattern and the preset nominal photoresist thickness are input into a pre-trained photoresist thickness distribution model to obtain the corresponding photoresist thickness distribution. Calculate the reflectivity distribution based on the photoresist thickness distribution; The incident light intensity distribution is calculated using the illumination light intensity and the current mask pattern; the initial current mask pattern is the same as the target wafer pattern; the current mask pattern includes multiple edge segments; The optical image distribution is calculated based on the reflectivity distribution and the incident light intensity distribution. The optical image distribution is corrected using a photoresist pattern to obtain a photoresist image; Calculate the edge placement error of the photoresist image corresponding to the target wafer pattern; Based on the edge placement error, the edge segments of the current mask pattern are corrected.

2. The optical proximity correction method as described in claim 1, characterized in that, Calculating the optical image distribution based on the reflectivity distribution and the incident light intensity distribution includes: The optical image distribution is calculated using the following formula: I f(x,y) =I i(x,y) ×(1-r eff(x,y) ); Among them, I f(x,y) The optical image distribution; r eff(x,y) The reflectivity distribution is I. i(x,y) The incident light intensity is denoted as .

3. The optical proximity correction method as described in claim 1, characterized in that, Calculating the optical image distribution based on the reflectivity distribution and the incident light intensity distribution includes: Under the conditions corresponding to the nominal photoresist thickness and the incident light intensity distribution, calculate the nominal reflectivity and nominal optical image distribution of the current mask pattern; The optical image distribution is calculated based on the nominal optical image distribution, the nominal reflectance, and the reflectance distribution.

4. The optical proximity correction method as described in claim 3, characterized in that, The nominal optical image distribution corresponding to the current mask pattern, calculated under the conditions of the nominal photoresist thickness and the incident light intensity distribution, includes: The nominal optical image distribution is calculated using the following formula: I f 0 (x,y) =I i(x,y) ×(1-r eff 0 (x,y) ); Among them, I f 0 (x,y) The nominal optical image distribution; r eff 0 (x,y) I is the nominal reflectivity; i(x,y) The incident light intensity is denoted as .

5. The optical proximity correction method as described in claim 3, characterized in that, The optical image distribution is calculated based on the nominal optical image distribution, the nominal reflectance, and the reflectance distribution, including: The optical image distribution is calculated using the following formula: I f(x,y) =I f 0 (x,y) ×[(1-r eff(x,y) ) / (1-r eff 0 (x,y) )]; Among them, I f(x,y) The optical image distribution; I f 0 (x,y) The nominal optical image distribution; r eff(x,y) The reflectivity distribution is r. eff 0 (x,y) The nominal reflectance is given.

6. The optical proximity correction method as described in claim 1, characterized in that, The optical image distribution is modified by photoresist patterning to obtain a photoresist image, including: The photoresist image is obtained using the following formula: ； Where R(x,y) is the photoresist image; F j It is the j-th photoresist model term based on the photoresist pattern correction; c j For F j The corresponding linear coefficients; I f(x,y) The optical image distribution is described above.

7. The optical proximity correction method as described in claim 1, characterized in that, Before correcting the edge segments of the current mask pattern based on the edge placement error, the method further includes: Determine whether the edge placement error of the photoresist image corresponding to the target wafer pattern exceeds a preset error tolerance threshold; Accordingly, the step of correcting the edge segments of the current mask pattern based on the edge placement error includes: When the edge placement error exceeds a preset error tolerance threshold, the edge segments of the current mask pattern are adjusted according to the edge placement error to obtain an updated current mask pattern, and the edge placement error corresponding to the updated current mask pattern is calculated until the edge placement error does not exceed the error tolerance threshold.

8. The optical proximity correction method as described in claim 1, characterized in that, The training method for the photoresist thickness distribution model includes: An uncorrected photoresist thickness model is established. The photoresist thickness model first performs a polygon expansion operation on the input wafer surface topography pattern to obtain a polygon expansion pattern. Then, it calculates the initial upper surface height of the polygon expansion pattern using a preset thickness increase coefficient. Finally, it performs convolution processing on the initial upper surface height using a preset Gaussian smoothing coefficient to obtain a smooth upper surface height. The photoresist thickness distribution is then determined using the smooth upper surface height. Multiple photoresist layup samples are prepared according to a variety of preset sample surface morphology patterns, and the photoresist thickness distribution of the sample corresponding to the photoresist layup sample is obtained. Using the sample surface topography pattern and the corresponding sample photoresist thickness distribution, the thickness increase coefficient, the Gaussian smoothing coefficient, and the polygon expansion operation are calibrated so that the difference between the simulated photoresist thickness distribution obtained by inputting the sample surface topography pattern and the corresponding sample photoresist thickness distribution is less than a preset calibration tolerance threshold.

9. The optical proximity correction method as described in claim 8, characterized in that, When the loss function value of the current expression is less than a preset calibration tolerance threshold, it is considered that the difference between the simulated photoresist thickness distribution obtained by inputting the sample surface morphology pattern and the corresponding sample photoresist thickness distribution is less than the preset calibration tolerance threshold. ； Where cost is the loss function value; W is the size of the protruding pattern in the sample surface topography pattern; and S is the spacing between the protruding patterns in the sample surface topography pattern. For the corresponding protruding pattern size and protruding pattern spacing, the difference between the photoresist thickness in the simulated photoresist thickness distribution and the photoresist thickness in the sample photoresist thickness distribution at the corresponding position of the protruding pattern in the sample surface topography pattern; For the corresponding protruding pattern size and protruding pattern spacing, in the polygonal extended pattern corresponding to the sample surface topography pattern, at the corresponding position between two adjacent extended protruding patterns, the difference between the photoresist thickness in the simulated photoresist thickness distribution and the photoresist thickness in the sample photoresist thickness distribution.

10. An optical proximity correction device, characterized in that, include: The receiving module is used to receive the wafer surface topography pattern, the target wafer pattern, and the intensity of the irradiated light. The simulation module is used to input the wafer surface topography pattern and the preset nominal photoresist thickness into a pre-trained photoresist thickness distribution model to obtain the corresponding photoresist thickness distribution. The reflectivity module is used to calculate the reflectivity distribution based on the photoresist thickness distribution; An incident light intensity module is used to calculate the incident light intensity distribution based on the illumination light intensity and the current mask pattern; the initial current mask pattern is the same as the target wafer pattern; the current mask pattern includes multiple edge segments; An optical image module is used to calculate an optical image distribution based on the reflectivity distribution and the incident light intensity distribution; A photoresist module is used to perform photoresist pattern correction on the optical image distribution to obtain a photoresist image; The EPE module is used to calculate the edge placement error of each edge segment on the target wafer pattern corresponding to the photoresist image; The correction module is used to correct the edge segments of the current mask pattern based on the edge placement error.

Images