Digital mask photoresist pattern matching method and system based on optical proximity correction

Abstract

The invention discloses a digital mask photoresist pattern matching method and system based on optical proximity correction. The method comprises the following steps: determining a modulation coefficient of a digital mask pattern; projecting the digital mask pattern based on a Hopkinson diffraction model and a photoresist model to obtain a photoresist pattern; performing optimization processing on the modulation coefficient through a steepest descent method to obtain a modulation coefficient after preliminary optimization; and calculating an error value between the photoresist pattern and a given target pattern through an image error method, and performing iterative optimization processing on the preliminarily optimized modulation coefficient according to the error value to obtain an optimized digital mask pattern. According to the invention, the matching rate of the exposure pattern and the target pattern can be improved by optimizing the modulation coefficient of the digital mask. The digital mask photoresist pattern matching method and system based on optical proximity correction can be widely applied to the technical field of photoresist image matching.

Description

Technical Field

[0001] The invention relates to the technical field of photoresist image matching, and in particular to a digital mask photoresist pattern matching method and system based on optical proximity correction. Background Art

[0002] Various devices with beam patterning functions have been widely used in micro-nano processing, such as diffractive optical elements, phase plates, spatial light modulators, and digital micromirror devices. Among them, laser modulation shaping based on spatial light modulators (SLMs) and digital micromirror devices (DMDs) is the most mainstream light modulation method. Computer-generated holograms (CGHs) are commonly used phase diagrams for wavefront modulation of light through SLMs, usually calculated by phase recovery algorithms. However, the speckle noise caused by the unconstrained phase of CGHs severely limits the quality of the produced graphics, making it challenging to produce submicron-scale features. On the other hand, SLMs can change the polarization state distribution of the light field by controlling the orientation direction of the liquid crystal molecules in each pixel. Therefore, using a simple binary phase diagram combined with a polarizer to modulate the wavefront intensity can achieve the effect of a physical mask and produce a uniform patterned light field with a consistent phase distribution. Compared with the fixed pattern of the physical mask, the production cost is high. The patterned laser lithography (PLL) technology of the digital mask based on SLM is flexible and programmable, capable of quickly switching the mask pattern with high precision. It provides considerable throughput and flexibility for micro-nano processing, such as slicing-based 3D additive manufacturing or femtosecond laser ablation of metal films to process metasurfaces.

[0003] Although PLL has advantages such as high efficiency and low cost to produce complex patterns, its disadvantages are inevitable. Previous work has verified that physical masks have obvious optical proximity effects (OPE) when manufacturing sub-micron scale patterns. This effect is also obvious for digital mask lithography based on SLM or DMD. For projection lithography based on physical masks, undesirable effects can be solved by optical proximity correction (OPC), such as adding sub-resolution auxiliary features (SRAF). However, for OPE in projection lithography of digital masks, the methods reported so far only focus on optimizing the grayscale values ​​of pixels within the target pattern, or only consider adding auxiliary feature patterns. These limited optimization strategies lead to significant deviations. Summary of the invention

[0004] In order to solve the above technical problems, the purpose of the present invention is to provide a digital mask photoresist pattern matching method and system based on optical proximity correction, which can improve the matching rate between the exposure pattern and the target pattern by optimizing the modulation coefficient of the digital mask.

[0005] The first technical solution adopted by the present invention is: a digital mask photoresist pattern matching method based on optical proximity correction, comprising the following steps:

[0006] determining a modulation coefficient of a digital mask pattern;

[0007] Projecting the digital mask pattern based on the Hopkins diffraction model and the photoresist model to obtain a photoresist pattern;

[0008] The modulation coefficient is optimized by the steepest descent method to obtain the initially optimized modulation coefficient;

[0009] The error value between the photoresist pattern and the given target pattern is calculated by the image error method, and the modulation coefficient after preliminary optimization is iteratively optimized according to the error value to obtain the optimized digital mask pattern.

[0010] Further, the step of determining the modulation coefficient of the digital mask pattern specifically includes:

[0011] Acquire a digital mask pattern, and determine the complex amplitude transmittance of the digital mask pattern;

[0012] According to the complex amplitude transmittance of the digital mask pattern, the modulation coefficient of the digital mask pattern is obtained by performing unconstrained optimization through trigonometric functions.

[0013] Furthermore, the expression of the complex amplitude transmittance of the digital mask pattern is specifically as follows:

[0014]

[0015] In the above formula, m a，b represents the modulation coefficient loaded on the pixel at row a and column b of the spatial light modulator, rect(·) represents the rectangular function, M(x, y) represents the complex amplitude transmittance function of a single pixel of the spatial light modulator, T x With T y represent the arrangement period of the spatial light modulator in the x direction and the y direction, respectively, and W x Represents the length of a single pixel of the spatial light modulator, W y Represents the width of a single pixel of the spatial light modulator.

[0016] Furthermore, the expression of the modulation coefficient of the digital mask pattern is specifically as follows:

[0017]

[0018] In the above formula, m a，b represents the modulation coefficient of the digital mask pattern, cosθ a，b Represents a transformation relationship.

[0019] Furthermore, the step of projecting the digital mask pattern based on the Hopkins diffraction model and the photoresist model to obtain the photoresist pattern specifically includes:

[0020] According to the complex amplitude transmittance of the digital mask pattern and the Hopkins diffraction model, the light intensity distribution on the photoresist is obtained;

[0021] By using the sigmoid function to approximate the hard threshold function, a photoresist model with a differentiable cost function is constructed;

[0022] Combined with the light intensity distribution on the photoresist, the digital mask pattern is projected through a photoresist model with a differentiable cost function to obtain a photoresist pattern.

[0023] Further, the expression of the photoresist pattern is specifically as follows:

[0024]

[0025] In the above formula, Z represents the photoresist pattern, sigmoid(·) represents the sigmoid function, I represents the light intensity distribution on the photoresist, and a r Represents the steepness of the sigmoid function, t r Represents the threshold of photoresist.

[0026] Furthermore, the step of optimizing the modulation coefficient by the steepest descent method to obtain a preliminary optimized modulation coefficient specifically includes:

[0027] determining a cost function based on a norm of a difference between the photoresist pattern and a given target pattern;

[0028] Set the optimization step size and calculate the gradient of the cost function;

[0029] The modulation coefficient is optimized according to the gradient of the cost function to obtain a preliminary optimized modulation coefficient.

[0030] Furthermore, the expression of the cost function is specifically as follows:

[0031]

[0032] In the above formula, F(·) represents the cost function, Z * represents a given target pattern, Z represents the photoresist pattern, θ represents the latent variable, and a r Represents the steepness of the sigmoid function, t r represents the threshold of the photoresist, N represents the Nth pixel, i represents the row, j represents the column, and h represents the i，j represents the point spread function of the imaging system.

[0033] Further, the step of calculating the error value between the photoresist pattern and the given target pattern by the image error method, and iteratively optimizing the modulation coefficient after preliminary optimization according to the error value to obtain the optimized digital mask pattern specifically includes:

[0034] Calculating the error value between the photoresist pattern and a given target pattern by an image error method;

[0035] comparing the error value with a preset threshold;

[0036] If the error value is greater than or equal to the preset threshold, the modulation coefficient is optimized by the steepest descent method in a loop to obtain the initially optimized modulation coefficient, until the error value is less than the preset threshold, and the optimized modulation coefficient is output;

[0037] The optimized modulation coefficient is mapped to the spatial light modulator, and the digital mask pattern is projected to obtain the optimized digital mask pattern.

[0038] The second technical solution adopted by the present invention is: a digital mask photoresist pattern matching system based on optical proximity correction, comprising:

[0039] A first module is used to determine a modulation coefficient of a digital mask pattern;

[0040] The second module is used to project the digital mask pattern based on the Hopkins diffraction model and the photoresist model to obtain the photoresist pattern;

[0041] The third module is used to optimize the modulation coefficient by the steepest descent method to obtain the modulation coefficient after preliminary optimization;

[0042] The fourth module is used to calculate the error value between the photoresist pattern and the given target pattern by using the image error method, and iteratively optimize the modulation coefficient after preliminary optimization according to the error value to obtain the optimized digital mask pattern.

[0043] The beneficial effects of the method and system of the present invention are as follows: the present invention determines the modulation coefficient of the digital mask pattern; further projects the digital mask pattern based on the Hopkins diffraction model and the photoresist model to obtain the photoresist pattern, and optimizes the modulation coefficient by the steepest descent method, optimizes the modulation coefficient of the digital mask, makes the photoresist pattern match the target pattern, and then uses the inverse pattern solving algorithm, i.e., the steepest descent method, to find a digital mask that closely matches the target layout, thereby improving the matching rate between the exposure pattern and the target pattern. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 It is a flow chart of the steps of the digital mask photoresist pattern matching method based on optical proximity correction of the present invention;

[0045] Figure 2 It is a structural block diagram of a digital mask photoresist pattern matching system based on optical proximity correction of the present invention;

[0046] Figure 3 is a schematic diagram of a simplified digital mask projection lithography system and OPC effect provided by a specific embodiment of the present invention;

[0047] Figure 4 It is a schematic diagram of the steps of optimizing the digital mask photoresist pattern provided by a specific embodiment of the present invention;

[0048] Figure 5 is a schematic diagram of error comparison before and after optimization of a photoresist pattern provided by a specific embodiment of the present invention;

[0049] Figure 6 It is a schematic diagram comparing the accuracy before and after the optimization of the photoresist pattern provided by a specific embodiment of the present invention. DETAILED DESCRIPTION

[0050] The present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. The step numbers in the following embodiments are only provided for the convenience of explanation and description, and the order between the steps is not limited in any way. The execution order of each step in the embodiment can be adaptively adjusted according to the understanding of those skilled in the art.

[0051] First of all, it should be noted that Figure 3 As shown, a simplified digital mask projection lithography system is shown, as well as a comparison of the effects of optimized and non-optimized digital masks. Due to the optical proximity effect (OPE), the patterned light field loaded based on the target pattern will cause the printed pattern to be deformed. Optical proximity correction (OPC) is an inverse lithography technology (ILT) widely used to solve OPE by pre-distorting the mask pattern to compensate for imaging distortion. In this case, the output and imaging system are known, but the input is unknown. Unlike the binary modulation of the light field by the physical mask, the mask function of the digital mask is realized by the modulation coefficient loaded onto the SLM. ILT is committed to finding the optimally distributed modulation coefficient m to ensure that the photoresist pattern is similar to the target layout.

[0052] Reference Figure 1 and Figure 4 The present invention provides a digital mask photoresist pattern matching method based on optical proximity correction, the method comprising the following steps:

[0053] S100, determining a modulation coefficient of a digital mask pattern;

[0054] S110, acquiring a digital mask pattern, and determining a complex amplitude transmittance of the digital mask pattern;

[0055] In this embodiment, in the DMPL digital mask projection lithography system proposed in this paper, the SLM (spatial light modulator) realizes the digital mask by controlling the orientation direction of the liquid crystal molecules in each pixel to modulate the grayscale amplitude, and the minimum modulation range is one pixel. For the complex amplitude transmittance of an N×N pixel digital mask to be optimized, its expression is:

[0056]

[0057] In the above formula, m a，b represents the modulation coefficient loaded on the pixel at row a and column b of the spatial light modulator, and the initial value is 0 or 1, which respectively represent extinction and light transmission, rect(·) represents the rectangular function, represents the complex amplitude transmittance function of a single pixel of the SLM, and for the convenience of discussion, the embodiment of the present invention sets its value to 1, M(x, y) represents the complex amplitude transmittance function of a single pixel of the spatial light modulator, T x With T y represent the arrangement period of the spatial light modulator in the x direction and the y direction, respectively, and W x Represents the length of a single pixel of the spatial light modulator, W y Represents the width of a single pixel of the spatial light modulator.

[0058] S120 , performing unconstrained optimization through trigonometric functions according to the complex amplitude transmittance of the digital mask pattern to obtain a modulation coefficient of the digital mask pattern.

[0059] In this embodiment, the value of the modulation coefficient m is limited to the range of 0 to 1. Due to the difficulty of constrained optimization, the problem is transformed into an unconstrained optimization problem using trigonometric functions. The modulation coefficient m can be described as:

[0060]

[0061] In the above formula, m a，b represents the modulation coefficient of the digital mask pattern, cosθ a，b Represents a transformation relationship.

[0062] It should be further explained that θ a，b ∈(-∞, +∞), so the modulation coefficient m is indirectly optimized by optimizing the value of the latent variable θ. * ,The purpose of inverse lithography technology ILT is to find the modulation coefficient m loaded on the SLM so that the distance between the ,resist pattern and the target layout is minimized.

[0063] S200, projecting the digital mask pattern based on the Hopkins diffraction model and the photoresist model to obtain a photoresist pattern;

[0064] S210, obtaining the light intensity distribution on the photoresist according to the complex amplitude transmittance of the digital mask pattern in combination with the Hopkins diffraction model;

[0065] In this embodiment, for the coherent imaging system, considering the Hopkins diffraction model, the light intensity distribution on the photoresist is expressed as:

[0066] I=|M*h| 2

[0067] In the above formula, I represents the light intensity distribution on the photoresist, h represents the imaging system point spread function (PSF), a Gaussian low-pass filter is used for approximation in the implementation of the present invention, * represents a convolution operation, and M represents a digital mask.

[0068] S220, constructing a photoresist model with a differentiable cost function by using a sigmoid function to approximate a hard threshold function;

[0069] S230 , combining the light intensity distribution on the photoresist, and projecting the digital mask pattern through a photoresist model with a differentiable cost function to obtain a photoresist pattern.

[0070] In this embodiment, the photoresist model used in the embodiment of the present invention adopts a constant threshold model (CTR). However, the expression of the hard threshold model is not differentiable, which poses a challenge to optimization algorithms such as gradient descent. In order to solve this problem, the cost function is usually made differentiable by using a sigmoid function to approximate the hard threshold function. Therefore, the output pattern on the photoresist is:

[0071]

[0072] In the above formula, Z represents the photoresist pattern, sigmoid(·) represents the sigmoid function, I represents the light intensity distribution on the photoresist, and a r Indicates the steepness of the sigmoid function, which is set to 80 in the embodiment of the present invention. r Represents the threshold of the photoresist, and the embodiment of the present invention sets its value to 0.5.

[0073] S300, optimizing the modulation coefficient by the steepest descent method to obtain a preliminarily optimized modulation coefficient;

[0074] S310, determining a cost function according to a norm of a difference between the photoresist pattern and a given target pattern;

[0075] In this embodiment, the cost function is defined as the norm of the difference between the output pattern Z and the desired pattern Z*, which is expressed as:

[0076]

[0077] In the above formula, F(·) represents the cost function, Z* represents the given target pattern, Z represents the photoresist pattern, θ represents the latent variable, and a r Represents the steepness of the sigmoid function, t r represents the threshold of the photoresist, N represents the Nth pixel, i represents the row, j represents the column, and h represents the i，j represents the point spread function of the imaging system.

[0078] S320, setting the optimization step size and calculating the gradient of the cost function;

[0079] S330 , optimizing the modulation coefficient according to the gradient of the cost function to obtain a preliminarily optimized modulation coefficient.

[0080] In this embodiment, the gradient of the cost function The calculation is as follows:

[0081]

[0082] At the k+1th iteration, its expression is:

[0083]

[0084] In the above formula, s represents the step length.

[0085] S400, calculating the error value between the photoresist pattern and the given target pattern by an image error method, and performing iterative optimization processing on the modulation coefficient after preliminary optimization according to the error value to obtain an optimized digital mask pattern.

[0086] Specifically, the error value between the photoresist pattern and the given target pattern is calculated by the image error method; the error value is compared with a preset threshold; if the error value is greater than or equal to the preset threshold, the modulation coefficient is optimized through the steepest descent method in a loop to obtain a preliminary optimized modulation coefficient, until the error value is less than the preset threshold, and the optimized modulation coefficient is output; the optimized modulation coefficient is mapped to the spatial light modulator, and the digital mask pattern is projected to obtain an optimized digital mask pattern.

[0087] In this embodiment, in order to facilitate the evaluation of the optimization effect of ILT, the embodiment of the present invention proposes to use the image error (PE) as a standard for evaluating the image quality of the lithography system. PE is defined as the total number of pixels that are not faithfully reproduced in the binary output mode, and its expression is:

[0088]

[0089] The smaller the PE value, the closer the printed pattern is to the target pattern. When the PE value is less than the set value, the algorithm stops iterating and outputs the optimized digital mask. It is worth noting that the modulation coefficient m of the output digital mask is between 0 and 1, and can only be loaded onto the SLM after being mapped to a grayscale value of 127-0 by the algorithm. Figure 5 (a) and Figure 5 (c) shows the digital mask actually loaded onto the SLM. Figure 5 (b) and Figure 5 (d) shows the photoresist pattern without optimized mask and the photoresist pattern with optimized mask, respectively. Figure 5 The optimization results in (d) show that the OPE problem is well solved. Figure 5 (e) shows the convergence process of PE of three printed images as the optimization algorithm is iterated, thus demonstrating the robustness and versatility of the algorithm.

[0090] Furthermore, in order to verify the effectiveness of the above optimization algorithm, the embodiment of the present invention conducted a digital OPC exposure experiment, such as Figure 6 As shown. When the digital mask size is 120 pixels, it is scaled to 1um by the photolithography system, resulting in a significant optical proximity effect. The line width of the selected pattern is between 200 and 900nm. Figure 6 (a) and Figure 6 Figures 1, 2, and 3 shown in (b) demonstrate the sensitivity of smaller line widths to corner rounding and proximity effects. We note that since the pattern of Figure 3 is denser, more diffracted light is superimposed, resulting in more obvious corner rounding and structural connection. After mask optimization, the theoretical ( Figure 6 (d)) and experiments ( Figure 6 (c)) The results all show that the line width closely matches the target pattern, which reduces corner rounding and structural connection problems to a certain extent. Among them, Figure 3 is the minimum period grating that we tested and can be processed, with a period of 669nm, a structural width of 390nm, and a spacing of 279nm. According to the Rayleigh resolution limit, the minimum structural period that can be reproduced on the photoresist is:

[0091]

[0092] Where: λ is the wavelength of the light source, NA is the numerical aperture on the wafer side, and k is the process constant that can be optimized by OPC technology. It can be obtained that the process constant k of this system after optimization is about 1.94. The above results verify that OPC based on the steepest descent method effectively reduces the OPE in SLM-DMPL.

[0093] In order to quantitatively analyze the optimization effect, the embodiment of the present invention uses image subtraction technology to calculate the matching rate (MR) between the exposure result and the target pattern. The matching rate is defined as:

[0094]

[0095] Where Z is the binary image of the exposure pattern and (i, j) are the coordinates of the pixels of the pattern. The MR results of Figures 1, 2 and 3 are shown in Figure 6 (e). The average MR of the optimized digital mask increased by about 16%. It is worth noting that for Figure 1 and Figure 2, which only have a single structure, the MR of the unoptimized mask reached 88.8% and 78.7%, respectively. The high MR values ​​show the consistency and accuracy of patterned light field processing for a single structure. Among all the graphics, the MR of Figure 3 increased most significantly, from 39.1% to 76.5%. The above graphics have different characteristics, and the optimization effect is better when the structure is denser. All results further confirm the effectiveness of OPC based on the steepest descent method for digital mask projection lithography.

[0096] In summary, the embodiment of the present invention establishes an optical proximity correction (OPC) based on the steepest descent method and is applied to the spatial light modulator (SLM) digital mask projection lithography technology, which can be used to optimize the modulation coefficient of the digital mask so that the photoresist pattern matches the target pattern. The technology uses a phase-type SLM to modulate the polarization state distribution of a continuous laser to obtain a patterned light field to achieve a digital mask. The embodiment of the present invention follows the inverse imaging mask design method to optimize the digital mask. First, a forward model of the distortion effect of the analog imaging system is established, including an optical model and a photoresist model. Then, the inverse pattern solving algorithm, i.e., the steepest descent method, is used to find a digital mask that closely matches the target layout. Finally, the optimized digital grayscale pattern is loaded into the SLM as a digital mask. The experimental results show that the exposure pattern closely matches the target pattern, and the matching rate between the exposure pattern and the target pattern is increased by up to 39.1%.

[0097] Reference Figure 2 , a digital mask resist pattern matching system based on optical proximity correction, comprising:

[0098] The first module 201 is used to determine the modulation coefficient of the digital mask pattern;

[0099] The second module 202 is used to project the digital mask pattern based on the Hopkins diffraction model and the photoresist model to obtain a photoresist pattern;

[0100] The third module 203 is used to optimize the modulation coefficient by the steepest descent method to obtain a preliminarily optimized modulation coefficient;

[0101] The fourth module 204 is used to calculate the error value between the photoresist pattern and the given target pattern by using the image error method, and iteratively optimize the modulation coefficient after preliminary optimization according to the error value to obtain the optimized digital mask pattern.

[0102] The contents of the above method embodiments are all applicable to the present system embodiments. The functions specifically implemented by the present system embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0103] The above is a specific description of the preferred implementation of the present invention, but the invention is not limited to the embodiments. Those skilled in the art may make various equivalent modifications or substitutions without violating the spirit of the present invention. These equivalent modifications or substitutions are all included in the scope defined by the claims of this application.

Claims

1. A digital mask photoresist pattern matching method based on optical proximity correction, characterized in that: The following steps are involved: determining a modulation coefficient of a digital mask pattern; Projecting the digital mask pattern based on the Hopkins diffraction model and the photoresist model to obtain a photoresist pattern; The modulation coefficient is optimized by the steepest descent method to obtain the initially optimized modulation coefficient; The error value between the photoresist pattern and the given target pattern is calculated by the image error method, and the modulation coefficient after preliminary optimization is iteratively optimized according to the error value to obtain the optimized digital mask pattern.

2. The digital mask photoresist pattern matching method based on optical proximity correction according to claim 1, characterized in that: The step of determining the modulation coefficient of the digital mask pattern specifically includes: Acquire a digital mask pattern, and determine the complex amplitude transmittance of the digital mask pattern; According to the complex amplitude transmittance of the digital mask pattern, the modulation coefficient of the digital mask pattern is obtained by performing unconstrained optimization through trigonometric functions.

3. The digital mask photoresist pattern matching method based on optical proximity correction according to claim 2, characterized in that: The expression of the complex amplitude transmittance of the digital mask pattern is specifically as follows: In the above formula, m a,b represents the modulation coefficient loaded on the pixel at row a and column b of the spatial light modulator, rect(·) represents the rectangular function, M(x,y) represents the complex amplitude transmittance function of a single pixel of the spatial light modulator, T x With T y represent the arrangement period of the spatial light modulator in the x direction and the y direction, respectively, and W x Represents the length of a single pixel of the spatial light modulator, W y Represents the width of a single pixel of the spatial light modulator.

4. The digital mask photoresist pattern matching method based on optical proximity correction according to claim 3, characterized in that: The expression of the modulation coefficient of the digital mask pattern is specifically as follows: In the above formula, m a,b represents the modulation coefficient of the digital mask pattern, cosθ a,b Represents a transformation relationship.

5. The digital mask photoresist pattern matching method based on optical proximity correction according to claim 4, characterized in that: The step of projecting the digital mask pattern based on the Hopkins diffraction model and the photoresist model to obtain the photoresist pattern specifically includes: According to the complex amplitude transmittance of the digital mask pattern and the Hopkins diffraction model, the light intensity distribution on the photoresist is obtained; By using the sigmoid function to approximate the hard threshold function, a photoresist model with a differentiable cost function is constructed; Combined with the light intensity distribution on the photoresist, the digital mask pattern is projected through a photoresist model with a differentiable cost function to obtain a photoresist pattern.

6. The digital mask photoresist pattern matching method based on optical proximity correction according to claim 5, characterized in that: The expression of the photoresist pattern is specifically as follows: In the above formula, Z represents the photoresist pattern, sigmoid(·) represents the sigmoid function, I represents the light intensity distribution on the photoresist, and a r Represents the steepness of the sigmoid function, t r Represents the threshold of photoresist.

7. The digital mask photoresist pattern matching method based on optical proximity correction according to claim 6, characterized in that: The step of optimizing the modulation coefficient by the steepest descent method to obtain a preliminary optimized modulation coefficient specifically includes: determining a cost function based on a norm of a difference between the photoresist pattern and a given target pattern; Set the optimization step size and calculate the gradient of the cost function; The modulation coefficient is optimized according to the gradient of the cost function to obtain a preliminary optimized modulation coefficient.

8. The digital mask photoresist pattern matching method based on optical proximity correction according to claim 7, characterized in that: The expression of the cost function is specifically as follows: In the above formula, F(·) represents the cost function, Z * represents a given target pattern, Z represents the photoresist pattern, θ represents the latent variable, and a r Represents the steepness of the sigmoid function, t r represents the threshold of the photoresist, N represents the Nth pixel, i represents the row, j represents the column, and h represents the i,j represents the point spread function of the imaging system.

9. The digital mask photoresist pattern matching method based on optical proximity correction according to claim 8, characterized in that: The step of calculating the error value between the photoresist pattern and the given target pattern by the image error method, and iteratively optimizing the modulation coefficient after preliminary optimization according to the error value to obtain the optimized digital mask pattern specifically includes: Calculating the error value between the photoresist pattern and a given target pattern by an image error method; comparing the error value with a preset threshold; If the error value is greater than or equal to the preset threshold, the modulation coefficient is optimized by the steepest descent method in a loop to obtain the initially optimized modulation coefficient, until the error value is less than the preset threshold, and the optimized modulation coefficient is output; The optimized modulation coefficient is mapped to the spatial light modulator, and the digital mask pattern is projected to obtain the optimized digital mask pattern.

10. A digital mask photoresist pattern matching system based on optical proximity correction, characterized in that: Includes the following modules: A first module is used to determine a modulation coefficient of a digital mask pattern; The second module is used to project the digital mask pattern based on the Hopkins diffraction model and the photoresist model to obtain the photoresist pattern; The third module is used to optimize the modulation coefficient by the steepest descent method to obtain the modulation coefficient after preliminary optimization; The fourth module is used to calculate the error value between the photoresist pattern and the given target pattern by using the image error method, and iteratively optimize the modulation coefficient after preliminary optimization according to the error value to obtain the optimized digital mask pattern.

Images