A method for improving the smoothness of DMD digital lithography based on curvature dynamic fuzzy exposure technology

Abstract

The present invention relates to the field of DMD device imaging exposure technology, and particularly to a method for improving the smoothness of DMD digital lithography based on curvature dynamic blur exposure technology, including: selecting an exposure image to be processed; segmenting the exposure image to be processed to obtain a plurality of sub-images; calculating the slope of the sub-images, and designing a displacement path based on the slope; calculating the local dilation of the sub-images based on the displacement path, and preprocessing the sub-images; exposing the preprocessed sub-images to obtain a complete exposure image. The present invention can effectively expose complex images with high fidelity.

Description

Technical Field

[0001] The present invention relates to the technical field of DMD device imaging exposure, and particularly to a method for improving the smoothness of DMD digital lithography based on curvature dynamic blur exposure technology. Background Art

[0002] As a core technology in the field of semiconductor processing, DMD digital lithography technology replaces traditional optical masks based on its characteristics such as low cost, high efficiency, and light weight, and is widely used in various fields such as industrial printed circuit board (PCB) manufacturing, tissue engineering, biomedicine, and 3D printing. For example, by combining DMD digital lithography technology with magnetic materials, micro medical robots can be fabricated, which can enter the human body through magnetic field control for microsurgery. The essence of DMD is composed of a square micro-mirror array, and this microscopic structure causes the edges of the exposed micro-structures in digital lithography to be serrated, greatly reducing the edge accuracy of the image and unable to meet the manufacturing requirements of high-precision micro-devices.

[0003] To solve this problem, the conventional method is to reduce the magnification of the projection objective lens to improve the optical resolution, but this greatly reduces the exposure area and the efficiency is extremely low. Gao Yiqing et al. reduced the DMD pixel error by designing the feature size of the mask as an integer multiple of the DMD pixel, but this method is limited by the design and processing of the micro-structure slice layer and is not very good in terms of processing flexibility. The spatio-temporal modulation technology proposed by Liu Hua et al. improves the smoothness of the image edge by the method of multiple sub-pixel overlaps, but a large number of sub-images need to be fabricated. The gray-scale exposure imaging technology improves the smoothness of the image edge by controlling the line width, but it requires a large amount of calculation, has high requirements for equipment, and high costs. It can be seen that the existing technologies for improving the smoothness of DMD digital lithography are difficult to meet the current requirements in high-precision systems. Therefore, the present invention proposes a method for improving the smoothness of DMD digital lithography based on curvature dynamic blur exposure technology. Summary of the Invention

[0004] The purpose of the present invention is to provide a method for improving the smoothness of DMD digital lithography based on curvature dynamic blur exposure technology to achieve high-quality optimization of the edges of the exposed patterns for the above technical problems.

[0005] To achieve the above purpose, the present invention provides the following solutions:

[0006] A method for improving the smoothness of DMD digital lithography based on curvature dynamic blur exposure technology, comprising:

[0007] Selecting the exposure image to be processed;

[0008] Segmenting the exposure image to be processed to obtain a plurality of sub-images;

[0009] Calculate the slope of the sub-image and design a displacement path based on the slope;

[0010] Calculate the local dilation of the sub-image based on the displacement path and preprocess the sub-image;

[0011] Expose the preprocessed sub-image to obtain a complete exposed image.

[0012] Optionally, selecting the to-be-processed exposed image includes:

[0013] Debug the DMD lithography system to the working state so that the ultraviolet light source can sequentially pass through the engineering diffuser, lens group, DMD chip, mirror, and projection objective in the DMD lithography system, and expose the target image on the photosensitive material on the three-dimensional nano-moving platform to obtain the to-be-processed exposed image.

[0014] Optionally, segmenting the to-be-processed exposed image to obtain a plurality of sub-images includes:

[0015] Segment the to-be-processed exposed image through an image curvature processing algorithm to obtain sub-images with different curvatures.

[0016] Optionally, calculating the slope of the sub-image includes:

[0017] Fit the sub-image into a straight line by the least squares method to obtain the slope.

[0018] Optionally, designing the displacement path based on the slope includes:

[0019] Design the moving path of the three-dimensional nano-moving platform based on the slope a of the sub-image. When a is greater than 1 or less than -1, the displacement distance on the x-axis is 1 / 2 pixel distance, and the y-axis moves a / 2 pixel distance; when a is greater than -1 and less than 1, the displacement distance on the y-axis is 1 / 2 pixel distance, and the x-axis moves a / 2 pixel distance; when a is equal to 1 or -1, the displacement distances on both the x-axis and y-axis are 1 / 2 pixel distance.

[0020] Optionally, calculate the local dilation of the sub-image based on the displacement path using a motion blur kernel function.

[0021] Optionally, preprocessing the sub-image includes:

[0022] Adopt a Python image processing algorithm to perform edge pixel elimination processing on the sub-image.

[0023] Optionally, exposing the preprocessed sub-image includes:

[0024] Expose the pre - processed sub - images using a simulation exposure model, where the simulation exposure model is the Hopkins exposure model.

[0025] Optionally, exposing the pre - processed sub - images further includes:

[0026] Expose the pre - processed sub - images in order of curvature sorting. The pre - processed sub - images with the same curvature are exposed simultaneously.

[0027] The beneficial effects of the present invention are:

[0028] The present invention analyzes and segments the image edges through curve fitting, calculates the motion trajectory of the three - dimensional nano - moving platform; calculates the local dilation size using the dynamic blur kernel function to process complex exposure images; by making full use of the flexibility of the DMD mode and the high precision of the three - dimensional nano - moving platform, not only the smoothness of the complex image edges is optimized, but also the line width is guaranteed, and complex images with high fidelity can be effectively exposed. Brief Description of the Drawings

[0029] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0030] Figure 1 It is a flowchart of a method for improving the smoothness of DMD digital lithography based on curvature dynamic blur exposure technology according to an embodiment of the present invention;

[0031] Figure 2 It is a schematic diagram of the experimental device according to an embodiment of the present invention;

[0032] Figure 3 It is a schematic diagram of image curvature segmentation according to an embodiment of the present invention;

[0033] Figure 4 It is a schematic diagram of sub - image pre - processing according to an embodiment of the present invention;

[0034] Figure 5 It is a schematic diagram of the displacement of the three - dimensional nano - moving platform according to an embodiment of the present invention;

[0035] Figure 6 It is a simulation diagram of exposure energy according to an embodiment of the present invention;

[0036] Figure 7 It is a comparison diagram of Python simulation before and after optimization according to an embodiment of the present invention;

[0037] Figure 8This is a comparison diagram of the experimental lithography effects before and after optimization in the embodiments of the present invention. Detailed implementation manners

[0038] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0039] To make the above objects, features, and advantages of the present invention more obvious and understandable, the present invention will be further described in detail below in conjunction with the accompanying drawings and specific implementation manners.

[0040] This embodiment provides a method for improving the smoothness of DMD digital lithography based on the curvature dynamic blur exposure technology, as Figure 1 shown, including:

[0041] Select the exposure image to be processed; segment the exposure image to be processed to obtain a plurality of sub-images; calculate the slope of the sub-images, and design a displacement path based on the slope; calculate the local dilation of the sub-images based on the displacement path, and preprocess the sub-images; expose the preprocessed sub-images to obtain a complete exposure image.

[0042] Further, selecting the exposure image to be processed includes:

[0043] Adjust the DMD lithography system to the working state, so that the ultraviolet light source can sequentially pass through the engineering diffuser, lens group, DMD chip, mirror, and projection objective in the DMD lithography system, and expose the target image on the photosensitive material on the three-dimensional nano-moving platform to obtain the exposure image to be processed.

[0044] Specifically, as Figure 2As shown, after the ultraviolet light source (LUMINUS UV-LED, CBT-39-UV-C32-FB400-22) is turned on, it is projected onto the DMD through a lens group, with a resolution of 1920×1080 (pixel size is 7.56μm). The typical peak wavelength of this light source is 405nm, but the beam divergence angle is relatively large, and the light intensity follows a Gaussian distribution. When the light source reaches the subsequent part of the DMD lithography system, a large amount of light energy is lost. In order to make the DMD lithography system have a high light source utilization rate, it is usually necessary to add a better lens group to reduce the loss of light energy. The lens group of the DMD lithography system in this embodiment uses an engineered diffuser (ED1-C20-MD, Thorlabs, USA) and lenses (1 plano-convex lens and 3 biconvex lenses are placed in sequence). The DMD controller model is DLP6500 (Texas Instruments, Texas, USA). The light beam of the photomask is precisely exposed on the photosensitive resin (W200 water-washable resin, eSUN, Shenzhen, China) on the three-dimensional nano-positioning platform (cormorrow, XYZ100, E70. D3S closed-loop piezoelectric controller).

[0045] Place a clean glass slide on the spin coater, use a micropipette to drop an appropriate amount of photoresist resin on the silicon substrate, set the rotation speed and time of the spin coater, and perform spin coating.

[0046] In this embodiment, the DMD lithography system is adjusted to the working state, so that the ultraviolet light source can accurately expose the required exposure image on the photosensitive material on the three-dimensional nano-moving platform through the engineered diffuser and lens group, DMD chip, mirror, and projection objective in the DMD system.

[0047] As an alternative embodiment, the three-dimensional nano-moving platform is a piezoelectric nano-ceramic moving platform.

[0048] Further, segment the to-be-processed exposure image to obtain a number of sub-images, including:

[0049] Segment the to-be-processed exposure image through an image curvature processing algorithm to obtain sub-images with different curvatures.

[0050] Further, calculate the slope of the sub-images, including:

[0051] Fit the sub-images into a straight line by the least squares method to obtain the slope.

[0052] Further, design a displacement path based on the slope, including:

[0053] Design the moving path of the three-dimensional nano-moving platform based on the slope a of the sub-image. When a is greater than 1 or less than -1, the displacement distance on the x-axis is 1 / 2 pixel distance, and the y-axis moves a / 2 pixel distance; when a is greater than -1 and less than 1, the displacement distance on the y-axis is 1 / 2 pixel distance, and the x-axis moves a / 2 pixel distance.

[0054] In this embodiment, the required exposure image is segmented into multiple sub-images through an image curvature processing algorithm, fitted into a straight line according to the least squares method, and its slope is obtained, and the relative moving path of the three-dimensional nano-moving platform and the exposure objective lens is designed, specifically including:

[0055] (1) Segment the exposure image through an image curvature processing algorithm to obtain sub-images with different curvatures, as Figure 3 shown;

[0056] (2) Fit each sub-image by the least squares method to obtain its slope. The least squares method is expressed as:

[0057]

[0058] The optimal estimated values of parameters a and b are obtained by taking the derivative:

[0059]

[0060] (3) Determine the moving path of the three-dimensional nano-moving platform. According to the slope a of the sub-image, when a is greater than 1 or less than -1, the displacement distance on the x-axis is 1 / 2 pixel distance, and the y-axis moves a / 2 pixel distance; when a is greater than -1 and less than 1, the displacement distance on the y-axis is 1 / 2 pixel distance, and the x-axis moves a / 2 pixel distance. When a is equal to 1 or -1, the displacement distances on both the x and y axes are 1 / 2 pixel distance, as Figure 5 shown.

[0061] As an alternative embodiment, the distance of one pixel is 1.89 micrometers.

[0062] Furthermore, calculate the local dilation of the sub-image based on the displacement path using a motion blur kernel function.

[0063] Furthermore, the preprocessing of the sub-image includes:

[0064] Use a Python image processing algorithm to perform edge pixel elimination processing on the sub-image.

[0065] In this embodiment, calculate the local dilation through a motion blur kernel function, and then perform preprocessing on the edge pixels of the sub-image through an image processing algorithm, specifically including:

[0066] (1) Determine the size of the local dilation and calculate the size of the local dilation according to the motion trajectory and the motion blur kernel function;

[0067] Represent the fuzzy direction and distance as a two-dimensional vector (x, y), which can be calculated in the following way:

[0068] length = sqrt(x 2 + y 2 );

[0069] Construct a matrix [kernel_size, kernel_size], and set the cells within the length range around the center of the matrix;

[0070] Each element M[i, j] in the matrix is located at the center of the matrix and is less than or equal to the length from the center. The mathematical formula can be expressed as follows:

[0071]

[0072] M[i, j] = 0;

[0073] The normalized matrix can be expressed as:

[0074]

[0075] In the formula, M is the motion blur kernel matrix, i and j are the indices of the elements in the matrix, length is the length of the motion blur, and sum(M) is the sum of all elements in the matrix.

[0076] The locality of the blur effect represents the local degree of the blur effect, that is, the ratio of the blur kernel size to the product of the blur length and the range of the blur kernel in the image. The larger the value, the more local the blur effect. The local effect generated in the actual exposure can be expressed as:

[0077]

[0078] A = rf;

[0079] Where, r represents the ratio of image magnification, A represents the amount of image magnification, v represents the speed of the target or object movement, T represents the exposure time, and f represents the focal length of the object.

[0080] (2) Use the Python image processing algorithm to perform edge pixel elimination processing on the sub-image to obtain the preprocessed sub-image, as Figure 4 shown.

[0081] Furthermore, the exposure of the preprocessed sub-image includes:

[0082] Use the simulation exposure model to expose the preprocessed sub-image, where the simulation exposure model is the Hopkins exposure model.

[0083] Furthermore, performing exposure on the preprocessed sub-images further includes:

[0084] Exposing the preprocessed sub-images in sequence according to curvature, and exposing the preprocessed sub-images with the same curvature simultaneously.

[0085] Specifically, in this embodiment, the preprocessed sub-images are exposed in sequence to obtain a complete exposed image, which can effectively optimize the edges of the exposed image with high quality. The processing process includes the following:

[0086] Create Python code through a dynamic blur kernel function, import the preprocessed sub-image, set the displacement data, and obtain a simulated exposure energy map after the dynamic blur effect, as Figure 6 shown;

[0087] Select the Hopkins exposure model as the simulation exposure model. The Hopkins exposure model is a mathematical model based on the Fourier transform and can be expressed as:

[0088] When performing lithography simulation, analyze the frequency responses of the mask and the light source using the Fourier transform. Given a function, its one-dimensional continuous Fourier transform is defined as:

[0089]

[0090] where f(x) is the function in the spatial domain; F(f) is the function in the frequency domain; x is the spatial variable; f is the frequency variable.

[0091] For a two-dimensional image, such as the mask image in DMD digital lithography simulation, its two-dimensional Fourier transform is defined as:

[0092]

[0093] where f(x,y) is the image in the two-dimensional spatial domain; F(f x +f y ) is the two-dimensional frequency domain image; (x,y) are the two-dimensional spatial coordinates; (f x ,f y ) are the two-dimensional frequency coordinates.

[0094] The transfer function describes the change of light after passing through the mask. For the Hopkins simulation, the transfer function is:

[0095]

[0096] λ is the wavelength of light; z is the distance from the light source to the object;, f x ,f y are the frequency coordinates respectively; n is the refractive index; j is the imaginary unit.

[0097] The Hopkins equation describes the image intensity in the lithography process and can be expressed for a simplified scenario as follows:

[0098] I(x,y) = F -1 {O(u,v)H(u,v)} 2 ;

[0099] where I(x,y) is the image intensity on the wafer plane; O(u,v) is the Fourier transform of the target function on the mask plane; H(u,v) is the transfer function; F -1 is the inverse Fourier transform operation; (x,y) are two-dimensional spatial coordinates; (u,v) are two-dimensional frequency coordinates.

[0100] Import the energy simulation sub-graph into the Hopkins python exposure model to obtain the optimized exposure simulation results and compare them with the exposure simulation results before optimization. See Figure 7 , and from the results of the simulation experiments, it can be known that the curvature blur dynamic exposure technology has an excellent optimization effect on the edges of the exposure of complex patterns.

[0101] Place the photosensitive resin and the substrate on the moving platform by the immersion method. Each sub-image is imported into the DMD controller and the time is set. The displacement data obtained from each sub-graph is imported into the moving platform and the time is set, and the sub-graph is consistent with the data sequence. When the light source is activated, the light beam irradiates on the photosensitive resin, and the platform starts to move according to the imported data, and the sub-images are exposed in sequence. When the light intensity is 8.11 mw / cm2, the exposure time of the sub-image is about 6 - 7 seconds. After the exposure of each sub-image is completed, both the DMD and the moving platform will pause for a short period of time to continue the process when the next sub-image and displacement data are ready. After all the sub-images are exposed, the optimized exposure experiment image is obtained and compared with the exposure experiment image before optimization. See Figure 8, To evaluate the optimization effect of this method, three different image exposure points were selected in the experiment to measure the line width. The measured values of the line width before optimization were recorded as 57.75μm, 37.00μm, and 30.25μm. After optimization, these numbers were refined to 57.62μm, 36.90μm, and 30.24μm respectively. This resulted in a line width error rate of only ±0.27%, proving the accuracy of the curvature blur dynamic exposure technology. In addition, to quantify the serrated defects, three measurement positions were selected again. Before optimization, the serrated defects were very obvious, being 1.63μm, 1.43μm, and 1.46μm respectively, with an average of 1.5μm. After applying the curvature blur dynamic exposure technology for optimization, these numbers were significantly reduced to 0.22μm, 0.28μm, and 0.14μm, and the average defect size was only 0.21μm. The comparative analysis of the images and data shows that this method not only effectively alleviates the edge defects in complex patterns, but also significantly reduces the impact of dynamic blur on the line width accuracy. This progress proves that this method can improve the fidelity and quality of the lithography process.

[0102] Specifically, in this embodiment, the image edge is analyzed and segmented by curve fitting, and the motion trajectory of the three-dimensional nano-moving platform is calculated; the local dilation size is calculated using the dynamic blur kernel function to process complex exposure images; by making full use of the flexibility of the DMD mode and the high precision of the three-dimensional nano-moving platform, not only the smoothness of the complex image edge is optimized, but also the line width is guaranteed, and complex images with high fidelity can be effectively exposed.

[0103] The embodiments described above are only descriptions of the preferred embodiments of the present invention, and do not limit the scope of the present invention. Without departing from the design spirit of the present invention, various deformations and improvements made by those of ordinary skill in the art to the technical solutions of the present invention shall fall within the protection scope determined by the claims of the present invention.

Claims

1. A method for improving the smoothness of DMD digital lithography based on curvature dynamic fuzzy exposure technology, characterized in that Including: Select the exposure image to be processed; Segment the exposure image to be processed to obtain a number of sub-images; Calculate the slope of the sub-images and design a displacement path based on the slope; Calculate the local dilation of the sub-images based on the displacement path and preprocess the sub-images; Expose the preprocessed sub-images to obtain a complete exposure image.

2. The method for improving the smoothness of DMD digital lithography based on the curvature dynamic blur exposure technology according to claim 1, wherein, Selecting the exposure image to be processed includes: Debug the DMD lithography system to the working state, so that the ultraviolet light source can sequentially pass through the engineering diffuser, lens group, DMD chip, mirror, and projection objective lens in the DMD lithography system, and expose the target image on the photosensitive material on the three-dimensional nano-moving platform to obtain the exposure image to be processed.

3. The method for improving the smoothness of DMD digital lithography based on the curvature dynamic blur exposure technology according to claim 1, wherein Segmenting the exposure image to be processed to obtain a number of sub-images includes: Segment the exposure image to be processed through an image curvature processing algorithm to obtain sub-images with different curvatures.

4. The method for improving the smoothness of DMD digital lithography based on the curvature dynamic fuzzy exposure technology according to claim 1, wherein Calculating the slope of the sub-images includes: Fit the sub-images into a straight line by the least squares method to obtain the slope.

5. The method for improving the smoothness of DMD digital lithography based on the curvature dynamic blur exposure technology according to claim 1, wherein Designing a displacement path based on the slope includes: Design the movement path of the three-dimensional nano-moving platform based on the slope a of the sub-images. When a is greater than 1 or less than -1, the displacement distance on the x-axis is 1 / 2 pixel distance, and the y-axis moves a / 2 pixel distance; when a is greater than -1 and less than 1, the displacement distance on the y-axis is 1 / 2 pixel distance, and the x-axis moves a / 2 pixel distance; when a is equal to 1 or -1, the displacement distances on the x-axis and y-axis are both 1 / 2 pixel distance.

6. The method for improving the smoothness of DMD digital lithography based on the curvature dynamic fuzzy exposure technology according to claim 1, wherein Use a motion blur kernel function to calculate the local dilation of the sub-images based on the displacement path.

7. The method for improving the smoothness of DMD digital lithography based on the curvature dynamic blur exposure technology according to claim 1, characterized in that Preprocessing the sub-images includes: Use the Python image processing algorithm to perform edge pixel elimination processing on the sub-images.

8. The method for improving the smoothness of DMD digital lithography based on the curvature dynamic fuzzy exposure technology according to claim 1, wherein Exposing the preprocessed sub-images includes: Use a simulation exposure model to expose the preprocessed sub-images, where the simulation exposure model is the Hopkins exposure model.

9. The method for improving the smoothness of DMD digital lithography based on the curvature dynamic blur exposure technology according to claim 8, wherein, Exposing the preprocessed sub-images also includes: Expose the preprocessed sub-images in order of curvature, and the preprocessed sub-images with the same curvature are exposed simultaneously.

Images