Two-exposure process preparation method
Through the multimodal sensor and dynamic compensation algorithm, the substrate position is adjusted in real time, and the problem of insufficient alignment error control of the existing two-exposure process is solved, and high-precision graphics superposition and complex structural processing are achieved.
Patent Information
- Application Number
- CN202510385445.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-03-29
- Publication Date
- 2025-06-13
AI Technical Summary
The existing two-exposure process has shortcomings in alignment error control, especially in large-size substrates or high-deep aspect ratio structures, which are difficult to achieve sub-nanometer alignment error control, resulting in pattern dislocation and reducing the functionality and reliability of the device.
Multimodal sensors are used to monitor substrate deformation and offset in real time, and combined with Kalman filter and neural network dynamic compensation algorithm, to generate displacement correction signals, drive nano-level precision displacement stages to adjust substrate position, and compensate for alignment errors caused by thermal drift and mechanical vibration.
Effectively suppress the influence of thermal drift, mechanical vibration and environmental disturbance, reduce the superposition alignment error of the two exposure patterns, ensure the uniformity of line width, and ensure the consistency of high-precision and complex graphic structures.
Smart Images
Figure CN120143558A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of printed circuit board manufacturing, and particularly to a preparation method for a double-exposure process. Background Art
[0002] In the fields of printed circuit boards and micro-nano processing, lithography technology is the core process for achieving high-precision pattern transfer. With the continuous reduction of device size and the improvement of integration, the requirements for pattern resolution, complex structures, and processing accuracy are becoming increasingly stringent. The traditional single-exposure process exposes and develops the photoresist through a single mask. Although the process is simple, its resolution is limited by the optical properties of the photoresist material and the physical limits of the exposure equipment (such as diffraction effects), making it difficult to meet the processing requirements for sub-micron or even nano-scale patterns. Especially when manufacturing multi-layer stacked structures, high-density interconnect lines, or devices with complex topological features, the single-exposure process often faces problems such as blurred pattern edges and uneven feature sizes, resulting in a decrease in the yield rate and performance loss. To overcome the limitations of single exposure, multiple-exposure processes have been proposed in the prior art. For example, the double-exposure technique combines two independent exposure steps to achieve higher-resolution patterns. However, the existing double-exposure techniques still face the following key problems in practical applications:
[0003] For double exposure, different masks are required for pattern overlay. However, existing alignment systems, such as optical alignment or mechanical alignment, are limited by factors such as equipment accuracy, environmental vibration, or thermal drift, and it is difficult to achieve sub-nanometer alignment error control. Especially for large-sized substrates or high aspect ratio structures, alignment offsets will cause pattern misalignment, significantly reducing the functionality and reliability of the device. Summary of the Invention
[0004] The purpose of the present invention is to provide a preparation method for a double-exposure process to solve the problems raised in the above background art.
[0005] To achieve the above purpose, the present invention provides the following technical solution: A preparation method for a double-exposure process, the process preparation method includes the following steps:
[0006] S1: Substrate pretreatment: Coat a photoresist layer on the surface of the substrate and perform pre-baking. A periodic calibration mark array is preset on the surface of the substrate, and the mark spacing is 2-5 times the minimum line width of the target pattern on the substrate;
[0007] S2: First exposure: Perform the first exposure on the photoresist layer through the first mask to form a basic pattern. The exposure energy is E 1 , and after exposure, scan the position of the calibration mark through an in-situ optical detection system and generate initial coordinate data;
[0008] S3: Dynamic alignment compensation: Transfer the substrate to the second exposure station, collect the substrate deformation data and calibration mark offset in real time through a multi-modal sensor, combine with the initial coordinate data, use the dynamic compensation algorithm to generate a displacement correction signal, and drive the nano-precision displacement stage to adjust the substrate position to compensate for the alignment error caused by thermal drift and mechanical vibration;
[0009] S4: Second exposure: At the compensated substrate position, perform a second exposure on the same photoresist layer through the second mask to form a superimposed pattern. A transition zone is set at the edge of the overlapping area between the second mask and the first mask, and the width of the transition zone is 1 / 10 - 1 / 5 of the minimum line width, and the second exposure energy is E2, satisfying E 2 = 0.6E 1 - 0.8E 1 ;
[0010] S5: Development and etching: Adopt a step-by-step development process. First, remove the photoresist in the unexposed area with the first developer concentration, and then remove the photoresist in the partially exposed area with the second developer concentration to form a composite pattern structure. Subsequently, perform plasma etching to transfer the pattern to the substrate;
[0011] S6: Post-processing: Optimize the surface topography of the pattern through atomic layer deposition technology.
[0012] Preferably, the periodic calibration mark array is a composite structure, including a base layer metal mark and a covering layer transparent medium mark. The mark shape is a concentric ring type, and each mark unit contains a sub-wavelength grating structure, and the grating period is 1 / 4 - 1 / 2 of the exposure wavelength.
[0013] Preferably, the dynamic compensation algorithm is based on a Kalman filter and a neural network model. The input parameters include the environmental temperature change rate, the substrate curvature radius, and the vibration spectrum data. The accuracy of the output displacement correction amount is not greater than 0.3 nm, and the compensation frequency is not less than 1 kHz.
[0014] Preferably, the substrate is fixed by a vacuum adsorption device and a multi-point temperature control device during the exposure process. The adsorption pressure is 0.1 - 0.5 MPa, the temperature control accuracy is ±0.1 °C, and a flexible buffer layer is provided at the edge of the substrate to suppress the deformation caused by stress concentration.
[0015] Preferably, the light intensity distribution in the transition zone is optimized by computational lithography technology, so that the light intensity gradient in the overlapping area is not greater than 3%, and the second exposure uses polarization modulation illumination, and the polarization direction is orthogonal to the first exposure to reduce interference noise.
[0016] Preferably, in the development and etching processes, the concentration of the first developer is a 0.2-0.5% tetramethylammonium hydroxide solution, and the development time is 30-60 seconds; the concentration of the second developer is a 0.8-1.2% tetramethylammonium hydroxide solution, and the development time is 10-20 seconds. Between the two development steps, deionized water rinsing and drying are performed.
[0017] Preferably, for the plasma etching, a pulsed radio frequency source is used, with a pulse frequency of 100-500 Hz and a duty cycle of 30-50%. The etching gas is a CF4 / O2 / Ar mixed gas with a mixed volume ratio of 10:1:5, and the etching selectivity is not less than 20:1.
[0018] Preferably, the multimodal sensor includes an infrared sensor, a laser interferometer, and a high-speed camera.
[0019] Preferably, the process preparation method is applicable to large-sized substrates with a diameter of not less than 300 nm, and the final pattern alignment error is not greater than 1.5 nm, and the line width uniformity is not greater than 2 nm.
[0020] Compared with the prior art, the beneficial effects of the present invention are as follows:
[0021] 1. The present invention uses a multimodal sensor to monitor the deformation and offset of the substrate in real time, and combines a Kalman filter and a neural network dynamic compensation algorithm to effectively suppress the influence of thermal drift, mechanical vibration, and environmental disturbance, reducing the pattern overlay alignment error between two exposures.
[0022] 2. The present invention adopts the optimization of the light intensity gradient in the transition region and the orthogonal polarization illumination technology to reduce the interference noise between two exposures, reduce the roughness of the pattern edge, ensure the line width uniformity, and ensure the consistency of high-precision complex pattern structures.
[0023] 3. The present invention fixes the substrate through vacuum adsorption and a multi-point temperature control device, and combines an edge flexible buffer layer to suppress the stress deformation of the substrate, ensuring the flatness of large-sized substrate processing; the stepwise gradient development process accurately controls the residual thickness of the photoresist, and combines pulsed radio frequency plasma etching to improve the pattern transfer fidelity and reduce sidewall defects. Description of the Drawings
[0024] Figure 1 It is a structural schematic diagram of the preparation method of the double-exposure process of the present invention.
[0025] Figure 2 It is a flowchart of the preparation method of the double-exposure process of the present invention.
[0026] Figure 3 It is a flowchart of the substrate pretreatment of the present invention.
[0027] Figure 4 It is a flowchart of the dynamic alignment compensation of the present invention. Detailed implementation manners
[0028] 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.
[0029] Please refer to Figures 1 to 4 , the present invention provides a technical solution: a preparation method of a double-exposure process, and this process preparation method includes the following steps:
[0030] S1: Substrate pretreatment: Coat a photoresist layer on the substrate surface and perform pre-baking. A periodic calibration mark array is preset on the substrate surface, and the mark spacing is 2-5 times the minimum line width of the target pattern on the substrate.
[0031] The substrate is a silicon substrate, a glass substrate or a flexible substrate, such as a polyimide film. The substrate diameter is not less than 300 mm, and the thickness is 0.5-1.5 mm. The photoresist coating device uses a spin coater, model SC-200N, with a rotation speed of 2000-4000 rpm, and coats a photoresist layer with a thickness of 100-500 nm, and is equipped with a vacuum adsorption table to fix the substrate. The pre-baking equipment uses a hot plate baking table, with a temperature of 90-110 °C and a time of 60-90 seconds, for removing the photoresist solvent.
[0032] A periodic calibration mark array is prepared on the substrate surface by electron beam lithography. The periodic calibration mark array is a composite structure, including a base layer metal mark and a covering layer transparent medium mark. The mark shape is a concentric ring type, and each mark unit contains a sub-wavelength grating structure, and the grating period is 1 / 4-1 / 2 of the exposure wavelength.
[0033] S2: First exposure: Perform the first exposure on the photoresist layer through the first mask, forming a basic pattern. The exposure energy is E 1 , and after the exposure, scan the calibration mark position through an in-situ optical detection system and generate initial coordinate data;
[0034] The first exposure uses a deep ultraviolet lithography machine, with a wavelength of 248 nm, an aperture of 0.75, equipped with the first mask, chromium-based, and the pattern line width is not greater than 50 nm. The exposure energy E 1 = 20 mJ / cm 2 .
[0035] S3: Dynamic alignment compensation: Transfer the substrate to the second exposure station. Real-time collect the substrate deformation data and the calibration mark offset through a multi-modal sensor. Combine with the initial coordinate data, and use the dynamic compensation algorithm to generate a displacement correction signal, driving the nano-precision displacement stage to adjust the substrate position to compensate for the alignment error caused by thermal drift and mechanical vibration;
[0036] The multi-modal sensor includes an infrared sensor, a laser interferometer and a high-speed camera. The infrared sensor monitors the temperature distribution of the substrate; the laser interferometer measures displacement; the high-speed camera captures the calibration mark offset.
[0037] The dynamic compensation algorithm is based on the Kalman filter and the neural network model. The input parameters include the environmental temperature change rate, the substrate curvature radius and the vibration spectrum data. The training data comes from 1000 groups of thermal deformation experiments. The accuracy of the output displacement correction amount is not greater than 0.3 nm, and the compensation frequency is not less than 1 kHz.
[0038] During the exposure process, the substrate is fixed by a vacuum adsorption device and a multi-point temperature control device. The adsorption pressure is 0.1 - 0.5 MPa, and the temperature control accuracy is ±0.1 °C. A flexible buffer layer is provided at the edge of the substrate, and the flexible buffer layer material is polyimide to inhibit the deformation caused by stress concentration.
[0039] S4: Second exposure: At the compensated substrate position, perform a second exposure on the same photoresist layer through the second mask to form a superimposed pattern. A transition zone is provided at the edge of the overlapping area between the second mask and the first mask, and the width of the transition zone is 1 / 10 - 1 / 5 of the minimum line width. And the second exposure energy is E2, satisfying E 2 = 0.6E 1 - 0.8E 1 ;
[0040] Use the same lithography machine as the first exposure, adopt the second mask, and the exposure energy E2 = 14 mJ / cm 2 .
[0041] The light intensity distribution of the transition zone is optimized by computational lithography technology, so that the light intensity gradient of the overlapping area is not greater than 3%, and the second exposure adopts polarization modulation illumination, and the polarization direction is orthogonal to the first exposure to reduce interference noise.
[0042] S5: Development and etching: Adopt a step-by-step development process. First, remove the photoresist in the unexposed area with the first developer concentration, and then remove the photoresist in the partially exposed area with the second developer concentration to form a composite pattern structure. Subsequently, perform plasma etching to transfer the pattern to the substrate;
[0043] In development and etching, the concentration of the first developer is a 0.2 - 0.5% tetramethylammonium hydroxide solution, and the development time is 30 - 60 seconds; the concentration of the second developer is a 0.8 - 1.2% tetramethylammonium hydroxide solution, and the development time is 10 - 20 seconds. Between the two developments, deionized water rinsing and drying are performed.
[0044] Plasma etching uses a pulsed radio frequency source with a pulse frequency of 100 - 500 Hz and a duty cycle of 30 - 50%. The etching gas is a CF4 / O2 / Ar mixed gas with a mixing volume ratio of 10:1:5, and the etching selectivity is not less than 20:1.
[0045] S6: Post - processing: Optimize the surface topography of the pattern through atomic layer deposition process. Deposit a 10nm Al 2 O 3 film to fill the side - wall defects of the pattern and reduce the surface roughness.
[0046] The process preparation method is applicable to large - size substrates with a diameter of not less than 300 nm, and the final pattern alignment error is not greater than 1.5 nm, and the line - width uniformity is not greater than 2 nm.
[0047] In summary, the present invention uses a multi - modal sensor to monitor the deformation and offset of the substrate in real - time, combines the Kalman filter and neural network dynamic compensation algorithms to effectively suppress the influence of thermal drift, mechanical vibration, and environmental disturbance, and reduces the superposition alignment error of the two exposure patterns; adopts the optimization of the light - intensity gradient in the transition zone and orthogonal polarization illumination technology to reduce the interference noise of the two exposures, reduce the roughness of the pattern edge, ensure the line - width uniformity, and ensure the consistency of high - precision complex pattern structures; fixes the substrate through vacuum adsorption and multi - point temperature control device, combines with an edge flexible buffer layer to suppress the stress deformation of the substrate and ensure the flatness of large - size substrate processing; uses a step - by - step gradient development process to accurately control the residual thickness of the photoresist, combines with pulsed radio - frequency plasma etching to improve the pattern transfer fidelity and reduce side - wall defects.
[0048] Although the embodiments of the present invention have been shown and described, for those of ordinary skill in the art, it can be understood that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.
Claims
1. A double exposure process preparation method, characterized in that: The process preparation method comprises the following steps: S1: substrate pretreatment: a photoresist layer is coated on the substrate surface and pre-baked, and a periodic calibration mark array is preset on the substrate surface, and the mark spacing is 2-5 times the minimum line width of the substrate target pattern; S2: First exposure: The photoresist layer is exposed for the first time through the first mask to form a basic pattern. The exposure energy is E1. After exposure, the calibration mark position is scanned by the in-situ optical detection system and the initial coordinate data is generated. S3: Dynamic alignment compensation: The substrate is transferred to the secondary exposure station, and the substrate deformation data and calibration mark offset are collected in real time by a multimodal sensor. Combined with the initial coordinate data, a displacement correction signal is generated using a dynamic compensation algorithm to drive the nano-level precision translation stage to adjust the substrate position and compensate for the alignment error caused by thermal drift and mechanical vibration. S4: secondary exposure: at the compensated substrate position, the same photoresist layer is exposed for the second time through the second mask to form a superimposed pattern, wherein a transition zone is set at the edge of the overlapping area between the second mask and the first mask, the width of the transition zone is 1 / 10-1 / 5 of the minimum line width, and the secondary exposure energy is E2, satisfying E2=0.6E1-0.8E1; S5: Development and etching: A step-by-step development process is used to first remove the photoresist in the unexposed area with a first developer concentration, and then remove the photoresist in the partially exposed area with a second developer concentration to form a composite pattern structure, followed by plasma etching to transfer the pattern to the substrate; S6: Post-processing: Optimizing the surface morphology of the pattern through atomic layer deposition process.
2. The double exposure process preparation method according to claim 1, characterized in that: The periodic calibration mark array is a composite structure, including a base layer metal mark and a cover layer transparent medium mark, the mark shape is a concentric ring type, and each mark unit includes a sub-wavelength grating structure, and the grating period is 1 / 4-1 / 2 of the exposure wavelength.
3. The double exposure process preparation method according to claim 1, characterized in that: The dynamic compensation algorithm is based on a Kalman filter and a neural network model. The input parameters include the ambient temperature change rate, the substrate curvature radius and the vibration spectrum data. The accuracy of the output displacement correction amount is no more than 0.3nm, and the compensation frequency is no less than 1kHz.
4. The double exposure process preparation method according to claim 1, characterized in that: The substrate is fixed by a vacuum adsorption device and a multi-point temperature control device during the exposure process. The adsorption pressure is 0.1-0.5MPa, the temperature control accuracy is ±0.1°C, and a flexible buffer layer is provided at the edge of the substrate to suppress deformation caused by stress concentration.
5. The double exposure process preparation method according to claim 1, characterized in that: The light intensity distribution in the transition zone is optimized by computational lithography technology so that the light intensity gradient in the overlapping area is no more than 3%, and the secondary exposure adopts polarization modulation illumination, and the polarization direction is orthogonal to the primary exposure to reduce interference noise.
6. The double exposure process preparation method according to claim 1, characterized in that: In the development and etching, the first developer concentration is 0.2-0.5% tetramethylammonium hydroxide solution, and the development time is 30-60 seconds; the second developer concentration is 0.8-1.2% tetramethylammonium hydroxide solution, and the development time is 10-20 seconds, and deionized water is used for rinsing and drying between the two developments.
7. The double exposure process preparation method according to claim 1, characterized in that: The plasma etching uses a pulsed radio frequency source with a pulse frequency of 100-500 Hz and a duty cycle of 30-50%. The etching gas is a CF4 / O2 / Ar mixed gas with a mixed volume ratio of 10:1:5 and an etching selectivity ratio of not less than 20:
1.
8. The double exposure process preparation method according to claim 1, characterized in that: The multimodal sensor includes an infrared sensor, a laser interferometer and a high-speed camera.
9. The double exposure process preparation method according to claim 1, characterized in that: The process preparation method is suitable for large-size substrates with a diameter of not less than 300 nm, and the final pattern alignment error is not greater than 1.5 nm, and the line width uniformity is not greater than 2 nm.