High-precision optical element processing system based on ion beam modification and adaptive detection
By combining the online detection of Hartmann wavefront sensor and wavelength tuning interferometer in the optical processing system, the ion beam residence time is predicted using a timing learning algorithm, real-time surface shape feedback and efficient surface shape convergence are achieved at the sub-nanometer level, and the problem of surface shape error and real-time control difficulties in traditional optical processing methods is solved.
Patent Information
- Application Number
- CN202510476960.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-04-16
- Publication Date
- 2025-05-30
AI Technical Summary
Traditional optical processing methods are prone to introduce medium and high-frequency surface shape errors when processing aspherical surfaces and free curved surfaces, making it difficult to achieve sub-nanometer-level accuracy, and lack real-time closed-loop control, making it difficult to control surface shape distortion in ultra-thin component processing.
The Hartmann wavefront sensor and wavelength tuning interferometer are used for online detection, real-time feedback at the sub-nanometer level is realized, and the physical model of material removal and thermal stress distribution is combined for reverse compensation. The timing learning algorithm is used to predict the ion beam dwell time and path planning, and the surface shape convergence speed is improved.
It realizes real-time surface shape feedback at the sub-nanometer level, improves the surface shape convergence speed by more than 40%, supports special-shaped component processing, solves the problem of ultra-thin component processing, and is adapted to high-precision optical component processing of different materials.
Smart Images

Figure CN120055906A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of optical processing, and specifically to a high-precision optical element processing system based on ion beam figuring and adaptive detection. Background Art
[0002] Medium and high-frequency error problems: When processing aspherical and free-form surfaces by traditional methods such as mechanical polishing and magnetorheological polishing, medium and high-frequency surface errors are easily introduced, resulting in low convergence efficiency and difficulty in achieving sub-nanometer accuracy (RMS≤λ / 50).
[0003] Lack of real-time closed-loop control: Existing technologies rely on off-line detection and manual intervention, and errors cannot be dynamically corrected during the processing.
[0004] Challenges in processing ultra-thin components: Optical components with a thickness less than 1 mm are prone to deformation due to thermal stress or mechanical stress during processing, and it is difficult to control surface shape distortion by traditional methods.
[0005] Driven by industry demands: With the surge in the demand for ultra-precision optical components in fields such as extreme ultraviolet lithography (EUV) and space optical remote sensing (such as surface shape accuracy RMS≤1nm, roughness Ra≤0.2nm), there is an urgent need for an efficient manufacturing method that can achieve a full closed-loop of real-time detection - processing - compensation.
[0006] Traditional optical processing methods have the following limitations: Medium and high-frequency errors are easily introduced when processing complex surface shapes, resulting in low surface shape convergence efficiency, lack of real-time surface shape detection and feedback during the processing, relying on manual intervention, and the surface shape accuracy of ultra-thin components is easily reduced due to stress deformation.
[0007] Therefore, there is an urgent need for an improved technology based on a high-precision optical element processing system of ion beam figuring and adaptive detection to solve this problem existing in the prior art. Summary of the Invention
[0008] The purpose of the present invention is to provide online detection through a Hartmann wavefront sensor and a wavelength-tuned interferometer to achieve sub-nanometer real-time feedback of processing errors (accuracy up to 0.1nm), break through the lag of traditional off-line detection, reverse-compensate the workpiece table pose based on the physical model of material removal amount and thermal stress distribution, solve the problem of deformation in the processing of ultra-thin components, use a sequential learning algorithm to predict the ion beam dwell time and path planning, increase the surface shape convergence speed by more than 40%, support the processing of special-shaped components such as aspherical surfaces, off-axis paraboloids, and microlens arrays, and the modular design allows the replacement of the detection module or the processing module to adapt to the processing of high-precision optical components of different materials, so as to solve the problems raised in the above background art.
[0009] To achieve the above object, the present invention provides the following technical solutions: A high-precision optical element processing system based on ion beam shaping and adaptive detection, including an optical element processing system, an ion beam processing module, an adaptive optical detection module, and an intelligent control module. The optical element processing system mainly includes an ion beam processing module, an adaptive optical detection module, and an intelligent control module;
[0010] The ion beam processing module includes an ion source system and a six-axis linkage workpiece table;
[0011] The adaptive optical detection module includes a wavelength-tunable interferometer and a Hartmann wavefront sensor;
[0012] The intelligent control module includes a processing path optimization unit and a stress compensation unit.
[0013] Preferably, the ion source system uses a radio frequency ion source, and the beam current density range is 0.1 - 5 mA / cm 2 ;
[0014] Dynamic adjustment mechanism for the incident angle of the radio frequency ion source: The incident angle of the ion beam is continuously adjustable from 0° to 60° through a piezoelectric ceramic actuator, and the angular resolution ≤ 0.1°;
[0015] Beam spot size control: Multi-pole magnetic lenses are used for focusing, and the beam spot diameter can be adjusted in the range of 50 μm - 2 mm.
[0016] Preferably, the six-axis linkage workpiece table has linear axes of motion freedom X, Y, Z. The stroke of the six-axis linkage workpiece table is ±200 mm, and the positioning accuracy formed is ±10 nm. The angular range of the A, B, C rotation axes on the six-axis linkage workpiece table is ±30°, and the resolution is 0.001°. The six-axis linkage workpiece table is equipped with an anti-vibration air-bearing base.
[0017] Preferably, the light source of the wavelength-tunable interferometer is a tunable laser, and the wavelength of the tunable laser is 632.8 nm, and the tuning range is ±5 nm;
[0018] The wavelength-tunable interferometer triggers a full-field scan every 5 minutes during the rough machining stage to generate a three-dimensional surface shape error map.
[0019] Preferably, the micro-lens array of the Hartmann wavefront sensor is 12×12 sub-apertures, the sampling frequency of the Hartmann wavefront sensor is 100 Hz, and the local slope error of the machining area is detected in real time to be 0.05 λ / mm;
[0020] The Hartmann wavefront sensor is synchronized with the ion beam processing, and continuously feeds back the wavefront aberration during the finish machining stage to trigger the dynamic adjustment of the ion beam incident angle.
[0021] Preferably, the processing path optimization unit is based on a time series prediction model of an LSTM neural network, with the input being historical surface shape error data and ion beam parameters (beam current, incident angle), and the output being the optimal dwell time distribution;
[0022] Hardware support: GPU acceleration computing card (NVIDIA A100), with the time consumption for single path planning < 10 seconds.
[0023] Preferably, the stress compensation unit
[0024] Thermal-mechanical coupling model: Predict the thermal stress distribution based on the material removal rate (calculation formula: Δh = k·I·t·cosθ, where k is the etching coefficient, I is the beam current, t is the dwell time, and θ is the incident angle);
[0025] Compensation strategy: Offset the deformation through the Z-axis micro-displacement of the six-axis workpiece stage (compensation amount Δz = α·ΔT·h, where α is the thermal expansion coefficient, ΔT is the temperature rise, and h is the component thickness).
[0026] Preferably, the processing method of the high-precision optical element processing system based on ion beam shaping and adaptive detection includes the following steps:
[0027] Use a wavelength-tuned interferometer to perform a full-field scan on the workpiece to generate an initial surface shape error map, PV value, RMS value, and mid-high frequency error distribution;
[0028] According to the error distribution, divide the processing area (for example, mark the area where PV > 100 nm as the "rough machining area");
[0029] Discretize the workpiece surface into grids with a grid spacing of 0.5 mm, calculate the optimal dwell time for each grid point based on the LSTM model, and generate an ion beam scanning path using a spiral filling trajectory to avoid edge effects;
[0030] First, perform rough machining on the mirror surface, adjust the ion beam mode, with a large beam current of (3 mA / cm 2 ), a large beam spot of (2 mm), a workpiece stage movement speed of 10 mm / s (X / Y axis), and a material removal rate > 50 nm / min;
[0031] Interrupt the machining every 5 minutes, trigger the interferometer detection, update the surface shape error map. If the residual error in a local area is detected to be > 30 nm, then perform a secondary scan on this area (extend the dwell time by 20%);
[0032] After the rough machining is completed, enter the finish machining stage, adjust the ion beam mode to a small beam current of (0.5 mA / cm 2 ), a small beam spot of (50 μm), a workpiece stage movement speed of 1 mm / s (X / Y axis), and a material removal rate < 5 nm / min;
[0033] The Hartmann sensor monitors the wavefront slope of the machining area in real time. If a local slope deviation > 0.1λ / mm is detected, the incident angle of the ion beam is adjusted, and the dwell time is dynamically corrected by the ribbon machine;
[0034] The temperature rise is calculated based on the cumulative material removal amount, and the deformation amount is predicted by finite element analysis (FEA) to generate a compensation displacement command;
[0035] Through the fine movement of the Z-axis and the tilting of the B / C axes of the six-axis workpiece stage, the thermal deformation is counteracted in the reverse direction. After compensation, the interferometer is triggered for verification until the change amount of the surface shape error < 1nm;
[0036] The Ra value is measured using a white light interferometer, and the scanning area is 5×5μm 2 , if Ra > 0.2nm, local ion beam shaping is triggered;
[0037] The full-field RMS value needs to meet λ / 50 (when λ = 632.8nm, RMS ≤ 12.6nm), and in the actual embodiment, RMS 1.1nm is achieved.
[0038] Compared with the prior art, the beneficial effects of the present invention are:
[0039] This system realizes sub-nanometer real-time feedback of machining errors (accuracy up to 0.1nm) through online detection by a Hartmann wavefront sensor and a wavelength-tuned interferometer, breaking through the lag of traditional offline detection. Based on the physical model of material removal amount and thermal stress distribution, the pose of the workpiece stage is compensated in the reverse direction to solve the problem of machining deformation of ultra-thin components. The dwell time and path planning of the ion beam are predicted using a sequential learning algorithm, which increases the surface shape convergence speed by more than 40%. It supports the machining of special-shaped components such as aspherical surfaces, off-axis paraboloids, and microlens arrays, and the modular design allows the replacement of the detection module or the machining module to adapt to different materials. BRIEF DESCRIPTION OF THE DRAWINGS
[0040] Figure 1 It is a schematic diagram of the optical element processing system of the present invention.
[0041] In the figure: 1. Optical element processing system; 2. Ion beam processing module; 3. Adaptive optical detection module; 4. Intelligent control module; 5. Ion source system; 6. Six-axis linkage workpiece stage; 7. Wavelength-tuned interferometer; 8. Hartmann wavefront sensor; 9. Machining path optimization unit; 10. Stress compensation unit. DETAILED DESCRIPTION OF THE INVENTION
[0042] 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.
[0043] Please refer to Figure 1 , the present invention provides a technical solution for a high-precision optical element processing system based on ion beam shaping and adaptive detection: 1. A high-precision optical element processing system based on ion beam shaping and adaptive detection, including an optical element processing system 1, an ion beam processing module 2, an adaptive optical detection module 3, and an intelligent control module 4. It is characterized in that: the optical element processing system 1 mainly includes an ion beam processing module 2, an adaptive optical detection module 3, and an intelligent control module 4. The ion beam processing module 2 includes an ion source system 5 and a six-axis linkage workpiece table 6. The adaptive optical detection module 3 includes a wavelength-tunable interferometer 7 and a Hartmann wavefront sensor 8. The intelligent control module 4 includes a processing path optimization unit 9 and a stress compensation unit 10.
[0044] The ion source system 5 uses a radio frequency ion source, and the beam current density range is 0.1 - 5 mA / cm 2 ;
[0045] Dynamic adjustment mechanism for the incident angle of the radio frequency ion source: The incident angle of the ion beam can be continuously adjusted from 0° to 60° through a piezoelectric ceramic driver, and the angle resolution ≤ 0.1°.
[0046] Beam spot size control: Focusing is achieved using a multipole magnetic lens, and the beam spot diameter can be adjusted within the range of 50 μm - 2 mm.
[0047] The six-axis linkage workpiece table 6 has translational degrees of freedom in the X, Y, and Z axes. The stroke of the six-axis linkage workpiece table 6 is ±200 mm, and the positioning accuracy formed is ±10 nm. The angular range of the A, B, and C rotational axes on the six-axis linkage workpiece table 6 is ±30°, and the resolution is 0.001°. The six-axis linkage workpiece table 6 is equipped with an anti-vibration air-bearing base.
[0048] The light source of the wavelength-tunable interferometer 7 is a tunable laser. The wavelength of the tunable laser is 632.8 nm, and the tuning range is ±5 nm. The wavelength-tunable interferometer 7 triggers a full-field scan every 5 minutes during the rough machining stage to generate a three-dimensional surface shape error map.
[0049] The Hartmann wavefront sensor 8 has a microlens array of 12×12 sub-apertures. The sampling frequency of the Hartmann wavefront sensor 8 is 100 Hz. The local slope error of the machining area is detected in real time to be 0.05λ / mm. The Hartmann wavefront sensor 8 is synchronized with the ion beam machining and continuously feeds back the wavefront distortion during the finish machining stage to trigger the dynamic adjustment of the ion beam incident angle.
[0050] The machining path optimization unit 9 is based on a time series prediction model of an LSTM neural network. The inputs are historical surface shape error data, ion beam parameters (beam current, incident angle), and the output is the optimal dwell time distribution.
[0051] Hardware support: GPU acceleration card (NVIDIA A100), and the time consumption for single path planning is <10 seconds.
[0052] The thermal-mechanical coupling model of the stress compensation unit (10): predicts the thermal stress distribution according to the material removal rate (calculation formula: Δh = k·I·t·cosθ, where k is the etching coefficient, I is the beam current, t is the dwell time, and θ is the incident angle).
[0053] Compensation strategy: offsets the deformation through the Z-axis micro-displacement of the six-axis workpiece stage (compensation amount Δz = α·ΔT·h, where α is the thermal expansion coefficient, ΔT is the temperature rise, and h is the component thickness).
[0054] The machining method of the high-precision optical element machining system based on ion beam figuring and adaptive detection:
[0055] Use a wavelength-tuned interferometer to perform a full-field scan on the workpiece to generate an initial surface shape error map, PV value, RMS value, and mid-high frequency error distribution. According to the error distribution, divide the machining area (for example, mark the area where PV > 100 nm as the "rough machining area").
[0056] Discretize the workpiece surface into grids with a grid spacing of 0.5 mm. Calculate the optimal dwell time for each grid point based on the LSTM model, and generate an ion beam scanning path using a spiral filling trajectory to avoid edge effects.
[0057] First, perform rough machining on the mirror surface, adjust the ion beam mode, with a large beam current of (3 mA / cm 2 ), a large beam spot of (2 mm), and the workpiece stage moving speed of 10 mm / s (X / Y axis), and the material removal rate > 50 nm / min.
[0058] Interrupt the machining every 5 minutes, trigger the interferometer detection, update the surface shape error map. If the residual error in the local area is detected to be > 30 nm, then perform a secondary scan on this area (extend the dwell time by 20%).
[0059] After the rough machining is completed, enter the finish machining stage, and adjust the ion beam mode to a small beam current of (0.5 mA / cm2 ), the small beam spot is (50 μm), the moving speed of the worktable is 1 mm / s (X / Y axis), and the material removal rate is < 5 nm / min.
[0060] The Hartmann sensor monitors the wavefront slope of the machining area in real time. If the detected local slope deviation > 0.1λ / mm, adjust the ion beam incident angle, and the ribbon machine dynamically corrects the dwell time.
[0061] Calculate the temperature rise based on the cumulative material removal amount, predict the deformation amount by finite element analysis (FEA), and generate a compensation displacement command.
[0062] Through the Z-axis fine movement and B / C-axis tilt of the six-axis worktable, reverse offset the thermal deformation. After compensation, trigger the interferometer for verification until the change in surface shape error < 1 nm.
[0063] Measure the Ra value using a white light interferometer, and the scanning area is 5×5 μm 2 , if Ra > 0.2 nm, trigger local ion beam shaping.
[0064] The full-field RMS value needs to meet λ / 50 (when λ = 632.8 nm, RMS ≤ 12.6 nm), and in the actual embodiment, RMS 1.1 nm is achieved.
[0065] This system realizes sub-nanometer real-time feedback of machining errors (accuracy up to 0.1 nm) through online detection by a Hartmann wavefront sensor and a wavelength-tuned interferometer, breaks through the lag of traditional offline detection, and based on the physical model of material removal amount and thermal stress distribution, reversely compensates the pose of the worktable to solve the problem of machining deformation of ultra-thin components. Using a sequential learning algorithm to predict the ion beam dwell time and path planning, the surface shape convergence speed is increased by more than 40%, supports the machining of special-shaped components such as aspherical surfaces, off-axis paraboloids, and microlens arrays, and the modular design allows replacing the detection module or machining module to adapt to different materials.
[0066] 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 high-precision optical element processing system based on ion beam shaping and adaptive detection, comprising an optical element processing system (1), an ion beam processing module (2), an adaptive optical detection module (3), and an intelligent control module (4), characterized in that: The optical element processing system (1) mainly comprises an ion beam processing module (2), an adaptive optical detection module (3), and an intelligent control module (4); The ion beam processing module (2) comprises an ion source system (5) and a six-axis linkage workpiece table (6); The adaptive optical detection module (3) comprises a wavelength tuning interferometer (7) and a Hartmann wavefront sensor (8); The intelligent control module (4) comprises a processing path optimization unit (9) and a stress compensation unit (10).
2. The high-precision optical element processing system based on ion beam shaping and adaptive detection according to claim 1, characterized in that: The ion source system (5) adopts a radio frequency ion source, and the beam density range is 0.1-5 mA / cm 2 ; Dynamic adjustment mechanism of incident angle of RF ion source: the incident angle of ion beam is continuously adjustable from 0° to 60° through piezoelectric ceramic driver, and the angle resolution is ≤ 0.1°; Beam spot size control: Multi-pole magnetic lens focusing is adopted, and the beam spot diameter can be adjusted in the range of 50μm-2mm.
3. The high-precision optical component processing system based on ion beam shaping and adaptive detection according to claim 1, characterized in that: The six-axis linkage workpiece platform (6) has the following degrees of freedom: X, Y, and Z linear axes; the travel of the six-axis linkage workpiece platform (6) is ±200 mm, and the resulting positioning accuracy is ±10 nm; the angle range of the A, B, and C rotation axes on the six-axis linkage workpiece platform (6) is ±30°, and the resolution is 0.001°; the six-axis linkage workpiece platform (6) is equipped with an anti-vibration air-floating base.
4. The high-precision optical element processing system based on ion beam shaping and adaptive detection according to claim 1, characterized in that: The light source of the wavelength tuning interferometer (7) is a tunable laser, the wavelength of the tunable laser is 632.8 nm, and the tuning range is ±5 nm; The wavelength tuning interferometer (7) triggers a full-field scan every 5 minutes during the rough processing stage to generate a three-dimensional surface error map.
5. The high-precision optical element processing system based on ion beam shaping and adaptive detection according to claim 1, characterized in that: The microlens array of the Hartmann wavefront sensor (8) has a 12×12 sub-aperture, the sampling frequency of the Hartmann wavefront sensor (8) is 100 Hz, and the local slope error of the real-time detection processing area is 0.05λ / mm; The Hartmann wavefront sensor (8) is synchronized with the ion beam processing, continuously feeds back the wavefront distortion during the finishing stage, and triggers the dynamic adjustment of the ion beam incident angle.
6. The high-precision optical element processing system based on ion beam shaping and adaptive detection according to claim 1, characterized in that: The process path optimization unit (9) is based on a time series prediction model of an LSTM neural network, the input of which is historical surface error data and ion beam parameters (beam current, incident angle), and the output of which is an optimal residence time distribution; Hardware support: GPU accelerated computing card (NVIDIA A100), single path planning takes less than 10 seconds.
7. The high-precision optical element processing system based on ion beam shaping and adaptive detection according to claim 1, characterized in that: The stress compensation unit (10) Thermal-mechanical coupling model: predicts thermal stress distribution based on material removal rate (calculation formula: Δh = k·I·t·cosθ, k is the etching coefficient, I is the beam current, t is the dwell time, and θ is the incident angle); Compensation strategy: offset the deformation through the Z-axis micro-displacement of the six-axis worktable (compensation amount Δz = α·ΔT·h, α is the thermal expansion coefficient, ΔT is the temperature rise, and h is the component thickness).
8. A method for realizing the high-precision optical element processing system based on ion beam shaping and adaptive detection as claimed in claim 1 comprises the following steps: Use a wavelength-tuned interferometer to perform full-field scanning of the workpiece to generate an initial surface error map, PV value, RMS value, and mid- and high-frequency error distribution; Divide the processing area according to the error distribution (for example, mark the area with PV>100nm as "rough processing area"); The workpiece surface is discretized into a grid with a grid spacing of 0.5 mm. The optimal dwell time of each grid point is calculated based on the LSTM model, and the ion beam scanning path is generated using a spiral filling trajectory to avoid edge effects. The mirror surface is first rough-machined, and the ion beam mode is adjusted, with a maximum beam current of (3mA / cm 2 ), the largest beam spot is (2mm), the workpiece stage movement speed is 10mm / s (X / Y axis), and the material removal rate is >50nm / min; The processing is interrupted every 5 minutes to trigger the interferometer detection and update the surface error map. If the residual error in the local area is detected to be greater than 30nm, a second scan is added to the area (the dwell time is extended by 20%). After the rough machining, the fine machining phase is carried out, and the ion beam mode is adjusted to a small beam current (0.5mA / cm 2 ), small beam spot size (50μm), workpiece stage movement speed 1mm / s (X / Y axis), material removal rate <5nm / min; The Hartmann sensor monitors the wavefront slope of the processing area in real time. If a local slope deviation of >0.1λ / mm is detected, the ion beam incident angle is adjusted and the draw frame dynamically corrects the dwell time; Calculate temperature rise based on the cumulative amount of material removed, predict deformation through finite element analysis (FEA), and generate compensation displacement instructions; Through the Z-axis micro-motion and B / C-axis tilting of the six-axis workpiece stage, the thermal deformation is reversed and the interferometer is triggered for verification after compensation until the surface error change is less than 1nm; Ra value was measured using white light interferometer, scanning area 5×5μm 2 , if Ra>0.2nm, trigger local ion beam modification; The full-field RMS value must satisfy λ / 50 (λ=632.8nm corresponds to RMS≤12.6nm), and in the actual embodiment, the RMS reaches 1.1nm.
Citation Information
Cited By
Self-adaptive optical polishing system and method based on dynamic temperature control and multi-modal detection
CN120663185A