Extreme ultraviolet light source chip removing device and design method and design system thereof

Abstract

The invention provides an extreme ultraviolet light source debris removal device and a design method and a design system thereof, and the design method comprises the steps: employing an Euler model to cooperate with a slippage boundary to establish a gas flow field model, and carrying out the gas flow field numerical simulation of background gas in a vacuum or low-pressure environment; establishing a particle motion model by adopting a Lagrange model, and performing motion trail tracking calculation on particles; carrying out coupling calculation on the gas flow field and the particle motion trail to obtain a particle motion distribution simulation result; performing experimental verification on the particle motion distribution simulation result to ensure that the particle motion simulation result is consistent with the actual motion, and correcting the model parameters through the experimental verification result; a Bayesian optimization algorithm is used, a Gaussian process regression model is used as an agent model, particle removal efficiency and energy consumption are used as optimization targets, structural design parameters of the debris removal device are globally optimized, and an optimal solution of the parameters is gradually approached and used for guiding actual device manufacturing and application.

Description

Technical Field

[0001] This invention relates to the field of extreme ultraviolet (EUV) light source debris removal technology, specifically to an extreme ultraviolet (EUV) light source debris removal device and its design method and system. Background Technology

[0002] Extreme ultraviolet (EUV) light sources are widely used in fields such as semiconductor lithography. During their operation, they inevitably generate metallic or non-metallic debris particles. These particles deposit on optical components or cavity walls, causing mirror damage, decreased reflectivity, and system contamination, severely affecting the output power and lifespan of the EUV light source.

[0003] To control the impact of debris, existing technologies have proposed various removal schemes. For example, foil trapping methods use physical barriers to intercept some particles; gas jet removal methods use argon or helium jets to drive particle migration; and vacuum pumping methods use mechanical or molecular pumps to reduce cavity pressure and assist particle migration. These methods alleviate the debris deposition problem to some extent, but they still have the following shortcomings:

[0004] Unstable removal efficiency: The removal effect varies greatly for particles of different sizes or at different speeds, and some particles are difficult to remove.

[0005] High energy consumption: The parameters of the gas jet and pumping unit have not been globally optimized, resulting in gas waste and excessive pump power.

[0006] Insufficient adaptability: When operating conditions (such as pressure range, gas type) change, the purging effect is difficult to maintain a stable level. Summary of the Invention

[0007] To address the aforementioned shortcomings in the prior art, this invention provides an extreme ultraviolet light source debris removal device, its design method, and its design system.

[0008] According to one aspect of the present invention, a design method for an extreme ultraviolet light source debris removal device is provided, comprising:

[0009] A gas flow field model was established using the Euler model in conjunction with slip boundary to perform numerical simulation of the gas flow field of background gas under vacuum or low-pressure conditions.

[0010] A particle motion model is established using the Lagrange model, and the particle motion trajectory is tracked and calculated.

[0011] Based on the gas flow field model and the particle motion model, the gas flow field and particle motion trajectory are coupled and calculated to obtain the simulation results of particle motion distribution;

[0012] The simulation results of particle motion distribution are experimentally verified to ensure that the simulation results of particle motion are consistent with the actual motion. The model parameters are then corrected based on the experimental verification results to obtain numerical simulation results of particle removal efficiency.

[0013] Using the numerical simulation results of particle removal efficiency as initial data, and based on the continuously updated particle removal efficiency under different design parameters, the design parameters of the debris removal device are globally optimized using a Bayesian optimization algorithm and a Gaussian process regression model as a surrogate model, with the optimization objectives of maximizing particle removal efficiency and minimizing energy consumption. The optimal solution of the parameters is gradually approached and then obtained, which is used to guide the manufacturing and application of actual devices.

[0014] According to another aspect of the present invention, a design system for an extreme ultraviolet light source debris removal device is provided, comprising:

[0015] The gas flow field simulation module uses the Euler model in conjunction with slip boundary to establish a gas flow field model and performs numerical simulation of the gas flow field of background gas under vacuum or low-pressure environment.

[0016] The particle motion simulation module uses the Lagrange model to establish a particle motion model and calculates the particle's motion trajectory.

[0017] The flow field-particle coupling module performs coupled calculations on the gas flow field and the particle motion model based on the gas flow field model and the particle motion model to obtain the simulation results of particle motion distribution.

[0018] The experimental verification module verifies the simulation results of particle motion distribution to ensure that the simulation results of particle motion are consistent with the actual motion, and corrects the model parameters based on the experimental verification results to obtain numerical simulation results of particle removal efficiency.

[0019] The device parameter optimization module is based on the continuously updated numerical simulation results of particle removal efficiency under different design parameters. It uses a Bayesian optimization algorithm and a Gaussian process regression model as a surrogate model. With the optimization objectives of maximizing particle removal efficiency and minimizing energy consumption, it performs global optimization of the design parameters of the debris removal device, gradually approaching the optimal solution of parameters and thus obtaining the optimal solution of design parameters, which is used to guide the actual device manufacturing and application.

[0020] According to a third aspect of the present invention, an extreme ultraviolet light source debris removal device is provided, comprising: a cavity, a jet structure, a pumping unit, and a control module; wherein:

[0021] The particle source is directly connected to the cavity;

[0022] The jet structure is connected to the gas source via a gas supply pipeline and is fixed at a predetermined position in the cavity;

[0023] The extraction unit is connected to the exhaust port of the cavity to form a stable pressure differential drive;

[0024] The control module is electrically connected to the jet structure and the extraction unit respectively, and obtains the optimal design parameters by executing the design method or running the design system described above, and adjusts the jet flow rate, jet angle and extraction speed in real time.

[0025] By adopting the above technical solution, the present invention has at least one of the following beneficial effects compared with the prior art:

[0026] The extreme ultraviolet light source debris removal device, its design method, and its design system provided by this invention can efficiently achieve the directional discharge and controlled removal of debris particles in vacuum or low-pressure environments. Specifically:

[0027] The debris removal efficiency is significantly improved. The design method and system of the extreme ultraviolet light source debris removal device provided by this invention introduces an Eulerian-Lagrange coupling model, which can simultaneously consider the gas flow field distribution and particle motion trajectory, accurately predict the transport path of particles in the cavity, and avoid the problem of "separation between airflow simulation and particle behavior" in traditional empirical design. It realizes the synergistic simulation of gas and solid phases, so the directional migration effect of particles is better and the removal efficiency is improved by more than 20%.

[0028] The accuracy of simulation predictions is improved. The design method and system for the extreme ultraviolet light source debris removal device provided by this invention introduces Maxwell slip boundary conditions in the gas flow field modeling and uses a rarefied gas correction factor in the particle model. This can accurately reflect the discontinuous effects of gas under low pressure conditions, ensuring the consistency between simulation results and the actual operating environment of the EUV device, thereby reducing the deviation between experiments and simulations.

[0029] Intelligent structural design optimization and reduced energy consumption. The design method and system for the extreme ultraviolet light source debris removal device provided by this invention introduce a Bayesian optimization algorithm and use debris removal efficiency and energy consumption as multi-objective optimization indicators. Compared with traditional manual parameter tuning or single-objective optimization, Bayesian optimization can quickly converge to the globally optimal design parameters under limited experimental data, thereby reducing gas consumption and pump power while maintaining high removal efficiency, achieving energy-saving effects.

[0030] The device is more adaptable. The extreme ultraviolet light source debris removal device provided by this invention adopts a combination design of jet structure and pumping unit. The jet gas and vacuum pumping work together and can be flexibly adjusted according to different pressures and particle sizes. It is suitable for debris removal needs in various low-pressure environments such as EUV light sources, plasma sources, and vacuum devices. Attached Figure Description

[0031] Other features, objects, and advantages of the present invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0032] Figure 1 This is a flowchart illustrating the design method of an extreme ultraviolet light source debris removal device in a preferred embodiment of the present invention.

[0033] Figure 2 This is a schematic diagram of the component modules of the design system of the extreme ultraviolet light source debris removal device in a preferred embodiment of the present invention.

[0034] Figure 3 This is a flowchart illustrating the overall workflow of the extreme ultraviolet light source debris removal device in a preferred embodiment of the present invention.

[0035] Figure 4 This is a schematic diagram of the composition of the extreme ultraviolet light source debris removal device in a preferred embodiment of the present invention; in the figure, 1 is the cavity, 2 is the jet structure, 3 is the pumping unit, 3-1 is the molecular pump, 3-2 is the mechanical pump, 4 is the control module, 5 is the silicon wafer, A is the particle source, B is the gas source, and C is the vacuum gauge. Detailed Implementation

[0036] The embodiments of the present invention are described in detail below: These embodiments are implemented based on the technical solution of the present invention, and provide detailed implementation methods and specific operation processes. It should be noted that those skilled in the art can make several modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention.

[0037] Existing technologies for debris removal from EUV light sources typically suffer from unstable removal efficiency, high energy consumption, and insufficient adaptability. To address these issues, one embodiment of this invention provides a design method for an EUV light source debris removal device. This method employs a debris removal structure design based on a combination of an Eulerian-Lagrange coupling model and Bayesian optimization. This approach can accurately predict the trajectory of debris particles under vacuum or low-pressure environments and uses intelligent optimization techniques to globally optimize key parameters of the debris removal device (such as nozzle diameter, spray angle, cavity geometry, and pumping velocity). This achieves efficient debris removal, low-energy operation, and adaptability to multiple operating conditions, effectively improving the stability and lifespan of the EUV light source system.

[0038] Specifically, such as Figure 1 As shown, the design method of the extreme ultraviolet light source debris removal device provided in this embodiment may include:

[0039] S1. A gas flow field model is established using the Euler model in conjunction with the slip boundary to perform numerical simulation of the gas flow field of the background gas under vacuum or low-pressure environment.

[0040] S2, a particle motion model is established using the Lagrange model, and the particle motion trajectory is tracked and calculated;

[0041] S3, based on the gas flow field model and the particle motion model, performs coupled calculations on the gas flow field and particle motion trajectory to obtain the simulation results of particle motion distribution;

[0042] S4. Experimentally verify the simulation results of particle motion distribution to ensure that the simulation results of particle motion are consistent with the actual motion, and correct the model parameters based on the experimental verification results to obtain the numerical simulation results of particle removal efficiency.

[0043] S5 uses the numerical simulation results of particle removal efficiency as initial data, and based on the continuously updated particle removal efficiency under different design parameters, it uses the Bayesian optimization algorithm and the Gaussian process regression model as a surrogate model. With the optimization objectives of maximizing particle removal efficiency and minimizing energy consumption, it performs global optimization on the design parameters of the debris removal device, gradually approximating the optimal solution of parameters and thus obtaining the optimal solution of design parameters, which is used to guide the manufacturing and application of actual devices.

[0044] In some preferred embodiments, the above-mentioned S1, which uses an Eulerian model in conjunction with a slip boundary to establish a gas flow field model, may further include:

[0045] Gas flow field models, including:

[0046] The mass conservation equation is expressed as:

[0047]

[0048] In the formula, ρ is the fluid density. Let t be the velocity vector and t be the time. Represents the vector differential operator;

[0049] The equation for the conservation of momentum is expressed as:

[0050]

[0051] In the formula, p is the pressure, and τ is the stress tensor. For volume force, τ ijLet v be the viscous stress acting in direction j in direction i, μ be the dynamic viscosity, and v be the viscous stress. i v is the velocity component in the direction i of the fluid. j Let x be the velocity component in the direction j of the fluid. j Let x be the spatial coordinate of direction j. i Let δ be the spatial coordinate of direction i. ij The symbol for Kronecker;

[0052] The energy conservation equation is expressed as:

[0053]

[0054] In the formula, h is the specific enthalpy, k is the thermal conductivity, Φ is the viscous dissipation term, and T is the temperature;

[0055] Maxwell's slip condition formula is expressed as:

[0056]

[0057] In the formula, u s Let λ be the wall slip velocity, λ be the path of freedom, and α be the slip coefficient, which is taken as 1 for complete diffusion. j Let be the unit vector in direction j;

[0058] Gas flow field models are used to obtain the three-dimensional distribution of gas velocity, pressure, and temperature fields.

[0059] In some preferred embodiments, the numerical simulation described above, S1, may further include:

[0060] Simulation conditions include: pressure ranging from 0.01 to 100 Pa;

[0061] The types of simulated gases include: argon and helium;

[0062] Numerical simulation methods include: finite volume method, finite difference method and commercial CFD solver software.

[0063] In some preferred embodiments, S2 above, which uses a Lagrange model to establish a particle motion model, may further include:

[0064] The particle motion model is represented as:

[0065]

[0066] In the formula, v p Where v is the particle velocity, v is the fluid velocity, and F is the fluid velocity. D (vv P ρ is the drag force per unit mass of the particle. P Where ρ is the particle density, ρ is the ambient gas density, and g is the particle density.x For gravitational acceleration, F x Other forces; μ is the dynamic viscosity of the fluid, d P Where is the particle diameter, and Cc is the correction for particle wall velocity slip considering rarefied gas mechanics;

[0067] The particle motion model is used to obtain the motion path of particles in the gas flow field and the final exit position.

[0068] In some preferred embodiments, the particle motion model S2 described above may further include:

[0069] Particle diameter d P The range is 0.01–10 μm;

[0070] Particle velocity v P The range is 1 to 500 m / s.

[0071] In some preferred embodiments, S3 above, which couples the gas flow field and particle motion trajectory based on the gas flow field model and the particle motion model to obtain the particle motion distribution simulation results, may further include:

[0072] Using unidirectional coupling technology, the gas flow field and particle trajectory are coupled and calculated. By repeatedly calculating the coupling relationship between the gas flow field and the particle trajectory, the dynamic equilibrium state of their interaction force is gradually approximated, and finally, the particle motion distribution map in the gas flow field is output.

[0073] In some preferred embodiments, S4 above, which involves experimentally verifying the simulation results of particle motion distribution to ensure that the simulation results of particle motion are consistent with the actual motion, and correcting the model parameters based on the experimental verification results to obtain numerical simulation results of particle removal efficiency, may further include:

[0074] An experimental setup was used to verify the simulation results of particle motion distribution. In the setup, the EUV light source was in operation, and the vacuum level of the cavity was 0.01–10 Pa. A silicon wafer was placed at the exit position, and the particle deposition amount was analyzed using SEM imaging to obtain the experimental verification results.

[0075] The experimental verification results were compared with the particle motion distribution simulation results to correct the parameters of the gas flow field and particle motion model.

[0076] The numerical simulation results of particle removal efficiency were obtained by calculating the particle deposition ratio.

[0077] In some preferred embodiments, S5 above uses the numerical simulation results of particle removal efficiency as initial data, and based on the continuously updated particle removal efficiency under different design parameters, utilizes a Bayesian optimization algorithm, employs a Gaussian process regression model as a surrogate model, and takes maximizing particle removal efficiency and minimizing energy consumption as optimization objectives to globally optimize the design parameters of the debris removal device, gradually approaching the optimal solution of parameters to obtain the optimal solution of design parameters. It may further include:

[0078] Using the Bayesian optimization algorithm and employing a Gaussian process regression model as a surrogate model, the optimization objective function is established as follows:

[0079] J(θ) = α·η C -β·Ec

[0080] In the formula, θ is the set of design parameters, and η C Ec represents particulate removal efficiency, which is the maximization indicator; Ec represents energy consumption, which is the minimization indicator; and α and β are weighting coefficients.

[0081] The numerical simulation results of particle removal efficiency verified by experiments are used as the initial training dataset for Bayesian optimization to find the optimal design parameters. After multiple optimizations, the current optimal design parameter data is selected for numerical simulation to obtain the particle removal efficiency under the new parameter set and added to the training dataset for the next Bayesian optimization.

[0082] After setting the number of iterations, the optimal design parameters for the debris removal device are output to guide the actual device manufacturing and application.

[0083] In some preferred embodiments, the design parameter set includes: nozzle diameter, injection angle, cavity geometry, and pumping flow velocity; wherein:

[0084] The design range for nozzle diameter is 0.5–5 mm;

[0085] The design range for the spray angle is 0–60°;

[0086] The design range for the pumping rate is 0.01 to 10 Pa·L / s.

[0087] Based on the same inventive concept, an embodiment of the present invention also provides a design system for an extreme ultraviolet light source debris removal device.

[0088] Specifically, such as Figure 2 As shown, the design system for the extreme ultraviolet light source debris removal device provided in this embodiment may include:

[0089] The gas flow field simulation module uses the Euler model in conjunction with slip boundary to establish a gas flow field model and performs numerical simulation of the gas flow field of background gas under vacuum or low-pressure environment.

[0090] The particle motion simulation module uses the Lagrange model to establish a particle motion model and calculates the particle's motion trajectory.

[0091] The flow field-particle coupling module is based on the gas flow field model and the particle motion model. It performs coupled calculations on the gas flow field and the particle motion trajectory to obtain the simulation results of particle motion distribution.

[0092] The experimental verification module verifies the simulation results of particle motion distribution to ensure that the simulation results of particle motion are consistent with the actual motion. The model parameters are then corrected based on the experimental verification results to obtain numerical simulation results of particle removal efficiency.

[0093] The device parameter optimization module uses the numerical simulation results of particle removal efficiency as initial data. Based on the continuously updated particle removal efficiency under different design parameters, it uses a Bayesian optimization algorithm and a Gaussian process regression model as a surrogate model. With the optimization objectives of maximizing particle removal efficiency and minimizing energy consumption, it performs global optimization of the design parameters of the debris removal device, gradually approaching the optimal solution of parameters and thus obtaining the optimal solution of design parameters, which is used to guide the actual device manufacturing and application.

[0094] The implementation methods of each functional module constituting the system provided in the above embodiments of the present invention will be further described in detail below with reference to preferred embodiments. Its overall workflow is as follows: Figure 3 As shown.

[0095] In some preferred embodiments, the gas flow field simulation module described above is used for gas flow field modeling. Its implementation further includes:

[0096] The simulation is performed by a computer using an Euler model in conjunction with a slip boundary to obtain the gas velocity field, pressure field and temperature field distribution under vacuum or low pressure conditions.

[0097] The simulation conditions range from 0.01 to 100 Pa, and the gas types include argon and helium. Numerical methods such as finite volume method, finite difference method, and commercial CFD solver can all be used. The output data of this module, including the three-dimensional distribution of velocity field, pressure field, and temperature field, serves as the input for subsequent particle calculations.

[0098] mass conservation equation:

[0099] ρ: fluid density;

[0100] ν: velocity vector;

[0101] t: time.

[0102] Momentum conservation equation:

[0103] p: pressure;

[0104] τ: Stress tensor;

[0105] F: Volume force.

[0106] Energy conservation equation:

[0107] h: enthalpy;

[0108] k: Thermal conductivity;

[0109] Φ: Viscous dissipation term.

[0110] Maxwell slip condition:

[0111] u s Wall sliding speed;

[0112] λ: free path;

[0113] α: Slip coefficient (taken as 1 for complete diffusion);

[0114] Let be the unit vector in direction j.

[0115] In some preferred embodiments, the aforementioned particle motion simulation module is used to establish a particle motion model. Its implementation further includes:

[0116] The computer executes the particle tracking calculations using a Lagrange model. Particle diameters range from 0.01 to 10 μm, and velocities range from 1 to 500 m / s. Typical materials include tungsten, silicon, and metal fragments. This module obtains the particle's path and final exit position in the gas flow field. In this module, the particle motion model can be represented as:

[0117] Stokes equations:

[0118] F D (vv P ): The drag force per unit mass of the particle;

[0119] Cc: Correction for particle wall velocity slip considering rarefied gas mechanics;

[0120] v P Particle velocity;

[0121] v is the fluid velocity;

[0122] ρ: Ambient gas density;

[0123] ρ p Particle density;

[0124] g x Gravitational acceleration;

[0125] F x Other forces.

[0126] In some preferred embodiments, the aforementioned flow field particle coupling module is used for gas-solid coupling simulation. Its specific implementation further includes:

[0127] The gas flow field obtained from the gas flow field simulation module is coupled with the particle motion trajectory obtained from the particle motion simulation module for calculation. Since the number of particles is relatively sparse compared with gas molecules, one-way coupling (unidirectional flow field driving particles) is used. The numerical iteration is performed until fluid-structure equilibrium is reached, and the final output is a particle motion distribution map.

[0128] In some preferred embodiments, the above-mentioned experimental verification module is used for experimental verification. Its specific implementation further includes:

[0129] The simulation results were validated using an experimental setup to ensure consistency between the model predictions and actual particle movement. The validation results were used to correct simulation parameters and improve model accuracy. Experimental conditions included EUV light source operation and cavity vacuum of 0.01–10 Pa. Validation method involved placing a silicon wafer at the exit position and analyzing particle deposition using SEM imaging. Validation outputs included comparing simulated and measured particle distributions to correct model parameters.

[0130] In some preferred embodiments, the above-mentioned device parameter optimization module is used for parameter optimization. Its specific implementation further includes:

[0131] A computer executes a Bayesian optimization algorithm, using a Gaussian process regression model as a surrogate model to progressively approximate the optimal solution. Global optimization is performed on design parameters such as nozzle diameter, injection angle, cavity geometry, and pumping velocity. The range of design variables is: nozzle diameter 0.5–5 mm, injection angle 0–60°, and pumping velocity 0.01–10 Pa·L / s.

[0132] The objective function can be expressed as: J(θ)=α·η C -β·Ec;

[0133] θ: Set of design parameters;

[0134] η c Particle removal efficiency (maximization metric);

[0135] E cEnergy consumption (minimization metric);

[0136] α, β: weighting coefficients.

[0137] The numerical simulation results of particle removal efficiency verified by experiments are used as training data for Bayesian optimization to find the optimal parameters. After multiple optimizations, the current optimal design parameter data is selected for numerical simulation. After a set number of iterations (e.g., 10-30 times), the computer outputs the optimal design parameters of the debris removal device to guide the actual device manufacturing and application.

[0138] It should be noted that the steps in the method provided by the present invention can be implemented using the corresponding components in the system. Those skilled in the art can refer to the technical solution of the system to implement the steps of the method, and can also refer to the technical solution of the method to implement the composition of the system. That is, the embodiments in the system and the embodiments in the method can be understood as preferred examples of each other, which will not be elaborated here.

[0139] Based on the design method or design system provided in the above embodiments of the present invention, an embodiment of the present invention also provides an extreme ultraviolet light source debris removal device.

[0140] Specifically, such as Figure 4 As shown, the extreme ultraviolet light source debris removal device provided in this embodiment may include: a cavity 1, a jet structure 2, an extraction unit 3, and a control module 4; wherein:

[0141] Particle source A is directly connected to cavity 1 and enters cavity 1 through differential aperture;

[0142] The jet structure 2 is connected to the gas source B (such as argon) through a gas supply pipeline and is fixed in a set position in the cavity 1;

[0143] The extraction unit 3 is connected to the exhaust port of the cavity 1 to form a stable pressure difference drive;

[0144] The control module 4 is electrically connected to the jet structure 2 and the extraction unit 3 respectively, and obtains the optimal design parameters by running the design method or design system provided in the above embodiments of the present invention, and adjusts the jet flow rate, jet angle and extraction speed in real time.

[0145] In some preferred embodiments, cavity 1 is used to accommodate the background gas flow field and the debris particle transport channel, while realizing the effective transmission of EUV light; wherein, cavity 1 can adopt a cylindrical or cuboid structure, with a typical size range of 0.1 to 1m.

[0146] In some preferred embodiments, the jet structure 2 is fixed to the wall of the cavity 1, and its jet outlet diameter ranges from 0.5 to 5 mm. It is used to spray argon or other inert gases to form a directional airflow zone. Its jet angle can be adjusted within the range of 0 to 60° to guide debris particles along different paths.

[0147] In some preferred embodiments, the pumping unit 3 is mainly composed of a mechanical pump 3-2 and a molecular pump 3-1, and is connected to the outlet of the cavity 1. It can maintain stable pumping in a low pressure or vacuum environment of 0.01 to 10 Pa, forming a directional pressure difference drive.

[0148] In some preferred embodiments, the control module 4 includes a computer and a storage unit for executing the simulation model and its coupled calculations and optimization algorithms, and is electrically connected to the jet structure 2 and the extraction unit 3 respectively. By obtaining the optimal design parameters, the jet flow rate (0-500 sccm), jet angle and extraction rate are adjusted in real time to obtain the optimal debris removal efficiency and reduce energy consumption.

[0149] In some preferred embodiments, characterizing the debris removal efficiency by observing the silicon wafer 5 placed at the end of the device using a scanning electron microscope may further include:

[0150] The silicon wafer surface was scanned before and after the particle source and debris removal device were turned on. The number and size of particles were recorded, and the debris removal efficiency was calculated using the following method:

[0151]

[0152] In the formula, η is the particle removal efficiency, n0 is the number of particles per unit area under the removal conditions, and n1 is the number of particles per unit area under the removal conditions.

[0153] In this embodiment, the cavity defines the interaction space between the airflow and the particles, providing simulation and experimental conditions; the jet structure drives the particles to migrate in a directional manner through the interaction between the high-speed airflow and the particles; the extraction unit controls the cavity pressure to achieve the directional discharge of particles with the flow; the control module combines model simulation and optimization results to dynamically control the jet and extraction parameters to maximize debris removal efficiency and reduce energy consumption.

[0154] The extreme ultraviolet (EUV) light source debris removal device, its design method, and its design system provided in the above embodiments of the present invention perform coupled calculations of the Eulerian and Lagrange models by computer to numerically simulate the gas flow field and particle motion under low pressure or vacuum conditions, and verify and correct the simulation parameters in conjunction with experiments. Furthermore, a Bayesian optimization algorithm is introduced to globally optimize key structural design parameters such as nozzle diameter, injection angle, cavity geometry, and extraction rate, in order to achieve high debris removal efficiency and low energy consumption discharge. The control module runs the above models and optimization algorithms to dynamically adjust the jet and extraction parameters, thereby achieving efficient debris removal and stable control under complex working conditions.

[0155] Compared with existing technologies, this invention can efficiently realize the structural parameter design of a device for the directional discharge and controlled removal of debris particles in vacuum or low-pressure environments. It has the following technical advantages: The introduction of an Eulerian-Lagrange coupled model allows for simultaneous consideration of gas flow field distribution and particle trajectory, accurately predicting the transport path of particles within the cavity. It avoids the problem of "separation between airflow simulation and particle behavior" in traditional empirical design, achieving synergistic simulation of the gas-solid two phases, thus resulting in better directional particle migration and a removal efficiency improvement of over 20%. The introduction of Maxwell slip boundary conditions in the gas flow field modeling and the use of rarefied gas correction factors in the particle model accurately reflect the discontinuous effects of gas under low-pressure conditions, ensuring consistency between simulation results and the actual operating environment of EUV devices, thereby reducing the deviation between experiments and simulations. A Bayesian optimization algorithm is introduced, using debris removal efficiency and energy consumption as multi-objective optimization indicators. Compared with traditional manual parameter tuning or single-objective optimization, Bayesian optimization can quickly converge to the globally optimal design parameters with limited experimental data, thereby maintaining high removal efficiency while reducing gas consumption and pump power, achieving energy-saving effects. The combined design of the jet structure and the extraction unit in the device enables the jet gas and vacuum extraction to work synergistically. It can be flexibly adjusted according to different pressures and particle sizes, and is suitable for debris removal needs in various low-pressure environments such as EUV light sources, plasma sources, and vacuum devices.

[0156] Any matters not covered in the above embodiments of the present invention are well-known in the art.

[0157] The specific embodiments of the present invention have been described above. It should be understood that the present invention is not limited to the specific embodiments described above, and those skilled in the art can make various modifications or variations within the scope of the claims, which do not affect the essence of the present invention.

Claims

1. A design method for an extreme ultraviolet light source debris removal device, characterized in that, include: A gas flow field model was established using the Euler model in conjunction with slip boundary to perform numerical simulation of the gas flow field of background gas under vacuum or low-pressure conditions. A particle motion model is established using the Lagrange model, and the particle motion trajectory is tracked and calculated. Based on the gas flow field model and the particle motion model, the gas flow field and particle motion trajectory are coupled and calculated to obtain the simulation results of particle motion distribution; The simulation results of particle motion distribution are experimentally verified to ensure that the simulation results of particle motion are consistent with the actual motion. The model parameters are then corrected based on the experimental verification results to obtain numerical simulation results of particle removal efficiency. Using the numerical simulation results of particle removal efficiency as initial data, and based on the continuously updated particle removal efficiency under different design parameters, the design parameters of the debris removal device are globally optimized using a Bayesian optimization algorithm and a Gaussian process regression model as a surrogate model, with the optimization objectives of maximizing particle removal efficiency and minimizing energy consumption. The optimal solution of the design parameters is obtained to guide the manufacturing and application of the actual device.

2. The design method of the extreme ultraviolet light source debris removal device according to claim 1, characterized in that, The method of establishing a gas flow field model using the Euler model in conjunction with slip boundaries includes: The gas flow field model includes: The mass conservation equation is expressed as: In the formula, ρ is the fluid density. Let t be the velocity vector and t be the time. Represents the vector differential operator; The momentum conservation equation is expressed as: In the formula, p is the pressure, and τ is the stress tensor. For volume force, τ ij Let v be the viscous stress acting in direction j in direction i, μ be the dynamic viscosity, and v be the viscous stress. i v is the velocity component in the direction i of the fluid. j Let x be the velocity component in the direction j of the fluid. j Let x be the spatial coordinate of direction j. i Let δ be the spatial coordinate of direction i. ij The symbol for Kronecker; The energy conservation equation is expressed as: In the formula, h is the specific enthalpy, k is the thermal conductivity, Φ is the viscous dissipation term, and T is the temperature; Maxwell's slip condition formula is expressed as: In the formula, u s Let λ be the wall slip velocity, λ be the path of freedom, and α be the slip coefficient, which takes the value of 1 for complete diffusion. j Let be the unit vector in direction j; The gas flow field model is used to obtain the three-dimensional distribution of the gas velocity field, pressure field, and temperature field; The numerical simulation includes: Simulation conditions include: pressure ranging from 0.01 to 100 Pa; The types of simulated gases include: argon and helium; Numerical simulation methods include: finite volume method, finite difference method and commercial CFD solver software.

3. The design method of the extreme ultraviolet light source debris removal device according to claim 1, characterized in that, The method of establishing a particle motion model using the Lagrange model includes: The particle motion model is expressed as follows: In the formula, v P Where v is the particle velocity, v is the fluid velocity, and F is the fluid velocity. D (vv p ρ is the drag force per unit mass of the particle. P Where ρ is the particle density, ρ is the fluid density, and g is the particle density. x For gravitational acceleration, F x For other forces, μ is the dynamic viscosity of the fluid, and d P Where is the particle diameter, and Cc is the correction for particle wall velocity slip considering rarefied gas mechanics; The particle motion model is used to obtain the motion path and final exit position of the particles in the gas flow field. in: Particle diameter d p The range is 0.01–10 μm; The particle velocity v P The range is 1 to 500 m / s.

4. The design method of the extreme ultraviolet light source debris removal device according to claim 1, characterized in that, The process involves coupled calculations of the gas flow field model and the particle motion model to obtain simulation results of particle motion distribution, including: Using unidirectional coupling technology, the gas flow field and particle trajectory are coupled and calculated. By repeatedly calculating the coupling relationship between the gas flow field and the particle trajectory, the dynamic equilibrium state of their interaction force is gradually approximated, and finally, the particle motion distribution map in the gas flow field is output.

5. The design method of the extreme ultraviolet light source debris removal device according to claim 1, characterized in that, The experimental verification of the particle motion distribution simulation results ensures that the simulation results are consistent with the actual motion, and the model parameters are corrected based on the experimental verification results to obtain numerical simulation results of particle removal efficiency, including: The simulation results of particle motion distribution were experimentally verified using an experimental setup. In the setup, the EUV light source was in operation, and the cavity vacuum was 0.01–10 Pa. A silicon wafer was placed at the outlet, and the particle deposition amount was analyzed using SEM imaging to obtain the experimental verification results. The experimental verification results were compared with the particle motion distribution simulation results to correct the parameters of the gas flow field and particle motion model. The numerical simulation results of particle removal efficiency were obtained by calculating the particle deposition ratio.

6. The design method of the extreme ultraviolet light source debris removal device according to claim 1, characterized in that, The process involves using the numerical simulation results of the particle removal efficiency as initial data, and based on the continuously updated particle removal efficiency under different design parameters, employing a Bayesian optimization algorithm and a Gaussian process regression model as a surrogate model. The optimization objectives are to maximize particle removal efficiency and minimize energy consumption. This process globally optimizes the design parameters of the debris removal device to obtain the optimal solution, including: Using the Bayesian optimization algorithm and employing a Gaussian process regression model as a surrogate model, the optimization objective function is established as follows: J(θ)=α·η c -b·Ec In the formula, θ is the set of design parameters, and η C Ec represents particulate removal efficiency, which is the maximization indicator; Ec represents energy consumption, which is the minimization indicator; and α and β are weighting coefficients. The numerical simulation results of particle removal efficiency verified by experiments are used as the initial training dataset for Bayesian optimization to find the optimal design parameters. After multiple optimizations, the current optimal design parameter data is selected for numerical simulation to obtain the particle removal efficiency under the new parameter set and added to the training dataset for the next Bayesian optimization. After a set number of iterations, the optimal design parameters for the debris removal device are output to guide the actual device manufacturing and application. in: The set of design parameters includes: nozzle diameter, injection angle, cavity geometry, and extraction / exhaust velocity; wherein: The design range for the nozzle diameter is 0.5–5 mm; The design range for the spray angle is 0–60°. The design range for the pumping rate is 0.01 to 10 Pa·L / s.

7. A design system for an extreme ultraviolet light source debris removal device, characterized in that, include: The gas flow field simulation module uses the Euler model in conjunction with slip boundary to establish a gas flow field model and performs numerical simulation of the gas flow field of background gas under vacuum or low-pressure environment. The particle motion simulation module uses the Lagrange model to establish a particle motion model and calculates the particle's motion trajectory. The flow field-particle coupling module performs coupled calculations on the gas flow field and the particle motion model based on the gas flow field model and the particle motion model to obtain the simulation results of particle motion distribution. The experimental verification module verifies the simulation results of particle motion distribution to ensure that the simulation results of particle motion are consistent with the actual motion, and corrects the model parameters based on the experimental verification results to obtain numerical simulation results of particle removal efficiency. The device parameter optimization module, based on the continuously updated numerical simulation results of the gas flow field under different design parameters, uses the Bayesian optimization algorithm and the Gaussian process regression model as a surrogate model. With the optimization objectives of maximizing particle removal efficiency and minimizing energy consumption, it performs global optimization of the design parameters of the debris removal device to obtain the optimal solution of the design parameters, which is used to guide the actual device manufacturing and application.

8. An extreme ultraviolet light source debris removal device, characterized in that, include: The cavity, jet structure, extraction and exhaust unit, and control module; among which: The particle source is directly connected to the cavity; The jet structure is connected to the gas source via a gas supply pipeline and is fixed at a predetermined position in the cavity; The extraction unit is connected to the exhaust port of the cavity to form a stable pressure differential drive; The control module is electrically connected to the jet structure and the extraction unit respectively, and obtains the optimal design parameters by executing the design method of any one of claims 1-6 or running the design system of claim 7, and adjusts the jet flow rate, jet angle and extraction speed in real time.

9. The extreme ultraviolet light source debris removal device according to claim 8, characterized in that, Also includes: The cavity is used to accommodate the background gas flow field and the debris particle transport channel, while realizing the effective transmission of EUV light; wherein, the cavity adopts a cylindrical structure and the diameter of the cavity cross section ranges from 0.1 to 1m; The jet structure is fixed to the wall of the cavity, and its jet outlet diameter ranges from 0.5 to 5 mm. It is used to spray argon or other inert gases to form a directional airflow zone. Its jet angle can be adjusted within the range of 0 to 60° to guide debris particles along different paths. The pumping unit is mainly composed of a mechanical pump and a molecular pump, and is connected to the outlet of the cavity. It can maintain stable pumping in a low pressure or vacuum environment of 0.01 to 10 Pa, forming a directional pressure difference drive. The control module includes a computer and a storage unit, which are used to execute simulation models and their coupled calculations and optimization algorithms. It is electrically connected to the jet structure and the extraction unit respectively. By obtaining the optimal design parameters, it adjusts the jet flow rate, jet angle and extraction rate in real time, thereby obtaining the optimal debris removal efficiency and reducing energy consumption.

10. The extreme ultraviolet light source debris removal device according to claim 9, characterized in that, The debris removal efficiency was characterized by observing the silicon wafer placed at the end of the device using a scanning electron microscope, including: The silicon wafer surface was scanned before and after the particle source and debris removal device were turned on. The number and size of particles were recorded, and the debris removal efficiency was calculated using the following method: In the formula, η is the particle removal efficiency, n0 is the number of particles per unit area under the removal conditions, and n1 is the number of particles per unit area under the removal conditions.

Images