Dual ion source parallel beam collaborative irradiation device and electrostatic lens beam shape optimization method

Abstract

This invention relates to the field of ion beam polishing equipment technology, and more particularly to a dual-ion-source parallel beam synergistic irradiation device and an electrostatic lens beam shape optimization method. The device includes a first and a second Penning ion source arranged opposite each other; a multi-pole array electrostatic lens, each located at its exit end, its toroidal surface physically divided into independent sector plates; a composite detection module comprising a three-dimensional displacement grid probe and a feature capture grid; and a control unit that, based on the actual main beam current density field, the secondary characteristic current density field, and the comprehensive ionization coefficient of the working gas, constructs a modified Poisson equation with dynamic charge perturbation compensation, and solves inversely to minimize the spatial density deviation based on the target distribution, obtaining an asymmetric bias voltage array which is independently applied to the corresponding sector plates. In this invention, the asymmetric electric field is used to compensate for mechanical assembly tolerances, offsetting the electric field perturbation caused by secondary particles, significantly improving the polishing smoothness and process consistency of micro / nano samples.

Description

Technical Field

[0001] This invention relates to the field of ion beam polishing equipment technology, and in particular to a dual-ion-source parallel beam synergistic irradiation device and an electrostatic lens beam shape optimization method. Background Technology

[0002] As micro- and nanoscale fabrication technologies continue to evolve towards deeper levels and smaller dimensions, cutting-edge fields such as fundamental research in materials science, failure analysis of high-order semiconductor chips, and advanced packaging defect detection are experiencing a surge in demand for extremely high-precision, non-destructive surface microstructures. In these advanced applications, the feature size of the target processing area is constantly shrinking, and the material structure is becoming increasingly complex. Traditional machining or extensive chemical etching is no longer sufficient, placing extremely stringent requirements on the processing accuracy, surface non-destructiveness, and process stability of ion beam polishing equipment.

[0003] In conventional polishing processes, the ion beam emitted from a Penning ion source typically exhibits a divergent distribution, meaning the beam density is high in the central region and low in the surrounding areas. This pattern, when directly applied to target surface polishing, easily introduces microscopic scratches. While in practice, extending processing time or rotating the sample can mitigate scratch formation, this significantly sacrifices processing efficiency. To improve efficiency and surface smoothness, employing dual ion beams for coordinated irradiation from different directions has become an effective technique. However, in actual engineering assembly, limitations imposed by machining tolerances make it difficult to achieve high-precision axis overlap and center alignment between dual ion sources, leading to deviations in the effective areas of the two ion beams. Summary of the Invention

[0004] To overcome the above deficiencies, this invention provides a dual-ion-source parallel beam synergistic irradiation device and an electrostatic lens beam shape optimization method, aiming to improve the problem of secondary particle disturbance of the spatial electric field.

[0005] In a first aspect, the present invention provides the following technical solution: a dual-ion-source parallel beam synergistic irradiation device, comprising:

[0006] An ion emission module includes a first Penning ion source and a second Penning ion source arranged opposite to each other; a multi-pole array electrostatic lens is respectively disposed at the beam extraction end of the first Penning ion source and the second Penning ion source, and the toroidal surface of the multi-pole array electrostatic lens is physically divided into multiple independent sector plates.

[0007] The composite detection module includes a three-dimensional displacement grid probe set in the dual-beam ion intersection region and connected to a three-dimensional displacement transmission mechanism, and a feature capture grid suspended above the processing target surface.

[0008] The control unit is connected to the ion emission module, the multi-electrode array electrostatic lens, and the composite detection module, respectively. Based on the actual main beam current density field measured by the three-dimensional displacement grid probe, the secondary characteristic current density field measured by the characteristic capture grid, and the comprehensive ionization coefficient of the working gas, the control unit constructs a modified Poisson equation with dynamic charge perturbation compensation. Under the constraint of the modified Poisson equation, with the target beam distribution state as the reference, the spatial density deviation is calculated and solved in reverse to obtain the asymmetric bias voltage array, and the asymmetric bias voltage array is independently applied to the corresponding sector plates.

[0009] Preferably, the feature capture grid is a metal mesh structure with an independent charge discharge loop. The feature capture grid is configured to intercept secondary particles ejected by the backsplash of the processing target surface to form a total secondary return current, thereby generating the secondary feature current density field.

[0010] Preferably, when constructing the modified Poisson equation, the control unit converts the actual main beam current density field into the primary charge density; and combines the secondary characteristic current density field with the comprehensive ionization coefficient to generate the dynamic charge perturbation compensation, so as to reconstruct the actual space electric field boundary model contaminated by secondary particles.

[0011] Preferably, it includes a dynamic gas mixing unit, which is connected to the inlet of the first Penning ion source and the second Penning ion source. The dynamic gas mixing unit is configured to adjust the mass flow rate of the mixed gas based on a proportional-integral-derivative control strategy and output the comprehensive ionization coefficient to the control unit in real time.

[0012] Secondly, the present invention provides the following technical solution: a method for optimizing the beam shape of an electrostatic lens, comprising the following steps:

[0013] Step S1: Obtain the overall ionization coefficient of the introduced working gas;

[0014] Step S2: The actual main beam density field is acquired by scanning with the three-dimensional displacement grid probe;

[0015] Step S3: The secondary particles rebounding from the processing target surface are intercepted by the feature capture grid, and the secondary feature current density field is calculated and generated.

[0016] Step S4: Using the actual main beam current density field, the secondary characteristic current density field, and the comprehensive ionization coefficient as input parameters, construct a modified Poisson equation with dynamic charge perturbation compensation;

[0017] Step S5: Under the constraint of the modified Poisson equation, based on the set target beam distribution state, calculate the spatial density deviation minimization and perform inverse solution to obtain the asymmetric bias voltage array, and apply the asymmetric bias voltage array independently to the corresponding sector plates.

[0018] Preferably, step S2 specifically includes:

[0019] The three-dimensional displacement transmission mechanism drives the three-dimensional displacement grid probe to traverse and scan the set three-dimensional spatial grid nodes to read discrete main beam current signals;

[0020] The discrete main beam current signal is fitted and mapped using a spatial interpolation algorithm to construct the actual main beam density field representing a spatially continuous distribution.

[0021] Preferably, step S3 specifically includes:

[0022] Record the total secondary return current formed after the feature-capturing grid intercepts secondary particles;

[0023] The secondary characteristic current density field is calculated and generated based on the total secondary return current, the effective projected area of ​​the feature capture grid, and the weighting function characterizing the spatial attenuation distribution of secondary particles above the processing target surface.

[0024] Preferably, the physical logic of the modified Poisson equation constructed in step S4 satisfies:

[0025] The second derivative of the total potential in space is equal to the sum of the original charge density in space and the compensation for the dynamic charge disturbance, divided by the negative value of the vacuum permittivity.

[0026] The primary charge density is derived from the actual main beam current density field.

[0027] Preferably, step S5 specifically includes:

[0028] Calculate the sum of squares of the differences between the actual main beam density field and the target beam distribution state within the effective processing space;

[0029] With the goal of minimizing the sum of squared differences, the solution is iteratively obtained within the safe power supply range of the plate limit voltage. When the sum of squared differences converges and stabilizes, the current voltage distribution result is extracted and output as the asymmetric bias voltage array.

[0030] Preferably, during the polishing cycle, the control unit executes steps S2 to S5 in a fixed time interval, updating the asymmetric bias voltage array according to the real-time changing processing environment.

[0031] Once the deviation between the actual and target spatial beam density reaches the convergence standard and the beam shape adjustment is completed, the sample protection baffle set above the processing target surface is driven to move to a non-obstructed position to perform formal irradiation processing on the processing target surface.

[0032] The present invention has the following beneficial effects:

[0033] 1. This invention configures a composite detection module to simultaneously acquire the current density fields of the main beam and the secondary return current. The control unit constructs a modified Poisson equation by combining the gas comprehensive ionization coefficient, solves the asymmetric bias voltage array in reverse, and directly applies it to the lens sector plates. This physically cancels the spatial electric field disturbance caused by secondary particles, achieving accurate feedback and dynamic correction of the dual ion beam distribution pattern, and significantly improving the polishing smoothness of the sample.

[0034] 2. The dynamic gas mixing unit of the present invention utilizes a proportional-integral-derivative control strategy to regulate multiple working gases in a closed loop, overcoming the flow fluctuation problem caused by fluid delay and ensuring that the formula gas is quickly and uniformly mixed at the inlet. The unit outputs the comprehensive ionization coefficient to the control unit in real time to participate in electric field reconstruction, and connects the gas source control with the electrostatic lens adjustment at the underlying data level, ensuring the stability of the beam output from the source.

[0035] 3. The control unit scans and fits the beam density field in a fixed cycle within the polishing cycle. It takes minimizing the sum of squares of the difference between the actual distribution and the target distribution as the objective for iterative convergence. When the beam shape does not reach the convergence standard, the sample protection baffle will physically block the processing target surface, avoiding the erroneous etching of the sample caused by ion beam divergence during equipment startup or process switching, thus effectively ensuring the processing yield. Attached Figure Description

[0036] Figure 1 This is a schematic diagram of the module of the dual-ion-source parallel beam synergistic irradiation device proposed in this invention;

[0037] Figure 2 This is a schematic diagram of the method steps for optimizing the electrostatic lens beam shape proposed in this invention. Detailed Implementation

[0038] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0039] Example 1:

[0040] In a first embodiment of the present invention, the present invention provides a dual-ion-source parallel beam synergistic irradiation device, such as... Figure 1 As shown, it includes:

[0041] The ion emission module includes a first Penning ion source and a second Penning ion source arranged opposite to each other; a multi-pole array electrostatic lens is respectively disposed at the beam extraction end of the first Penning ion source and the second Penning ion source, and the toroidal surface of the multi-pole array electrostatic lens is physically divided into multiple independent sector plates.

[0042] The composite detection module includes a three-dimensional displacement grid probe set in the dual-beam ion intersection region and connected to a three-dimensional displacement transmission mechanism, and a feature capture grid suspended above the processing target surface.

[0043] The control unit is connected to the ion emission module, the multi-electrode array electrostatic lens, and the composite detection module, respectively. Based on the actual main beam current density field measured by the three-dimensional displacement grid probe, the secondary characteristic current density field measured by the characteristic capture grid, and the comprehensive ionization coefficient of the working gas, the control unit constructs a modified Poisson equation with dynamic charge perturbation compensation. Under the constraint of the modified Poisson equation, the spatial density deviation is minimized by calculating the inverse solution based on the target beam distribution state to obtain the asymmetric bias voltage array, and the asymmetric bias voltage array is independently applied to the corresponding sector plates.

[0044] Specifically, the ion emission module includes a first Penning ion source and a second Penning ion source, which are arranged at a set angle relative to each other in the vacuum chamber to emit two independent ion beams into a preset confluence space.

[0045] A multi-pole array electrostatic lens is coaxially disposed at the beam extraction ends of the first and second Penning ion sources, respectively. The annular conductive cross section of the multi-pole array electrostatic lens is physically divided into multiple independent sector plates by an insulating medium. Each sector plate is arranged in a circular array in space and is connected to an independent high-voltage power supply interface to receive different analog voltage levels, thereby generating a spatial asymmetric electric field with a specific potential gradient inside the lens.

[0046] The composite detection module comprises a three-dimensional displacement grid probe and a feature capture grid. The three-dimensional displacement grid probe is positioned at the intersection of the dual ion beams above the processing target surface, and its physical pins are connected to a three-dimensional displacement transmission mechanism. The three-dimensional displacement transmission mechanism receives drive pulses and drives the three-dimensional displacement grid probe to perform discrete traversal scanning on the set three-dimensional spatial grid nodes to acquire the main beam current signal at each coordinate point. The feature capture grid is suspended and fixed at a preset height section above the processing target surface by an insulating bracket, used to intercept the secondary particle stream generated by the physical recoil of the processing target surface.

[0047] The control unit establishes electrical connections with the power output terminal of the ion emission module, the input terminal of the multi-pole array electrostatic lens, and the signal output terminal of the composite detection module through the internal system data bus. In the system operation state, the analog-to-digital conversion module of the control unit synchronously acquires the main beam current signal measured by the three-dimensional displacement grid probe and the total secondary return current signal measured by the feature capture grid.

[0048] The control unit uses the physical quantities collected above to generate the actual main beam current density field and the secondary characteristic current density field through spatial interpolation and weighted attenuation calculation. The control unit combines the comprehensive ionization coefficient of the introduced working gas to construct a modified Poisson equation with dynamic charge disturbance compensation. This partial differential equation unifies the space charge effect of the main beam, the ionization interference of recoil secondary particles on the residual gas, and the thermal deformation factor of the electrode plate into the space electric field boundary constraint condition.

[0049] The control unit uses the modified Poisson equation as a mathematical constraint and the preset target beam distribution state matrix in the system's non-volatile memory as a reference input to calculate the sum of squares of spatial density deviations between the current actual main beam density field and the target beam distribution state. The control unit executes an inverse iterative solution algorithm with the goal of minimizing this spatial density deviation. When the sum of squares of deviations meets the preset convergence tolerance threshold, the control unit extracts the voltage allocation result of the current iteration step to obtain a set of asymmetric bias voltage arrays containing multiple discrete voltage values.

[0050] The control unit converts the asymmetric bias voltage array into multiple physical control signals, which are independently applied to the corresponding sector plates of the multi-pole array electrostatic lens via a high-voltage amplification circuit. To address the dynamic electric field distortion caused by secondary particle recoil ionization and the basic electrostatic field offset caused by mechanical assembly tolerances in traditional dual-beam synergistic irradiation, the aforementioned composite detection and asymmetric reverse compensation feedback mechanism actively constructs a spatially asymmetric compensation electrostatic field inside the multi-pole array electrostatic lens. The Coulomb force generated by this compensation electrostatic field can accurately counteract the dynamic charge disturbance and initial electric field deviation in the intersection region, limiting the trajectory divergence of the ion beam during transmission and intersection. This ensures that the relative standard deviation between the actual beam current distribution spatial density superimposed on the processing target surface and the preset target beam current distribution spatial density remains within the set convergence constant range, achieving quantitative control and physical conformal preservation of the dual ion beam morphology characteristics.

[0051] The feature capture grid is a metal mesh structure with an independent charge discharge loop. The feature capture grid is configured to intercept secondary particles ejected by the backflow sputtering of the processing target to form a total secondary return current, thereby generating a secondary feature current density field.

[0052] Specifically, in the physical process of dual-ion beam co-bombardment of the target surface, the target material undergoes a collisional cascade effect, releasing a swarm of secondary particles ejected by recoil sputtering into the space above the vacuum chamber. A feature-capturing grid is configured to physically intercept these upward-spreading secondary particles within its covered spatial projection area. The intercepted charged particles accumulate charge on the metal grid contact surface. Through a configured independent charge discharge circuit, the accumulated charge on the grid is conducted in real time to an external current measurement circuit in the form of a directional charge flow. After hardware filtering and analog-to-digital conversion, the total secondary return current is accumulated within a set control cycle.

[0053] To address the problem of local charge accumulation and electric field distortion caused by the disordered residence of secondary particles recoiled from the target surface in the processing space, the feature capture grid and its related mechanisms in this embodiment produce a specific physical compensation effect. The independent charge discharge circuit continuously discharges the intercepted charge, enabling the feature capture grid to maintain its electrostatic balance while continuously intercepting particles, thus avoiding the grid itself becoming a new source of stray electric field emission due to the accumulation of static charge.

[0054] After acquiring the total secondary return current, the control unit combines the effective projected area parameter of the feature capture grid with the secondary particle spatial decay weighting function to calculate and generate a secondary feature current density field data matrix in a three-dimensional spatial coordinate system. This execution step converts the discretely distributed target recoil secondary particles into continuous, quantifiable space charge density field values.

[0055] The generated secondary characteristic current density field is input into the computational memory of the control unit as a quantitative boundary condition, providing basic data support for the subsequent derivation of the dynamic charge disturbance compensation density and the construction of the modified Poisson equation. This numerical conversion mechanism directly eliminates the uncertainty of the secondary particle distribution, enabling the control unit to calculate the asymmetric bias voltage array to offset the disturbance. Ultimately, the relative standard deviation between the actual main beam spatial density and the target beam spatial density in the processing intersection area is stabilized within the preset convergence tolerance threshold constant.

[0056] When constructing the modified Poisson equation, the control unit transforms the actual main beam current density field into the primary charge density; and combines the secondary characteristic current density field with the comprehensive ionization coefficient to generate dynamic charge perturbation compensation, so as to reconstruct the actual space electric field boundary model contaminated by secondary particles.

[0057] Specifically, during the calculation of the modified Poisson equation, the control unit first uses the actual main beam density field data matrix cached by the system as the basic input parameter. The control unit then combines the physical property parameters of the monomer ions in the introduced working gas with the current electrical operating parameters of the equipment to execute a physical algebraic transformation program. This program directly transforms the actual main beam density field in the three-dimensional spatial coordinate system into a primary charge density data matrix characterizing the space charge distribution of the beam itself in an ideal vacuum.

[0058] The control unit retrieves the secondary characteristic current density field data matrix detected and generated by the feature capture grid. The control unit then performs a multiplication-addition operation on this secondary characteristic current density field with the pre-configured integrated ionization coefficient, the secondary particle recoil escape characteristic velocity constant, and the secondary electron emission rate calibration constant. Through this calculation logic, the control unit generates a dynamic charge perturbation compensation density data matrix distributed across the three-dimensional spatial grid nodes.

[0059] To address the spatial electric field distortion and local charge ionization interference caused by secondary particles recoiling from the target surface within the dual-ion beam processing region, the aforementioned transformation and combination mechanism produces a specific physical reconstruction effect. Traditional Poisson equation boundary models are constructed based solely on the primary charge density, without including the additional electric field vector parameters generated by recoil particles. The control unit reconstructs the actual spatial electric field boundary model, incorporating secondary particle ionization interference, by physically superimposing the primary charge density and dynamic charge perturbation compensation density calculated within the mathematical boundary model.

[0060] The reconstructed actual spatial electric field boundary model serves as the source term parameter input for the partial differential equation, enabling the final modified Poisson equation to quantitatively characterize the actual electrostatic field topology under dynamic processing conditions. This provides a computational benchmark, including environmental disturbance variables, for the subsequent control unit to inversely solve the asymmetric bias voltage array, eliminating the calculation deviation of the electrode voltage compensation caused by neglecting the space charge effect of secondary particles. Based on the boundary conditions output by this reconstructed model, the compensation electric field generated by the electrodes of the multi-electrode array electrostatic lens can accurately counteract the space charge disturbance in the intersection region, ensuring that the relative standard deviation of the spatial beam density is continuously limited within the system's preset convergence tolerance threshold constant.

[0061] It includes a dynamic gas mixing unit, which is connected to the inlet of the first Penning ion source and the second Penning ion source. It is configured to adjust the mass flow rate of the mixed gas based on a proportional-integral-derivative control strategy and output the comprehensive ionization coefficient to the control unit in real time.

[0062] Specifically, the dynamic gas mixing unit is equipped with a multi-channel mass flow controller, a fluid pressure sensing component, and an independent microprocessor. During the dual-beam ion synergistic irradiation process, the dynamic gas mixing unit is configured to regulate the mass flow rate of the mixed gas introduced into the Penning ion source based on a proportional-integral-derivative (PID) control strategy. Specifically, the microprocessor receives real-time pressure and velocity feedback signals from the fluid pressure sensing component, calculates the deviation data between the current fluid physical quantities and the set target values, and outputs an analog control voltage signal using a proportional-integral-derivative (PID) algorithm. This control voltage signal drives the solenoid valves inside the multi-channel mass flow controller to adjust their physical opening, thereby performing physical closed-loop regulation of the working gas flow rate.

[0063] During the synchronous process of executing flow closed-loop regulation, the microprocessor of the dynamic gas mixing unit calculates the comprehensive ionization coefficient under the current operating condition based on the currently set gas type ratio parameters and real-time flow and pressure sensor data, using the built-in thermodynamic and fluid dynamic mapping relationship. The digital data packet of this comprehensive ionization coefficient is output in real time through the internal data bus according to the system's set sampling clock cycle and overwritten to the system variable register of the control unit.

[0064] To address the technical problem of unstable plasma discharge state caused by working gas pressure fluctuations and physical drift in the ratio during continuous operation of traditional dual-ion source equipment, which leads to spatial distortion of the main beam current density distribution, the dynamic gas mixing unit in this embodiment provides a clear physical adjustment and data compensation mechanism. On the one hand, the physical flow regulation mechanism based on the proportional-integral-derivative control strategy limits the dynamic overshoot and steady-state control error of the mixed gas mass flow rate entering the Penning ion source, ensuring the physical stability of the ionization gas pressure inside the ion emission module and providing a fluid hardware foundation for the stable extraction of dual-beam ions.

[0065] The dynamic gas mixing unit outputs the comprehensive ionization coefficient to the control unit in real time, achieving temporal synchronization of the previously isolated external gas fluid state with the spatial electric field calculation model within the control unit. This real-time updated comprehensive ionization coefficient, as a dynamic boundary variable, directly participates in the system calculations of the control unit to construct the modified Poisson equation and derive the dynamic charge perturbation compensation density. This data transmission and real-time update mechanism eliminates the computational lag and model deviation caused by using static fixed gas parameters, ensuring that the control unit's inverse solution process for the asymmetric bias voltage array can accurately respond to the real-time gas ionization environment within the processing chamber. Ultimately, this mechanism enables the compensation electric field output by the multi-pole array electrostatic lens to precisely constrain ion trajectories, ensuring that the deviation between the actual spatial beam density and the target state in the intersection region strictly converges and remains within the system's preset tolerance threshold constant range.

[0066] Example 2:

[0067] In a first embodiment of the present invention, the present invention provides a method for optimizing the beam shape of an electrostatic lens, such as... Figure 2 As shown, it includes the following steps:

[0068] Step S1: Obtain the overall ionization coefficient of the introduced working gas;

[0069] Specifically, the control unit sends preset working gas formula parameters to the dynamic gas mixing unit. The internal control loop of the dynamic gas mixing unit operates based on a proportional-integral-derivative (PID) optimization control strategy. This unit receives gases from multiple individual gas storage tanks and continuously adjusts the opening of each pneumatic regulating valve through a current feedback loop to control the mass flow rate of different individual gases entering the dynamic mixer for physical mixing.

[0070] At the output of the dynamic mixer, the gas mass and concentration detection module collects the volume concentration ratio of each component gas in the working gas after mixing in real time at a fixed sampling frequency. Simultaneously, fluid pressure and temperature sensors installed at the inlets of the first and second Penning ion sources synchronously collect the real-time absolute pressure and real-time thermodynamic temperature of the working gas before it enters the ion source chamber.

[0071] The programmable logic controller (PLC) within the dynamic gas mixing unit receives the aforementioned concentration, pressure, and temperature data, and calculates the overall ionization coefficient of the working gas introduced under the current conditions based on the physical properties of each component gas. The formula for calculating the overall ionization coefficient is as follows:

[0072] ;

[0073] in, Indicates the overall ionization coefficient of the introduced working gas; This indicates the total number of individual gases contained in the mixed working gas; Indicates the first in the mixed working gas Real-time volume concentration ratio of the individual gases; Indicates the first The intrinsic ionization coefficient constant of a single gas under standard atmospheric pressure and standard thermodynamic temperature conditions; This indicates the real-time absolute pressure at the air inlet as measured by the sensor. This represents the standard atmospheric pressure value; This indicates the real-time thermodynamic temperature at the air inlet as measured by the sensor. It represents the standard thermodynamic temperature value.

[0074] After calculating the current comprehensive ionization coefficient using the above formula, the dynamic gas mixing unit sends this value to the control unit in real time via its internal communication bus. The control unit receives the comprehensive ionization coefficient and writes it into its internal system variable register, using it as input parameters for subsequent calculations of dynamic charge disturbance compensation and the construction of the modified Poisson equation. Whenever the dynamic gas mixing unit adjusts the pneumatic regulating valve, causing changes in gas flow or concentration, the programmable logic controller synchronously updates the calculated result of the comprehensive ionization coefficient and overwrites it to the control unit.

[0075] Step S2: Acquire the actual main beam density field by scanning with a three-dimensional displacement grid probe;

[0076] Step S2 specifically includes:

[0077] The driving three-dimensional displacement transmission mechanism drives the three-dimensional displacement grid probe to traverse and scan the set three-dimensional spatial grid nodes to read discrete main beam current signals;

[0078] By fitting and mapping the discrete main beam current signal using a spatial interpolation algorithm, an actual main beam density field representing the spatially continuous distribution pattern is constructed.

[0079] Specifically, the control unit outputs drive control commands containing spatial position coordinates to the three-dimensional displacement transmission mechanism. Upon receiving these commands, the three-dimensional displacement mesh probe traverses and scans the designated three-dimensional spatial mesh nodes. The designated sequence of three-dimensional spatial mesh node coordinates is generated by the control unit during the device initialization phase and includes discrete coordinate points in three orthogonal directions: the X-axis, Y-axis, and Z-axis.

[0080] When the three-dimensional displacement transmission mechanism moves and positions the three-dimensional displacement mesh probe to the target three-dimensional spatial mesh node, the conductive mesh inside the probe receives the ion charge flow at that spatial location. The current-to-voltage conversion circuit connected to the probe converts the intercepted ion current into a voltage signal. The analog-to-digital conversion module in the control unit reads this signal according to a preset sampling period, thereby obtaining the discrete main beam current signal at that spatial node. The control unit binds the amplitude of the read main beam current signal with the coordinates of the corresponding three-dimensional spatial mesh node and stores it in its internal storage array.

[0081] After completing the traversal scan of the set spatial region, the control unit calls the discrete main beam current signals of each grid node in the storage array, and uses a spatial interpolation algorithm to fit and map the discrete main beam current signals to construct an actual main beam density field that represents the continuous spatial distribution.

[0082] Specifically, the control unit uses a three-dimensional inverse distance weighted interpolation algorithm to calculate the main beam current density at any continuous coordinate point within the dual-beam ion intersection region. The formula for calculating the actual main beam current density field is as follows:

[0083] ;

[0084] ;

[0085] in, Represents the coordinates in three-dimensional space as The actual main beam current density at the location; This represents the total number of adjacent 3D spatial grid nodes involved in the interpolation calculation; This indicates that the three-dimensional displacement mesh probe is in the first... The amplitude of the discrete main beam current signal measured at each three-dimensional spatial grid node; This represents the effective cross-sectional receiving area of ​​the three-dimensional displacement mesh probe; Indicates the first Spatial geometric coordinates of each three-dimensional spatial grid node; Indicates the interpolation target point With the Three-dimensional spatial grid nodes The straight-line distance between them; This represents the preset distance decay weighting index.

[0086] The control unit converts the main beam detection results, which contain a finite number of discrete coordinate points, into continuous three-dimensional density scalar field data covering the space above the processing target surface. This actual main beam density field data is output and written to the system buffer of the control unit, serving as the basic spatial parameters for deriving the native charge density and subsequently constructing the modified Poisson equation.

[0087] Step S3: The secondary particles rebounding from the processing target surface are intercepted by the feature capture grid, and the secondary feature current density field is calculated and generated.

[0088] Step S3 specifically includes:

[0089] Record the total secondary return current formed after the secondary particles are captured by the feature-capturing grid;

[0090] The secondary characteristic current density field is calculated and generated based on the total secondary return current, the effective projected area of ​​the feature capture grid, and the weighting function that characterizes the spatial attenuation distribution of secondary particles above the processing target surface.

[0091] Specifically, during the dual-beam ion synergistic irradiation process of the target surface, ion bombardment causes physical recoil in the target material, generating discrete recoil sputtering secondary particles. A feature trapping grid is fixedly installed at a predetermined height above the target surface in a horizontal spatial section to physically intercept the secondary particles that disperse upwards. The feature trapping grid is woven from a conductive metal material, and its physical leads are connected to the signal input terminal of an external current measurement circuit via a low-impedance coaxial cable.

[0092] The current measurement circuit continuously acquires the induced charge flow generated by the feature capture grid due to the interception of secondary particles, linearly amplifies this physical quantity, and converts it into an analog voltage signal. The analog-to-digital conversion module in the control unit reads this analog voltage signal at a set sampling frequency, performs digital low-pass filtering to filter out background electromagnetic white noise, and then accumulates and calculates the total secondary return current in a single control cycle.

[0093] After obtaining the total secondary return current, the control unit retrieves the pre-configured feature capture grid geometric parameters, i.e., the effective projected area of ​​the feature capture grid, from the system's non-volatile memory. Simultaneously, the control unit calls a weighting function characterizing the spatial attenuation distribution of secondary particles above the processing target surface. To avoid excessively long formulas leading to incomplete document rendering, the control unit introduces a spatial linear distance parameter to reduce the dimensionality of the weighting function's calculation formula. The specific calculation formula is as follows:

[0094] ;

[0095] ;

[0096] ;

[0097] in, Represents the coordinates in three-dimensional space as The secondary characteristic current density at the location; This represents the total secondary return current amplitude obtained from the analog-to-digital conversion module. This represents the effective projected area of ​​the feature capture grid on the orthographic projection plane; The weighting function represents the attenuation distribution of secondary particles in the space above the processing target surface; This represents the spatial collision attenuation constant of secondary particles within the vacuum chamber. This constant is calculated and set by the device based on the current vacuum level of the chamber and the mean free path of the residual gas molecules. The geometrical spatial coordinates of the point where the two ion beams converge on the processing target surface are represented. Represents the geometric space coordinates of the target node to be calculated; This represents the absolute height difference between the target node to be calculated and the bombardment intersection center point in the direction perpendicular to the Z-axis; Indicates the target node to be calculated Center point of convergence with the bombardment The straight-line distance between them; It is the spatial solid angle direction factor for the emission of secondary particles according to the cosine law.

[0098] The control unit iterates through all the set discrete coordinate points within the three-dimensional spatial grid using control loop statements, mapping and binding the calculated secondary characteristic current density values ​​with the corresponding spatial coordinate points to generate a complete secondary characteristic current density field data matrix. This data matrix is ​​then transferred to the control unit's cache register to characterize the dynamic charge spatial distribution state of the recoil particle swarm above the processing area, serving as the basic input data for subsequent superposition onto the main beam current density field and establishment of the modified Poisson equation.

[0099] Step S4: Using the actual main beam current density field, the secondary characteristic current density field, and the comprehensive ionization coefficient as input parameters, construct a modified Poisson equation with dynamic charge perturbation compensation.

[0100] The physical logic of the modified Poisson equation constructed in step S4 satisfies:

[0101] The second derivative of the total potential in space is equal to the sum of the original charge density in space and the compensation for dynamic charge disturbances, divided by the negative value of the vacuum permittivity.

[0102] The primary charge density is derived from the actual main beam current density field.

[0103] Specifically, the control unit reads the actual main beam current density data matrix and the secondary characteristic current density data matrix of the three-dimensional spatial grid nodes in parallel from the system cache register. Simultaneously, the control unit retrieves the comprehensive ionization coefficient stored in the system variable register. For the intersection space of the dual ion beams within the vacuum chamber, the traditional Poisson equation only considers the static space charge under ideal conditions. To correct for electric field distortion during actual processing, the control unit introduces native charge density and dynamic charge perturbation compensation, constructing a modified Poisson equation.

[0104] To avoid excessively long partial differential equation formulas leading to incomplete display, the control unit introduces the Laplace operator to reduce the dimensionality of the three-dimensional second-order partial derivative of the total spatial potential. The physical logic of the modified Poisson equation constructed in step S4 satisfies: the result of the Laplace operator operation on the total spatial potential equals the sum of the original spatial charge density and the compensation for dynamic charge perturbation divided by the negative value of the vacuum permittivity. The mathematical expression of this modified Poisson equation is as follows:

[0105] ;

[0106] ;

[0107] ;

[0108] The native charge density is derived from the actual main beam current density field by the control unit, and its derivation formula is as follows:

[0109] ;

[0110] The dynamic charge disturbance compensation is calculated by the control unit based on the secondary characteristic current density field and the comprehensive ionization coefficient. The derived calculation formula is as follows:

[0111] ;

[0112] In the above formula, Represents the Laplace operator; It represents the sum of the second-order partial derivatives of the total potential in space in three-dimensional space; Represents the coordinates in three-dimensional space as The total potential of the space at that location; Represents the coordinates in three-dimensional space as The native charge density at that location; Represents the coordinates in three-dimensional space as Dynamic charge disturbance compensation density at the location; It represents the vacuum permittivity.

[0113] In the derivation formula of primary charge density, The coordinates generated by the interpolation fitting are: The actual main beam current density at the location; This represents the monomer ion mass constant of the introduced working gas; This represents the fundamental charge constant carried by a single working gas ion; This indicates the real-time ion acceleration voltage amplitude output from the high-voltage power supply module to the Penning ion source.

[0114] In the derivation formula for dynamic charge disturbance compensation, The coordinates captured and calculated by the feature capture grid are: The secondary characteristic current density at the location; This represents the characteristic velocity constant for secondary particle recoil escape set within the system; This represents the overall ionization coefficient of the introduced working gas, calculated in advance using fluid pressure and temperature data. This represents the preset secondary electron emission rate calibration constant within the control unit.

[0115] The control unit transforms discrete flow field detection data into a continuous electric field boundary condition model using the aforementioned partial differential equations and algebraic relationships. This modified Poisson equation, as the core mathematical model, is written into the central processing unit's memory to characterize the comprehensive spatial electric field distribution, including the space charge effect of the main beam and the ionization interference of secondary particles recoiling from the target surface. The establishment of this mathematical model provides a computational benchmark for the subsequent inverse solution of the asymmetric bias voltage array that minimizes spatial density deviation by the control unit.

[0116] Step S5: Under the constraint of the modified Poisson equation, based on the set target beam distribution state, calculate the spatial density deviation minimization and solve in reverse to obtain the asymmetric bias voltage array, and apply the asymmetric bias voltage array independently to the corresponding sector plates.

[0117] Step S5 specifically includes:

[0118] Calculate the sum of squares of the differences between the actual main beam density field and the target beam distribution state within the effective processing space;

[0119] With the goal of minimizing the sum of squared differences, the solution is iteratively obtained within the safe power supply range of the plate limit voltage. When the sum of squared differences converges and stabilizes, the current voltage distribution result is extracted and output as an asymmetric bias voltage array.

[0120] Specifically, the control unit retrieves the preset effective processing space coordinates and the target beam distribution state matrix from its internal non-volatile memory. The effective processing space consists of a set of grid nodes containing three-dimensional coordinate points. The control unit uses the modified Poisson equation as the boundary constraint condition for the electric field and ion trajectory, and establishes a numerical model for the mapping relationship between the plate voltage of the multi-pole array electrostatic lens and the main beam density in space.

[0121] In this numerical model, the control unit calculates the sum of squares of the differences between the actual main beam density field and the target beam distribution state within the effective processing space at the current iteration step. The control unit uses this sum of squares as the objective function, and the formula for calculating the sum of squares is as follows:

[0122] ;

[0123] in, This represents the sum of squares of the differences between the actual main beam density field and the target beam distribution state within the effective processing space; This represents the total number of discrete three-dimensional mesh nodes set within the effective processing space; Indicates the index number of the grid node; This represents the asymmetric bias voltage array used in the current iterative calculation, which includes... Each independent voltage value, i.e. , The total number of physically separated, independent sector plates; This indicates that under the constraints of the modified Poisson equation, when a voltage array is applied... At that time, the coordinates calculated by the numerical model are The actual main beam current density at the location; This indicates that the preset coordinates are... The target beam current density at the location.

[0124] The control unit is configured to make the sum of squares of the above differences The optimization objective is to minimize the voltage array, and the gradient descent algorithm or the finite element adjoint state method is used for iterative solution. During the iterative solution process, the control unit controls the voltage array. Each voltage component is range-limited to ensure it fluctuates within the safe supply range of the plate's limiting voltage. Simultaneously, the control unit evaluates the convergence status of the sum of squared differences after each iteration. The relevant constraints and convergence criteria are as follows:

[0125] ;

[0126] ;

[0127] in, Represents an asymmetric bias voltage array The corresponding number in the middle The bias voltage value of each sector plate. The value range is 1 to Positive integers; This indicates the system's preset lower limit for the safe power supply voltage to the electrode plates; This indicates the system's preset upper limit for the safe power supply voltage to the electrode plates; This indicates the current iteration count; Indicates the first The sum of squared differences calculated in each iteration; Indicates the first The sum of squared differences calculated in each iteration; This represents the system's preset convergence tolerance threshold constant.

[0128] When the convergence tolerance constraint condition is met consecutively, i.e. the sum of squared differences converges and stabilizes, the control unit terminates the iterative calculation program, extracts and outputs the voltage allocation result corresponding to the current iteration step. This serves as the final asymmetric bias voltage array.

[0129] Subsequently, the control unit sends the digital signal of the asymmetric bias voltage array to a multi-channel digital-to-analog converter (DAC) via its internal data bus. The DAC converts this signal into corresponding multi-channel analog voltage control signals, which are then transmitted to independent multi-channel high-voltage amplifiers. The high-voltage amplifiers output the actual high-voltage level according to the set amplification factor and apply it independently to each sector plate on the multi-electrode array electrostatic lens via high-voltage shielded cables. By applying independent and asymmetric voltages to each sector plate, an asymmetric compensation electric field with a specific gradient is formed within the lens's internal space to counteract spatial electric field disturbances caused by mechanical assembly tolerances and secondary particles.

[0130] During the polishing cycle, the control unit executes steps S2 to S5 in a fixed time interval, updating the asymmetric bias voltage array according to the real-time changing processing environment.

[0131] Once the deviation between the actual and target spatial beam density reaches the convergence standard and the beam shape adjustment is completed, the sample protection baffle set above the processing target surface is driven to move to an unobstructed position to perform formal irradiation processing on the processing target surface.

[0132] Specifically, the control unit is equipped with a clock signal generator that outputs fixed-period trigger pulses as the system's time cycle. At the rising edge of each time cycle, the control unit triggers the three-dimensional displacement transmission mechanism to perform a new round of three-dimensional spatial mesh scanning and simultaneously reads the current signal of the feature capture grid acquired by the multi-channel digital-to-analog converter circuit. By sequentially executing the logical operations in steps S2 to S4, the control unit reconstructs the modified Poisson equation, which includes the latest dynamic charge perturbation compensation. This real-time updated partial differential equation is used to characterize the changes in the spatial physical electric field within the vacuum chamber caused by the thermal expansion deformation of the electrode plates, fluctuations in the working gas pressure, and transient changes in the secondary particle density.

[0133] Based on the updated modified Poisson equation within each time cycle, the control unit repeatedly executes the inverse numerical solution process in step S5. By comparing the spatial density deviation state of the current calculation cycle, the control unit dynamically calculates the new voltage distribution result and outputs the digital quantity of the asymmetric bias voltage array to a multi-channel digital-to-analog converter. This, in turn, updates the actual physical applied voltage on each independent sector plate of the multi-electrode array electrostatic lens through a multi-channel high-voltage amplifier. The update frequency of this cyclic execution mechanism is set by a system-preset sampling period constant.

[0134] Before formally performing irradiation processing on the target surface, the control unit needs to continuously monitor and determine the spatial density deviation within the effective processing space. The control unit compares the actual main beam density field data matrix, obtained through actual measurement and spatial interpolation, with the system's preset target beam distribution state data matrix. To avoid excessively long nested formulas leading to incomplete document rendering and layout, the control unit establishes the relative standard deviation of the spatial beam density and the convergence judgment logic step by step. The physical judgment logic and calculation formula for the actual spatial beam density deviation from the target reaching the convergence standard are as follows:

[0135] ;

[0136] ;

[0137] in, A measure representing the relative standard deviation of the space beam density; This represents the total number of discrete three-dimensional mesh nodes set within the effective processing space; This represents the index number of the grid node, with a value ranging from 1 to... Positive integers; This indicates that, under the action of the asymmetric bias voltage array in the current cycle, the coordinates obtained after actual measurement and spatial interpolation by the three-dimensional displacement grid probe are: The actual beam current density at the location; This indicates that the coordinates are preset. The target beam current density at the location; This represents the convergence tolerance threshold constant pre-configured in the system's non-volatile memory.

[0138] The control unit executes the above mathematical formula for determination within each time cycle. When the system logic determines... The value is continuously less than or equal to within the set time window. At this time, the control unit records the status flags to confirm that the deviation between the actual and target spatial beam density has reached the convergence criterion, meaning that the beam shape adjustment process has been completed under the current working environment. After determining convergence, the control unit uses external... The interface sends a position switching command signal to the sample protection baffle drive mechanism.

[0139] The sample protection baffle drive mechanism includes a servo motor and a linear guide transmission assembly. Upon receiving a position switching command signal, the servo motor operates, driving the sample protection baffle, made of metal, to perform a linear horizontal displacement via the linear guide, moving it from a shielded position covering the processing target surface to an unshielded position deviating from the main beam emission path. When the photoelectric limit switch within the system detects that the sample protection baffle has reached the limit physical coordinates of the unshielded position, it sends a closed electrical signal back to the control unit. Upon receiving this signal, the control unit directs the ion beam downwards to bombard the processing target surface, and the device officially begins co-irradiation processing of the target surface.

[0140] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A dual-ion-source parallel beam synergistic irradiation device, characterized in that, include: An ion emission module includes a first Penning ion source and a second Penning ion source arranged opposite to each other; a multi-pole array electrostatic lens is respectively disposed at the beam extraction end of the first Penning ion source and the second Penning ion source, and the toroidal surface of the multi-pole array electrostatic lens is physically divided into multiple independent sector plates. The composite detection module includes a three-dimensional displacement grid probe set in the dual-beam ion intersection region and connected to a three-dimensional displacement transmission mechanism, and a feature capture grid suspended above the processing target surface. The control unit is connected to the ion emission module, the multi-electrode array electrostatic lens, and the composite detection module, respectively. Based on the actual main beam current density field measured by the three-dimensional displacement grid probe, the secondary characteristic current density field measured by the characteristic capture grid, and the comprehensive ionization coefficient of the working gas, the control unit constructs a modified Poisson equation with dynamic charge perturbation compensation. Under the constraint of the modified Poisson equation, with the target beam distribution state as the reference, the spatial density deviation is calculated and solved in reverse to obtain the asymmetric bias voltage array, and the asymmetric bias voltage array is independently applied to the corresponding sector plates.

2. The dual-ion-source parallel beam synergistic irradiation device according to claim 1, characterized in that, The feature capture grid is a metal mesh structure with an independent charge discharge loop. The feature capture grid is configured to intercept secondary particles ejected by the backsplash of the processing target surface to form a total secondary return current, thereby generating the secondary feature current density field.

3. The dual-ion-source parallel beam synergistic irradiation device according to claim 2, characterized in that, When constructing the modified Poisson equation, the control unit transforms the actual main beam current density field into the primary charge density; and combines the secondary characteristic current density field with the comprehensive ionization coefficient to generate the dynamic charge perturbation compensation, so as to reconstruct the actual space electric field boundary model contaminated by secondary particles.

4. The dual-ion-source parallel beam synergistic irradiation device according to claim 1, characterized in that, It includes a dynamic gas mixing unit, which is connected to the inlet of the first Penning ion source and the second Penning ion source. The dynamic gas mixing unit is configured to adjust the mass flow rate of the mixed gas based on a proportional-integral-derivative control strategy and output the comprehensive ionization coefficient to the control unit in real time.

5. An electrostatic lens beam shape optimization method, applied to a dual-ion-source parallel beam synergistic irradiation device as described in any one of claims 1 to 4, characterized in that, Includes the following steps: Step S1: Obtain the overall ionization coefficient of the introduced working gas; Step S2: The actual main beam density field is acquired by scanning with the three-dimensional displacement grid probe; Step S3: The secondary particles rebounding from the processing target surface are intercepted by the feature capture grid, and the secondary feature current density field is calculated and generated. Step S4: Using the actual main beam current density field, the secondary characteristic current density field, and the comprehensive ionization coefficient as input parameters, construct a modified Poisson equation with dynamic charge perturbation compensation; Step S5: Under the constraint of the modified Poisson equation, based on the set target beam distribution state, calculate the spatial density deviation minimization and perform inverse solution to obtain the asymmetric bias voltage array, and apply the asymmetric bias voltage array independently to the corresponding sector plates.

6. The electrostatic lens beam shape optimization method according to claim 5, characterized in that, Step S2 specifically includes: The three-dimensional displacement transmission mechanism drives the three-dimensional displacement grid probe to traverse and scan the set three-dimensional spatial grid nodes to read discrete main beam current signals; The discrete main beam current signal is fitted and mapped using a spatial interpolation algorithm to construct the actual main beam density field representing a spatially continuous distribution.

7. The electrostatic lens beam shape optimization method according to claim 5, characterized in that, Step S3 specifically includes: Record the total secondary return current formed after the feature-capturing grid intercepts secondary particles; The secondary characteristic current density field is calculated and generated based on the total secondary return current, the effective projected area of ​​the feature capture grid, and the weighting function characterizing the spatial attenuation distribution of secondary particles above the processing target surface.

8. The electrostatic lens beam shape optimization method according to claim 5, characterized in that, The physical logic of the modified Poisson equation constructed in step S4 satisfies: The second derivative of the total potential in space is equal to the sum of the original charge density in space and the compensation for the dynamic charge disturbance, divided by the negative value of the vacuum permittivity. The primary charge density is derived from the actual main beam current density field.

9. The electrostatic lens beam shape optimization method according to claim 5, characterized in that, Step S5 specifically includes: Calculate the sum of squares of the differences between the actual main beam density field and the target beam distribution state within the effective processing space; With the goal of minimizing the sum of squared differences, the solution is iteratively obtained within the safe power supply range of the plate limit voltage. When the sum of squared differences converges and stabilizes, the current voltage distribution result is extracted and output as the asymmetric bias voltage array.

10. The electrostatic lens beam shape optimization method according to claim 5, characterized in that, During the polishing cycle, the control unit executes steps S2 to S5 in a fixed time interval, and updates the asymmetric bias voltage array according to the real-time changing processing environment. Once the deviation between the actual and target spatial beam density reaches the convergence standard and the beam shape adjustment is completed, the sample protection baffle set above the processing target surface is driven to move to a non-obstructed position to perform formal irradiation processing on the processing target surface.

Images