Ion beam energy adaptive regulation method and system for fib device

Abstract

The application provides an ion beam energy adaptive regulation method and system for a FIB device, which comprises the following steps: obtaining an energy distribution parameter of an ion beam in a focusing area; determining a target energy regulation range according to the energy distribution parameter and a preset ion beam energy reference model; generating a multi-stage energy regulation instruction based on the target energy regulation range; controlling an output parameter of an ion beam energy regulator through the multi-stage energy regulation instruction, so that the current energy distribution of the ion beam enters the target energy regulation range; detecting energy fluctuation data of the adjusted ion beam on a sample surface in real time; and adjusting feedback control parameters of the ion beam energy regulator according to the energy fluctuation data and a preset adaptive compensation algorithm, so that the energy distribution of the ion beam is maintained in the target energy regulation range. The application can achieve better processing surface quality and process consistency in industrial processing.

Description

Technical Field

[0001] This invention relates to the fields of data processing and intelligent control technology, and in particular to an adaptive control method and system for ion beam energy in FIB equipment. Background Technology

[0002] Ion beam energy control is a core technology for achieving nanoscale processing precision in focused ion beam (FIB) equipment. Existing methods typically rely on static threshold control of energy parameters (such as beam current or magnetic field strength), using a fixed adjustment model to linearly regulate the ion beam energy. However, such methods are prone to decreased energy distribution uniformity and focus point drift, especially in long-term, high-precision processing tasks, which can lead to increased surface roughness and structural distortion. Furthermore, existing technologies cannot dynamically optimize the adjustment strategy based on the nonlinear relationship between energy decay gradient and diffusion rate, resulting in adjustment lag and overshoot oscillations. Moreover, frequent interruptions for manual calibration are required when environmental parameters change abruptly or hardware performance fluctuates, severely limiting processing efficiency and process consistency. Summary of the Invention

[0003] In view of this, embodiments of the present invention provide a method and system for adaptive control of ion beam energy in FIB equipment. The technical solution of the present invention is implemented as follows:

[0004] On one hand, embodiments of the present invention provide an adaptive energy control method for an ion beam in a FIB device, comprising the following steps: acquiring energy distribution parameters of the ion beam in the focusing region, the energy distribution parameters including ion beam current intensity, energy attenuation gradient, and spatial distribution characteristics; determining a target energy adjustment range based on the energy distribution parameters and a preset ion beam energy reference model, the target energy adjustment range including an allowable peak energy deviation threshold and a minimum energy gradient change rate; generating multi-level energy adjustment commands based on the target energy adjustment range, controlling the output parameters of the ion beam energy regulator through the multi-level energy adjustment commands to bring the current energy distribution of the ion beam into the target energy adjustment range; real-time detection of energy fluctuation data of the adjusted ion beam on the sample surface, the energy fluctuation data including instantaneous energy offset and focusing point energy diffusion coefficient; and adjusting the feedback control parameters of the ion beam energy regulator based on the energy fluctuation data and a preset adaptive compensation algorithm to maintain the energy distribution of the ion beam within the target energy adjustment range.

[0005] On the other hand, embodiments of the present invention provide a computer system including a memory and a processor, wherein the memory stores a computer program that can run on the processor, and the processor executes the program to implement the steps in the above-described method.

[0006] The ion beam energy adaptive control method for FIB equipment provided by this invention acquires the energy distribution parameters of the ion beam in the focusing region in real time and dynamically determines the target energy adjustment range based on the ion beam energy reference model. It generates multi-level energy adjustment commands to control the output parameters of the regulator, detects energy fluctuation data after adjustment in real time, and dynamically adjusts the feedback control parameters through an adaptive compensation algorithm. This enables a multi-dimensional synergistic adjustment mechanism of ion beam intensity, energy attenuation gradient, and spatial distribution characteristics, effectively suppressing random fluctuations and systematic drift of ion beam energy in dynamic working environments, and significantly improving focusing accuracy and energy distribution uniformity. By using the product of the energy attenuation gradient and the diffusion radius as the calculation basis for focusing accuracy compensation, it achieves a balance between spatial distribution characteristics and temporal attenuation characteristics. Coupled control ensures strict synchronization between the magnetic field gradient compensation command and the pulse timing adjustment signal, resolving the timing conflict problem of multi-parameter adjustment in traditional methods. Furthermore, based on the nonlinear feedback control cycle compression mechanism and the adaptive compensation algorithm with exponential decay rules, the adjustment step size and control frequency can be dynamically adjusted according to the error intensity, avoiding overshoot oscillations while achieving rapid convergence, thus ensuring system stability in high-precision machining scenarios. In addition, through intelligent matching of environmental parameter snapshots and adjustment performance fingerprints in the historical adjustment database, parameter configuration can be quickly reused under complex working conditions, significantly shortening the adjustment time of repetitive tasks and enhancing the equipment's adaptability to different vacuum environments and ion source decay states. This results in better surface quality and process consistency in applications such as etching and 3D reconstruction.

[0007] It should be understood that the above general description and the following detailed description are merely exemplary and explanatory, and are not intended to limit the technical solutions of the present invention. Attached Figure Description

[0008] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present invention and, together with the specification, serve to explain the technical solutions of the present invention.

[0009] Figure 1 This is a schematic diagram illustrating the implementation process of an adaptive control method for ion beam energy in a FIB device, provided by an embodiment of the present invention.

[0010] Figure 2 This is a schematic diagram of the hardware entity of a computer system provided in an embodiment of the present invention. Detailed Implementation

[0011] To make the objectives, technical solutions, and advantages of the present invention clearer, the technical solutions of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. The described embodiments should not be regarded as limitations on 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.

[0012] In the following description, references to "some embodiments" describe a subset of all possible embodiments; however, it is understood that "some embodiments" may be the same or different subsets of all possible embodiments and may be combined with each other without conflict. The terms "first / second / third" are used merely to distinguish similar objects and do not represent a specific ordering of objects. It is understood that "first / second / third" may be interchanged in a specific order or sequence where permissible, so that the embodiments of the invention described herein can be implemented in an order other than that illustrated or described herein.

[0013] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein is for descriptive purposes only and is not intended to limit the scope of the invention.

[0014] This invention provides a method for adaptive control of ion beam energy in a FIB (Fiber Injection Block) device, which can be executed by a processor of a computer system. The computer system can refer to a computer system embedded in or connected to the FIB device.

[0015] Figure 1 This is a schematic diagram illustrating the implementation process of an adaptive ion beam energy control method for FIB equipment provided in an embodiment of the present invention, as shown below. Figure 1 As shown, the method includes:

[0016] Step S100: Obtain the energy distribution parameters of the ion beam in the focusing region. The energy distribution parameters include the ion beam current intensity, energy attenuation gradient, and spatial distribution characteristics.

[0017] In this step, the ion beam current intensity refers to the number of ions passing through a certain cross-section per unit time. It reflects the energy carrying capacity of the ion beam; the greater the ion beam current intensity, the more energy is transferred in the same amount of time. The energy attenuation gradient refers to the rate attenuation of ion beam energy with distance or time during propagation. It reflects the stability and uniformity of ion beam energy distribution, and the magnitude of the energy attenuation gradient affects the effect of the ion beam in the focusing region. Spatial distribution characteristics describe the distribution of ion beam energy in the three-dimensional space of the focusing region, including the central location of energy concentration and the range of energy diffusion, facilitating precise control of the ion beam's effective location and range.

[0018] To obtain these energy distribution parameters, the specific steps include:

[0019] Step S110: The original energy signals of the ion beam at different angles in the focusing area are synchronously acquired by a ring-shaped multi-axis energy detection array. Each detection unit of the multi-axis energy detection array corresponds to a preset radial detection angle, and the signal acquisition areas of adjacent detection units partially overlap.

[0020] A multi-axis energy detection array is a device composed of multiple detection units arranged in a ring, capable of detecting the ion beam in a focused region from different angles. Each detection unit has a preset radial detection angle, allowing it to acquire signals from the ion beam in a specific direction. The signal acquisition areas of adjacent detection units partially overlap to ensure comprehensive and accurate acquisition of the energy signal in the focused region, avoiding blind spots. For example, when the energy distribution of the ion beam in the focused region is uneven, the overlapping acquisition areas ensure that energy signals at all locations can be detected.

[0021] Step S120: Perform time-domain denoising processing on the original energy signal to eliminate high-frequency electromagnetic interference components and generate a denoised energy waveform cluster. The denoised energy waveform cluster contains waveform data and spatial position labels corresponding to each detection unit.

[0022] The raw energy signal may be affected by high-frequency electromagnetic interference during acquisition, which can make the signal inaccurate. Therefore, time-domain denoising is necessary. Time-domain denoising is a method of filtering the signal in the time domain, which can identify and eliminate high-frequency electromagnetic interference components, thereby obtaining a cleaner energy signal. The denoised signal generates a denoised energy waveform cluster, where each waveform data corresponds to a detection unit and also contains a spatial location label for that detection unit. These labels are used to determine the specific location of each waveform data in the focusing area. For example, by performing a Fast Fourier Transform on the acquired raw energy signal to convert it to the frequency domain, then removing the high-frequency interference components, and finally using an Inverse Fourier Transform to convert the signal back to the time domain, the denoised energy waveform cluster can be obtained.

[0023] Step S130: Select the reference waveform with the largest energy integral value among all detection units.

[0024] The energy integral value refers to the cumulative value of the energy signal over a period of time, reflecting the total energy of the ion beam received by the detector unit. The reference waveform with the largest energy integral value is selected because it represents the most concentrated energy portion of the focusing region, and subsequent parameter calculations will be based on this reference waveform. For example, the energy signal of each detector unit is integrated, the magnitudes of the integral values ​​are compared, and the waveform with the largest integral value is selected as the reference waveform.

[0025] Step S140: Extract the energy sustaining interval from the steady-state phase of the reference waveform, and calculate the average value of the waveform peaks within the energy sustaining interval as the ion beam intensity.

[0026] The steady-state phase of a reference waveform refers to a period in which the waveform remains relatively stable, during which the energy output of the ion beam is relatively stable. The energy sustaining interval is selected to perform parameter calculations within this stable phase, resulting in a more accurate ion beam intensity. By calculating the average value of the waveform peaks within the energy sustaining interval, a value representing the ion beam intensity can be obtained. For example, in a reference waveform, a relatively stable time period is defined as the energy sustaining interval. The peak values ​​within this interval are summed, and then divided by the number of peaks; the resulting average value is the ion beam intensity.

[0027] Step S150: Divide the attenuation phase of the reference waveform into multiple sampling windows with equal time intervals.

[0028] The decay phase of the reference waveform refers to the stage where the ion beam energy begins to gradually decrease. Dividing the waveform into equally timed sampling windows allows for detailed analysis of energy changes during the decay phase. By calculating the rate of energy change within different sampling windows, information about the energy decay gradient can be obtained. For example, based on a preset time interval, the decay phase of the reference waveform can be divided into multiple sampling windows of equal length, each containing multiple energy data points.

[0029] Step S160: Calculate the rate of decrease of the energy peak within each sampling window, and take the weighted average of the rate of decrease of all sampling windows as the energy decay gradient.

[0030] The rate of decrease of the energy peak within each sampling window reflects the decay of the ion beam energy during that time period. By calculating the weighted average of the decrease rates across all sampling windows, a comprehensive energy decay gradient can be obtained, which more accurately describes the trend of ion beam energy change throughout the decay phase. For example, for each sampling window, the difference between the energy peaks at the start and end of the window is calculated and then divided by the window duration to obtain the rate of decrease of the energy peak within that window. Then, different weights are assigned to each sampling window based on its importance or representativeness, and a weighted average of all decrease rates is calculated to obtain the energy decay gradient.

[0031] Step S170: Perform spatial interpolation calculations between the spatial location labels of each detection unit and the corresponding waveform energy integral values ​​to generate a three-dimensional energy density surface covering the focused area.

[0032] Spatial interpolation is a method for estimating information about unknown points based on information from known points. In this step, the spatial location labels of each detector unit and the energy integral value of the corresponding waveform are used as known points. A spatial interpolation algorithm is then used to calculate the energy density at other locations within the focused area, generating a three-dimensional energy density surface covering the entire focused area. This surface visually displays the spatial distribution of ion beam energy within the focused area. For example, using the Kriging interpolation algorithm, energy density is estimated at other locations within the focused area based on the spatial location and energy integral value of each detector unit, ultimately generating a three-dimensional energy density surface. The Kriging interpolation algorithm is a statistically based spatial interpolation method that uses the spatial location and corresponding energy integral value of each detector unit as the basic data when generating the three-dimensional energy density surface. The spatial correlation of these known data points is analyzed, that is, the relationship between the spatial distance between data points and the difference in energy integral values ​​is determined. Based on this correlation, a semi-variogram model is constructed, which describes the spatial variation of energy. Then, for unknown locations within the focused area where energy density needs to be estimated, the Kriging algorithm uses a weighted combination of known data points for estimation. The weights are determined based on the previously constructed semi-variogram model; data points that are closer to the unknown point and have a stronger correlation receive a greater weight. By estimating the energy density at all unknown locations within the focusing region, these estimates are then integrated to generate a three-dimensional energy density surface covering the entire focusing region, visually representing the spatial distribution of the ion beam energy.

[0033] Step S180: Identify the spatial coordinates of the energy density extreme point in the three-dimensional energy density surface, and perform a density gradient scan from the extreme point to the surrounding area to determine the boundary surface range where the energy density drops to the preset critical value.

[0034] Energy density extrema points refer to the points in a three-dimensional energy density surface where the energy density is at its maximum or minimum. These points represent key locations where energy is concentrated or dispersed within the focusing region. Scanning the density gradient outwards from these extrema points determines the rate and range of energy density decrease from the extrema point. When the energy density decreases to a preset critical value, the corresponding surface range is the boundary surface range to be determined. This boundary surface range clearly defines the effective range of the ion beam energy within the focusing region. For example, by differentiating the three-dimensional energy density surface, the energy density extrema point is found. Then, starting from this point, a point-by-point scan is performed outwards, calculating the energy density gradient. When the energy density decreases to the preset critical value, the corresponding point is recorded. Connecting these points forms the boundary surface range.

[0035] Step S190: Define the spatial coordinates of the extreme point as the center position of the spatial distribution feature, and define the maximum radial extension distance of the boundary surface range as the diffusion radius of the spatial distribution feature.

[0036] The spatial coordinates of the extreme point represent the location where the ion beam energy is most concentrated in the focusing region, and are therefore defined as the center of the spatial distribution characteristic. The maximum radial extension distance of the boundary surface range reflects the degree of diffusion of the ion beam energy in space, and is defined as the diffusion radius of the spatial distribution characteristic. Using the center location and the diffusion radius, the spatial distribution characteristics of the ion beam energy in the focusing region can be described concisely and accurately. For example, in a three-dimensional coordinate system, the coordinates of the energy density extreme point are recorded as the center location, and the radial distances of the boundary surface range in various directions are measured; the maximum value is taken as the diffusion radius.

[0037] Step S1100: Combine the ion beam intensity, energy attenuation gradient, and spatial distribution characteristics including the center position and diffusion radius into energy distribution parameters.

[0038] Through the preceding steps, the ion beam current intensity, energy attenuation gradient, and spatial distribution characteristics were obtained. Combining these parameters forms a complete set of energy distribution parameters. These parameters comprehensively describe the energy distribution of the ion beam in the focusing region, providing crucial information for subsequent energy adjustment.

[0039] Step S200: Determine the target energy adjustment range based on the energy distribution parameters and the preset ion beam energy reference model. The target energy adjustment range includes the allowable peak energy deviation threshold and the minimum energy gradient change rate.

[0040] In this step, the preset ion beam energy reference model is a pre-established model that includes the energy distribution characteristics and related parameters of the ion beam under ideal conditions, such as the reference current intensity and gradient history curve. By comparing and analyzing the acquired energy distribution parameters with this reference model, the difference between the current ion beam energy and the ideal state can be determined, thereby determining the target energy adjustment range. The allowable peak energy deviation threshold refers to the maximum range by which the peak energy of the ion beam is allowed to deviate from the reference value in the reference model. It is used to constrain the amplitude variation range of the ion source current adjustment command, ensuring that the peak energy of the ion beam is within an acceptable error range. The minimum energy gradient change rate refers to the minimum allowable change rate of the ion beam energy decay gradient. It is used to define the step accuracy threshold of the focusing lens magnetic field adjustment, ensuring that the ion beam energy decay process meets expectations.

[0041] The specific steps for determining the target energy adjustment range are as follows:

[0042] Step S210: Compare the deviation between the ion beam current intensity and the reference current intensity recorded in the ion beam energy reference model to generate the current intensity deviation rate, and select the initial peak energy deviation threshold from the preset dynamic adjustment rule table according to the current intensity deviation rate.

[0043] The current intensity deviation rate refers to the proportion of the deviation between the ion beam current intensity and the reference current intensity, reflecting the degree of difference between the current ion beam current intensity and the ideal state. The preset dynamic adjustment rule table is a pre-defined table that corresponds to different initial peak energy deviation thresholds based on different current intensity deviation rates. By matching the current intensity deviation rate with the dynamic adjustment rule table, a suitable initial peak energy deviation threshold can be selected. For example, the difference between the ion beam current intensity and the reference current intensity is calculated, and then divided by the reference current intensity to obtain the current intensity deviation rate. Then, the initial peak energy deviation threshold corresponding to this deviation rate is found in the dynamic adjustment rule table.

[0044] Step S220: Based on the energy decay gradient and the gradient history curve stored in the ion beam energy reference model, calculate the gradient deviation, which is the maximum difference between the current gradient and the historical gradient curve within the same time window.

[0045] The gradient history curve records the changes in the ion beam energy decay gradient over a period of time. By comparing the current energy decay gradient with the gradient history curve within the same time window, the maximum difference between them is calculated, yielding the gradient deviation. The gradient deviation reflects the degree of difference between the current energy decay gradient and historical data, and it helps determine the stability of the current ion beam energy decay. For example, within the same time window, the value of the current energy decay gradient is compared with the value at the corresponding time point on the gradient history curve, and the largest difference is identified as the gradient deviation.

[0046] Step S230: Perform nonlinear correction on the initial peak energy deviation threshold according to the gradient deviation. If the gradient deviation exceeds the preset gradient stability range, reduce the initial peak energy deviation threshold according to the square of the gradient deviation to generate the corrected peak energy deviation threshold.

[0047] The preset gradient stability interval is a pre-defined range. When the gradient deviation falls within this range, it indicates that the ion beam energy decay gradient is relatively stable. When the gradient deviation exceeds this range, it indicates that the energy decay gradient is unstable, and the initial peak energy deviation threshold needs to be corrected. Reducing the initial peak energy deviation threshold by the square of the gradient deviation is to more strictly constrain the peak energy of the ion beam, thereby improving the stability of the ion beam energy. For example, if the gradient deviation is x and the preset gradient stability interval is [a, b], when x > b, the initial peak energy deviation threshold is multiplied by (1 / x). 2 ), thus obtaining the corrected peak energy deviation threshold.

[0048] Step S240: Extract the diffusion radius from the spatial distribution features, input it into the pre-configured focusing accuracy compensator, output the accuracy compensation coefficient which is negatively correlated with the diffusion radius, and scale the preset minimum energy gradient change rate in the ion beam energy reference model according to the accuracy compensation coefficient.

[0049] The pre-configured focusing accuracy compensator is a pre-set module that outputs a corresponding accuracy compensation coefficient based on the input diffusion radius.

[0050] The focusing accuracy compensator is essentially a functional module built upon algorithms and models. Its core purpose is to adjust focusing accuracy-related parameters based on the spatial distribution characteristics of ion beam energy. From its input perspective, it receives the diffusion radius, a key indicator describing the degree of ion beam energy diffusion in space. A larger diffusion radius means more dispersed ion beam energy and lower focusing accuracy; conversely, a smaller diffusion radius indicates more concentrated ion beam energy and higher focusing accuracy. The focusing accuracy compensator processes the input diffusion radius using its pre-defined internal algorithms and models. These algorithms and models are built upon extensive experimental data and theoretical analysis, reflecting the intrinsic relationship between diffusion radius and focusing accuracy. The output of the focusing accuracy compensator is a precision compensation coefficient that is negatively correlated with the diffusion radius. That is, when the diffusion radius increases, the precision compensation coefficient decreases accordingly; when the diffusion radius decreases, the precision compensation coefficient increases. This negative correlation is designed to effectively compensate for focusing accuracy. For example, when the diffusion radius is large, it indicates that the focusing effect of the ion beam is poor. In this case, a smaller precision compensation coefficient is needed to adjust the subsequent adjustment parameters to enhance the adjustment of focusing precision. Conversely, when the diffusion radius is small, the focusing effect of the ion beam is good. In this case, a larger precision compensation coefficient can be used to appropriately relax the adjustment requirements for focusing precision.

[0051] Taking a specific focusing accuracy compensator implementation as an example, it employs a polynomial fitting algorithm. Based on extensive experimental data, it was found that the diffusion radius r and the accuracy compensation coefficient k approximately satisfy k = ab * r. 2 The relationship is given by (where a and b are constants obtained by fitting experimental data). When the diffusion radius r is input, the focusing accuracy compensator calculates the corresponding accuracy compensation coefficient k according to this polynomial formula.

[0052] After obtaining the accuracy compensation coefficient, it is used to scale the preset minimum energy gradient change rate in the ion beam energy reference model. The preset minimum energy gradient change rate is an important parameter for ion beam energy regulation, specifying the minimum allowable rate of change of the energy decay gradient. By scaling it using the accuracy compensation coefficient, the regulation precision of the energy gradient can be dynamically adjusted according to the actual focusing conditions of the ion beam. For example, if the accuracy compensation coefficient is 0.8 and the preset minimum energy gradient change rate is G0, then the scaled minimum energy gradient change rate becomes 0.8*G0. In this way, during subsequent energy regulation, the change in the energy decay gradient can be controlled more precisely according to the focusing state of the ion beam, thereby improving the focusing precision and accuracy of energy regulation.

[0053] Step S250: Combine the corrected peak energy deviation threshold with the scaled minimum energy gradient change rate to form the target energy adjustment range, wherein the peak energy deviation threshold is used to constrain the amplitude variation range of the ion source current adjustment command, and the minimum energy gradient change rate is used to define the step accuracy threshold of the focusing lens magnetic field adjustment.

[0054] Through the preceding steps, the corrected peak energy deviation threshold and the scaled minimum energy gradient change rate were obtained. Combining them forms the target energy adjustment range. This range clarifies the target and constraints for ion beam energy adjustment, providing a basis for subsequent energy adjustment command generation.

[0055] Step S300: Generate multi-level energy regulation commands based on the target energy regulation range, and control the output parameters of the ion beam energy regulator through the multi-level energy regulation commands so that the current energy distribution of the ion beam enters the target energy regulation range.

[0056] Multi-level energy regulation commands are a series of instructions used to adjust the energy of the ion beam. They are executed sequentially and with priority, gradually adjusting the energy distribution of the ion beam to achieve the target energy regulation range. An ion beam energy regulator is a device used to control the energy of the ion beam. It adjusts its output parameters, such as ion source current, pulse timing, and focusing lens magnetic field, according to the multi-level energy regulation commands, thereby achieving precise regulation of the ion beam energy.

[0057] The specific steps for generating multi-level energy regulation commands are as follows:

[0058] Step S310: Decompose the target energy adjustment range into multiple energy adjustment stages, each energy adjustment stage corresponding to a different energy adjustment priority, wherein: the first adjustment stage prioritizes adjusting the ion beam current intensity to a preset reference intensity range; the second adjustment stage adjusts the ion beam pulse width according to the energy decay gradient so that the energy decay rate matches the minimum energy gradient change rate; the third adjustment stage adjusts the magnetic field strength of the focusing lens based on the diffusion radius of the spatial energy distribution matrix so that the position coordinates of the maximum energy density region coincide with the preset focusing center point.

[0059] The preset reference intensity range is a pre-defined range. The goal of the first adjustment stage is to adjust the ion beam intensity within this range to ensure that the energy carrying capacity of the ion beam meets the requirements. The second adjustment stage controls the energy decay rate by adjusting the ion beam pulse width to match the minimum energy gradient change rate, thereby ensuring a stable energy decay process. The third adjustment stage adjusts the magnetic field strength of the focusing lens according to the diffusion radius of the spatial energy distribution matrix, so that the position of the maximum energy density region coincides with the preset focusing center point, improving the focusing accuracy of the ion beam. For example, in the first adjustment stage, the ion beam intensity is gradually brought closer to the reference intensity range by increasing or decreasing the ion source current; in the second adjustment stage, the ion beam pulse width is adjusted according to the magnitude of the energy decay gradient to ensure that the energy decay rate meets the requirements; in the third adjustment stage, the excitation current of the focusing lens is adjusted according to the size of the diffusion radius to change the magnetic field strength, so that the maximum energy density region moves to the preset focusing center point.

[0060] Step S320: Allocate a corresponding time window for each energy regulation stage, and generate a corresponding regulation command sequence within each time window. The regulation command sequence includes:

[0061] Ion source current adjustment command, used to control the output current of the ion emitter;

[0062] Pulse timing adjustment commands are used to modify the rise time and duration of the ion beam pulse;

[0063] The magnetic field gradient control command is used to change the excitation current of the focusing lens to adjust the magnetic field distribution.

[0064] Allocating time windows for each energy conditioning stage ensures sufficient time for each stage to be adjusted and executed sequentially. Within each time window, a corresponding sequence of conditioning commands is generated to control the output parameters of the ion beam energy conditioner. Ion source current adjustment commands adjust the ion beam intensity by controlling the output current of the ion emitter; pulse timing adjustment commands control the energy decay rate by modifying the rise time and duration of the ion beam pulses; and magnetic field gradient control commands adjust the magnetic field distribution by changing the excitation current of the focusing lens, thereby regulating the ion beam focusing position.

[0065] The specific steps for allocating a time window and generating a sequence of regulation commands for each energy regulation stage are as follows:

[0066] Step S321: Determine the priority order of each energy adjustment stage based on the adjustment urgency weights of the peak energy deviation threshold and the minimum energy gradient change rate within the target energy adjustment range. The adjustment urgency weights are calculated by multiplying the influence coefficient of the peak energy deviation threshold on system stability and the sensitivity coefficient of the minimum energy gradient change rate on focusing accuracy.

[0067] The adjustment urgency weight reflects the importance and urgency of each adjustment parameter to system performance. The adjustment urgency weight is obtained by multiplying the influence coefficient of the peak energy deviation threshold on system stability by the sensitivity coefficient of the minimum energy gradient change rate on focusing accuracy. The priority order of each energy adjustment stage is determined based on the adjustment urgency weight, with higher priority stages adjusted first. For example, if the peak energy deviation threshold has a significant impact on system stability, and the minimum energy gradient change rate also has a significant impact on focusing accuracy, then the adjustment stages corresponding to these two parameters will have higher priority.

[0068] The peak energy deviation threshold refers to the maximum allowable deviation of the ion beam peak energy from the reference value in the reference model, and it is related to the stability of the ion beam energy. In an ion beam energy regulation system, if the peak energy of the ion beam deviates too much from the reference value, it will significantly affect the stability of the entire system. For example, when the peak energy is too high, the ion beam may exert an excessively strong force on the sample, causing sample damage and reduced processing accuracy; while when the peak energy is too low, the ion beam may fail to achieve the expected processing or analysis results, affecting the normal operation of the system. To quantify this impact, an influence coefficient is introduced. It can be obtained through experimental analysis and reflects the extent to which the system stability is affected by each unit change in the peak energy deviation threshold. Specifically, under different peak energy deviation threshold conditions, various performance indicators of the system (such as the stability of the ion beam current, the consistency of processing accuracy, etc.) can be monitored and analyzed, and then the influence coefficient can be calculated based on the changes in these indicators. For example, after a series of experiments, it was found that when the peak energy deviation threshold increases by 1%, the consistency of the system's processing accuracy decreases by 0.5%, so the influence coefficient can be determined based on this relationship.

[0069] The minimum energy gradient change rate refers to the minimum allowable rate of change in the energy attenuation gradient of an ion beam, primarily affecting the focusing accuracy of the ion beam. In semiconductor chip manufacturing, precise focusing ensures the ion beam accurately targets the desired position, improving processing precision and quality. The magnitude of the minimum energy gradient change rate directly influences the energy attenuation during ion beam propagation, thus affecting the focusing effect. Excessive or insufficient minimum energy gradient change rate can lead to issues such as ion beam focus point shift and uneven energy distribution, reducing focusing accuracy. The sensitivity coefficient measures the impact of minute changes in the minimum energy gradient change rate on focusing accuracy. This sensitivity coefficient is also determined through experiments and data analysis. Under different minimum energy gradient change rate conditions, parameters such as the ion beam's focus point position and energy distribution can be measured, and the variation of these parameters with the minimum energy gradient change rate can be analyzed to calculate the sensitivity coefficient. For example, experiments have shown that for every 0.1 increase in the minimum energy gradient change rate, the focus point shift increases by 0.2 micrometers; based on this relationship, the corresponding sensitivity coefficient can be determined.

[0070] Multiplying these two coefficients yields the adjustment urgency weight. This calculation method is used because system stability and focusing accuracy are both indispensable factors in ion beam energy regulation, and they are interrelated and mutually influential. By combining their influence, the importance and urgency of each regulation parameter can be more comprehensively assessed. The larger the adjustment urgency weight, the more significant the impact of that parameter on the overall system performance, and the higher the priority of the corresponding energy regulation stage. For example, if the peak energy deviation threshold has an influence coefficient of 0.6 on system stability, and the minimum energy gradient change rate has a sensitivity coefficient of 0.8 on focusing accuracy, then the adjustment urgency weight is 0.6 × 0.8 = 0.48. When determining the priority of the energy regulation stage, parameters with higher adjustment urgency weights are adjusted first based on this weight value to ensure that the system can quickly and effectively reach the target energy regulation range while maintaining system stability and focusing accuracy.

[0071] Step S322: Allocate an initial time window for the first adjustment stage based on priority order. The duration of the initial time window is inversely proportional to the difference between the peak energy deviation threshold and the current ion beam intensity. Allocate a dynamic time window for the second adjustment stage. The start time of the dynamic time window is determined by the sum of the end time of the initial time window and the preset buffer interval.

[0072] The duration of the initial time window is determined by the difference between the peak energy deviation threshold and the current ion beam intensity. A larger difference indicates a greater gap between the ion beam intensity and the target value, requiring more time for adjustment, thus resulting in a longer initial time window. Conversely, a smaller difference leads to a shorter initial time window. The dynamic time window begins after a preset buffer interval following the end of the initial time window, preventing interference between different adjustment stages. For example, if the peak energy deviation threshold is A, the current ion beam intensity is B, and the difference is |AB|, then the duration T1 of the initial time window can be expressed as T1 = k / |AB|, where k is a constant. With a preset buffer interval of t, the start time of the dynamic time window is the end time of the initial time window plus t.

[0073] Step S323: Based on the diffusion radius of the spatial distribution characteristics and the maximum adjustment rate of the focusing lens, calculate the time window length of the third adjustment stage. The time window length is proportional to the square root of the diffusion radius, and the start time is immediately after the end time of the dynamic time window.

[0074] A larger diffusion radius indicates a greater degree of energy diffusion in space, requiring more time to adjust the magnetic field strength of the focusing lens to move the region of maximum energy density to the preset focusing center point. The maximum adjustment rate of the focusing lens limits the speed of magnetic field strength adjustment; therefore, the time window length of the third adjustment stage is proportional to the square root of the diffusion radius. For example, if the diffusion radius is r and the maximum adjustment rate of the focusing lens is v, then the time window length T3 of the third adjustment stage can be expressed as T3 = m * √r / v, where m is a constant. The start time of the third adjustment stage immediately follows the end time of the dynamic time window to ensure the continuity of the adjustment process.

[0075] Step S324: Generate an ion source current adjustment command sequence within the initial time window of the first adjustment stage. The step size of the ion source current adjustment command sequence is dynamically adjusted by the difference between the peak energy deviation threshold and the current ion beam intensity, and the interval between adjacent commands is inversely proportional to the rate of change of the difference.

[0076] The step size of the ion source current adjustment command sequence is dynamically adjusted based on the difference between the peak energy deviation threshold and the current ion beam intensity. A larger difference results in a larger step size to more quickly approach the target value; a smaller difference results in a smaller step size to improve adjustment accuracy. The interval between adjacent commands is inversely proportional to the rate of change of the difference. A larger rate of change indicates a faster change in the ion beam intensity, requiring more frequent adjustment commands, thus resulting in a shorter interval between adjacent commands; conversely, a smaller rate of change results in a longer interval. For example, if the peak energy deviation threshold is A, the current ion beam intensity is B, the difference is |AB|, and the rate of change of the difference is Δ|AB| / Δt, then the step size ΔI of the ion source current adjustment command sequence can be expressed as ΔI = n * |AB|, where n is a constant, and the interval ΔT between adjacent commands can be expressed as ΔT = p / (Δ|AB| / Δt), where p is a constant.

[0077] Step S325: Generate a pulse timing adjustment command sequence within the dynamic time window of the second adjustment stage. The trigger interval of the pulse timing adjustment command is determined by the ratio of the energy decay gradient to the minimum energy gradient change rate, and the duration of each pulse timing adjustment command is proportional to the length of the dynamic time window.

[0078] The trigger interval for pulse timing adjustment commands is determined by the ratio of the energy decay gradient to the minimum energy gradient change rate. A larger ratio indicates a greater gap between the current energy decay gradient and the target value, requiring more frequent pulse timing adjustments, thus resulting in a shorter trigger interval. Conversely, a smaller ratio results in a longer trigger interval. The duration of each pulse timing adjustment command is proportional to the length of the dynamic time window. A longer dynamic time window provides more time for pulse timing adjustments, thus extending the duration of each command. For example, if the energy decay gradient is G and the minimum energy gradient change rate is G0, the trigger interval ΔT2 for pulse timing adjustment commands can be expressed as ΔT2 = q / (G / G0), where q is a constant. The length of the dynamic time window is T2, and the duration Δt of each pulse timing adjustment command can be expressed as Δt = r * T2, where r is a constant.

[0079] Step S326: Generate a magnetic field gradient control command group within the time window of the third adjustment stage. The amplitude of each control signal in the magnetic field gradient control command group is determined by the diffusion radius and the magnetic field response sensitivity of the focusing lens, and the alternation frequency of the control signal is proportional to the reciprocal of the time window length.

[0080] The amplitude of each control signal in the magnetic field gradient control command group is determined based on the diffusion radius and the magnetic field response sensitivity of the focusing lens. A larger diffusion radius indicates a greater change in magnetic field strength is needed to adjust the focusing position, thus resulting in a larger control signal amplitude. Conversely, a higher magnetic field response sensitivity of the focusing lens means a greater change in magnetic field strength under the same control signal, allowing for a relatively smaller control signal amplitude. The alternation frequency of the control signals is proportional to the reciprocal of the time window length. A shorter time window requires a faster adjustment of the magnetic field strength, resulting in a higher control signal alternation frequency; conversely, a longer time window results in a lower control signal alternation frequency. For example, if the diffusion radius is r and the magnetic field response sensitivity of the focusing lens is s, then the amplitude ΔB of each control signal in the magnetic field gradient control command group can be expressed as ΔB = s * r * k1, where k1 is a constant. The time window length for the third adjustment stage is T3, so the alternation frequency f of the control signals can be expressed as f = k2 / T3, where k2 is a constant.

[0081] Step S327: Monitor the fluctuation amplitude of ion beam intensity after command execution within each time window in real time. If the fluctuation amplitude exceeds the preset ratio of the peak energy deviation threshold, shorten the current time window and insert a transition buffer command. The duration of the transition buffer command is inversely proportional to the square root of the fluctuation amplitude exceeding the value.

[0082] Real-time monitoring of ion beam intensity fluctuations ensures the stability of the regulation process. If the fluctuation exceeds a preset percentage of the peak energy deviation threshold, it indicates a potential anomaly in the regulation process, requiring adjustment measures. Shortening the current time window accelerates regulation, while inserting a transition buffer command makes the process smoother. The duration of the transition buffer command is inversely proportional to the square root of the excess fluctuation value; the larger the excess value, the more severe the anomaly, requiring a shorter transition buffer time to restore stability. For example, if the peak energy deviation threshold is A, the preset percentage is α, and the ion beam intensity fluctuation is ΔI, when ΔI > α*A, the current time window is shortened, and a transition buffer command is inserted. The duration Δt of the transition buffer command can be expressed as Δt = m / √(ΔI - α*A), where m is a constant.

[0083] Step S328: After all time windows have completed the execution of the instructions, the minimum energy gradient change rate is recalibrated based on the change in diffusion radius of the adjusted spatial distribution characteristics. If the calibrated change rate exceeds the allowable error of the target energy adjustment range, the time window reallocation process is triggered.

[0084] The change in the diffusion radius of the adjusted spatial distribution characteristics reflects the impact of the adjustment process on the ion beam focusing. Recalibrating the minimum energy gradient change rate based on this change allows for more accurate energy adjustment to adapt to the actual conditions of the ion beam. If the calibrated change rate exceeds the allowable error within the target energy adjustment range, it indicates that the current time window allocation may be unreasonable, requiring a time window reallocation process to ensure the adjustment process reaches the target energy adjustment range. For example, if the diffusion radius before adjustment is r1 and the diffusion radius after adjustment is r2, the change in diffusion radius is Δr = |r2 - r1|. The minimum energy gradient change rate G' is recalculated based on Δr. If |G' - G0| > ε, where G0 is the minimum energy gradient change rate within the target energy adjustment range and ε is the allowable error, then the time window reallocation process is triggered.

[0085] Step S330: During the execution of the multi-level energy regulation command, the completion status of each regulation stage is monitored in real time. If the regulation error of a certain stage exceeds the stage allowable error threshold, the subsequent stages are paused and the error compensation subprocess is started.

[0086] Real-time monitoring of the completion status of each adjustment stage is crucial to ensuring the accuracy and stability of the adjustment process. The permissible error threshold for each stage is a pre-set allowable error range. If the adjustment error of a certain stage exceeds this threshold, it indicates a potential problem with the adjustment in that stage, requiring the suspension of subsequent stages and the initiation of an error compensation subprocess to correct the error. For example, in the first adjustment stage, the permissible error threshold is set to β. If the adjustment error of the ion beam intensity exceeds β, i.e., |I-I0|>β, where I is the current ion beam intensity and I0 is the target value of the preset reference intensity range, then the adjustment of subsequent stages is suspended, and the error compensation subprocess is initiated.

[0087] Step S400: Real-time detection of energy fluctuation data of the adjusted ion beam on the sample surface. The energy fluctuation data includes instantaneous energy offset and energy diffusion coefficient at the focal point.

[0088] Instantaneous energy offset refers to the difference between the energy of the ion beam on the sample surface at a certain moment and the expected energy; it reflects the instantaneous stability of the ion beam energy. The focal point energy diffusion coefficient is an indicator that measures the degree of energy diffusion of the ion beam around the focal point; it reflects the focusing accuracy of the ion beam. By real-time monitoring of these energy fluctuation data, the energy distribution of the ion beam on the sample surface can be understood promptly, providing a basis for subsequent adaptive compensation.

[0089] The specific steps for detecting energy fluctuation data are as follows:

[0090] Step S410: Arrange an array of energy sensors on the sample surface. The array of energy sensors consists of multiple micro-sensor units distributed in a grid pattern, and each micro-sensor unit covers a preset detection area.

[0091] An array-type energy sensor is a device used to detect the energy of an ion beam. It consists of multiple miniature sensor units arranged in a grid on the sample surface, enabling comprehensive and accurate detection of the energy distribution of the ion beam on the sample surface. Each miniature sensor unit covers a pre-defined detection area, ensuring energy detection at every location on the sample surface. For example, miniature sensor units are arranged at intervals on the sample surface to form a two-dimensional grid structure, with each unit responsible for detecting the ion beam energy in its designated area.

[0092] Step S420: Synchronously acquire the output signals of all micro-sensor units to generate an energy distribution heat map.

[0093] Synchronously acquiring the output signals of all microsensor units ensures that the acquired energy data is obtained at the same time, thus accurately reflecting the energy distribution of the ion beam on the sample surface. An energy distribution heatmap is a color-coded graph that visually displays the energy distribution of the ion beam on the sample surface. For example, by simultaneously acquiring the output signals of all microsensor units through a data acquisition system, converting these signals into corresponding energy values, and then displaying them on the heatmap using different colors based on the energy value—higher energy areas are darker, and lower energy areas are lighter.

[0094] Step S430: Perform edge detection on the energy distribution heatmap to identify the location and shape characteristics of energy anomaly areas.

[0095] Edge detection is an image processing technique that identifies areas of significant energy variation in an energy distribution heatmap, i.e., energy anomaly regions. Edge detection determines the location and shape characteristics of these anomaly regions, providing a basis for subsequent analysis. For example, using the Canny edge detection algorithm to process an energy distribution heatmap, it identifies the edges of energy changes, thereby determining the boundaries of energy anomaly regions and obtaining their location and shape features.

[0096] Step S440: Extract the instantaneous energy offset from the energy anomaly region. The instantaneous energy offset is the difference between the maximum and minimum energy values ​​within the current detection period.

[0097] In the energy anomaly region, energy fluctuations are more pronounced. By extracting the difference between the maximum and minimum energy values ​​within the current detection cycle in this region, the instantaneous energy offset can be obtained. This value reflects the instantaneous change in ion beam energy within this region. For example, within one detection cycle, the maximum energy value Emax and the minimum energy value Emin in the energy anomaly region are recorded, and the instantaneous energy offset ΔE = Emax - Emin.

[0098] Step S450: Calculate the variance of the energy distribution heatmap and normalize the variance to the energy diffusion coefficient at the focal point.

[0099] The variance of an energy distribution heatmap reflects the dispersion of energy distribution on the sample surface. A larger variance indicates a more dispersed energy distribution and greater energy diffusion at the focal point. Normalizing the variance to a focal point energy diffusion coefficient makes this coefficient comparable. For example, first calculate the variance σ of the energy values ​​of all pixels in the energy distribution heatmap. 2 Then, the variance value is converted into the focal point energy diffusion coefficient C using a normalization algorithm, for example, C = σ 2 / (Emax-Emin).

[0100] Step S460: Match the shape characteristics of the energy anomaly region with the preset defect pattern library. If a known energy diffusion pattern is matched, generate the corresponding diffusion type identifier and associate the diffusion type identifier with the energy fluctuation data.

[0101] A pre-defined defect pattern library stores various known energy diffusion patterns and their corresponding shape features. By matching the shape features of an energy anomaly region with the defect pattern library, the current energy diffusion pattern can be identified. If a match is successful, a corresponding diffusion type identifier is generated and associated with the energy fluctuation data, allowing for more accurate analysis of ion beam energy fluctuations. For example, the defect pattern library stores energy diffusion patterns of different shapes, such as circles, ellipses, and rectangles, along with their corresponding identifiers. When the shape of an energy anomaly region matches a certain pattern, the identifier for that pattern is extracted and associated with the energy fluctuation data.

[0102] Step S500: Based on the energy fluctuation data and the preset adaptive compensation algorithm, adjust the feedback control parameters of the ion beam energy regulator to maintain the energy distribution of the ion beam within the target energy regulation range.

[0103] The preset adaptive compensation algorithm is a pre-designed algorithm that automatically adjusts the feedback control parameters of the ion beam energy regulator based on energy fluctuation data to achieve adaptive regulation of the ion beam energy. Feedback control parameters include the ion source current adjustment step size, feedback period, and magnetic field response coefficient. By adjusting these parameters, the energy distribution of the ion beam can be more stably maintained within the target energy regulation range.

[0104] The process of adjusting the feedback control parameters using the adaptive compensation algorithm is as follows:

[0105] Step S510: The instantaneous energy offset is compared with the peak energy deviation threshold in the target energy adjustment range using an adaptive compensation algorithm. When the absolute value of the offset exceeds the preset offset threshold, an error polarity identifier and an amplitude weight are generated. The error polarity identifier indicates the direction of energy offset, and the amplitude weight is the ratio of the absolute value of the offset to the offset threshold.

[0106] The preset offset threshold is a pre-defined allowable offset range. When the absolute value of the instantaneous energy offset exceeds this threshold, it indicates a significant deviation in the ion beam energy, requiring compensation. The error polarity indicator is used to indicate the direction of the energy offset, such as whether it is a positive or negative offset. The amplitude weight reflects the relative relationship between the magnitude of the offset and the offset threshold, and it will be used in subsequent parameter adjustment calculations. For example, if the instantaneous energy offset is ΔE, the peak energy deviation threshold in the target energy adjustment range is A, and the preset offset threshold is γ, when |ΔE|>γ, if ΔE>0, the error polarity indicator is positive, and the amplitude weight w=|ΔE| / γ; if ΔE<0, the error polarity indicator is negative, and the amplitude weight is also w=|ΔE| / γ.

[0107] Step S520: Determine the direction of ion source current adjustment based on the error polarity identifier, and calculate the current adjustment step size according to the amplitude weight. The current adjustment step size is proportional to the square root of the amplitude weight, and the feedback control cycle is shortened based on the reciprocal of the amplitude weight, so that the current adjustment rate in the feedback control parameters is nonlinearly related to the error intensity.

[0108] The error polarity indicator signifies the direction of energy shift. Based on this indicator, the direction of ion source current adjustment can be determined. For example, when the energy shifts positively, the ion source current needs to be decreased; when the energy shifts negatively, the ion source current needs to be increased. The current adjustment step size is calculated based on the square root of the amplitude weight. A larger amplitude weight indicates a larger shift, requiring a larger adjustment step size to quickly correct the error. Shortening the feedback control cycle based on the reciprocal of the amplitude weight allows for more frequent adjustments when the error is large, thus creating a non-linear relationship between the current adjustment rate and the error intensity, improving adjustment efficiency. For example, if the error polarity indicator is positive, the ion source current needs to be decreased, and the current adjustment step size ΔI = k * √w, where k is a constant, and the feedback control cycle T = T0 / w, where T0 is the initial feedback control cycle.

[0109] Step S530: Input the focal point energy diffusion coefficient into the diffusion trend prediction model in the adaptive compensation algorithm, and output the diffusion runaway index. When the diffusion runaway index exceeds the minimum energy gradient change rate constraint of the target energy adjustment range, the magnetic field gradient compensation mechanism is triggered.

[0110] The diffusion trend prediction model is a sub-model within the adaptive compensation algorithm. It predicts the trend of ion beam energy diffusion based on the focal point energy diffusion coefficient and outputs a diffusion runaway index. The diffusion runaway index reflects the severity of ion beam energy diffusion. When it exceeds the minimum energy gradient change rate constraint of the target energy adjustment range, it indicates that the energy diffusion has exceeded an acceptable range, requiring the triggering of a magnetic field gradient compensation mechanism for adjustment. For example, the diffusion trend prediction model can be a machine learning-based model that establishes a relationship between the diffusion coefficient and the diffusion runaway index by learning from historical focal point energy diffusion coefficient data. When the current focal point energy diffusion coefficient is input, the corresponding diffusion runaway index is output.

[0111] Step S540: In the magnetic field gradient compensation mechanism, a gradient control command containing magnetic field intensity increment and phase delay is generated based on the diffusion type identifier matching magnetic field waveform correction strategy, and the magnitude of magnetic field intensity increment and phase delay is scaled by the magnetic field response coefficient in the feedback control parameters.

[0112] The magnetic field waveform correction strategy is a series of pre-defined strategies for adjusting the magnetic field waveform based on different diffusion type identifiers. After matching the corresponding strategy to the diffusion type identifier, gradient control commands containing magnetic field strength increments and phase delays are generated. These commands are used to adjust the magnetic field distribution of the focusing lens. By scaling the magnitudes of the magnetic field strength increments and phase delays using the magnetic field response coefficients in the feedback control parameters, precise adjustments can be made based on the actual response of the system. For example, when the diffusion type identifier is a certain type, the corresponding magnetic field waveform correction strategy is matched, and the generated gradient control commands contain a magnetic field strength increment of ΔB and a phase delay of... After scaling by the magnetic field response coefficient s, the actual magnetic field strength increment is s*ΔB, and the phase delay is...

[0113] Step S550: Real-time acquisition of the instantaneous energy offset and the energy diffusion coefficient at the focal point after adjustment, and calculation of the composite convergence index. The composite convergence index is composed of the product of the offset decrease rate and the diffusion coefficient fluctuation decay rate.

[0114] Real-time acquisition of the adjusted instantaneous energy offset and the focal point energy diffusion coefficient is used to evaluate the adjustment effect. The composite convergence index is a comprehensive indicator, composed of the product of the offset decrease rate and the diffusion coefficient fluctuation decay rate, reflecting the convergence of the ion beam energy in terms of stability and focusing accuracy. The faster the offset decrease rate and the faster the diffusion coefficient fluctuation decay rate, the better the adjustment effect, and the larger the composite convergence index. For example, within a time period, the change in instantaneous energy offset is recorded, and the offset decrease rate v1 = (ΔE1 - ΔE2) / Δt is calculated, where ΔE1 and ΔE2 are the instantaneous energy offsets at the beginning and end of the time period, respectively, and Δt is the length of the time period. Similarly, the fluctuation decay rate of the focal point energy diffusion coefficient v2 = (C1 - C2) / Δt is calculated, where C1 and C2 are the focal point energy diffusion coefficients at the beginning and end of the time period, respectively. The composite convergence index C = v1 * v2.

[0115] Step S560: When the composite convergence index is lower than the preset index threshold, iteratively update the current adjustment step size and the magnetic field response coefficient. The step size is reduced according to the exponential decay rule, and the magnetic field response coefficient is linearly adjusted according to the decreasing rate of the diffusion runaway index.

[0116] The preset threshold is a pre-defined standard for judging whether the regulation effect is good. When the composite convergence index is below this threshold, it indicates that the regulation effect is not ideal and further adjustments to the feedback control parameters are needed. The current regulation step size decreases according to the exponential decay rule. As the number of iterations increases, the step size gradually decreases to improve the regulation accuracy. The magnetic field response coefficient is adjusted linearly according to the rate of decrease of the diffusion runaway index. When the diffusion runaway index decreases rapidly, the magnetic field response coefficient increases accordingly to accelerate the magnetic field regulation speed; when the diffusion runaway index decreases slowly, the magnetic field response coefficient decreases accordingly to avoid over-regulation. For example, assuming the initial current regulation step size is ΔI0, the number of iterations is n, and the exponential decay rule is ΔIn = ΔI0 * e (-λn) , where λ is the decay coefficient. The rate of decrease of the diffusion runaway exponent is v3, and the adjustment formula for the magnetic field response coefficient is s=s0+k*v3, where s0 is the initial magnetic field response coefficient and k is the adjustment coefficient.

[0117] Step S570: When the instantaneous energy offset is within the peak energy deviation threshold and the diffusion runaway index is lower than the minimum energy gradient change rate constraint in N consecutive iterations, freeze the current adjustment step size, feedback period and magnetic field response coefficient in the feedback control parameters, and generate a lock signal, N>1.

[0118] If, in N consecutive iterations, the instantaneous energy offset remains within the peak energy deviation threshold and the diffusion runaway index is below the minimum energy gradient rate of change constraint, it indicates that the energy distribution of the ion beam has stabilized within the target energy adjustment range. At this point, the current adjustment step size, feedback period, and magnetic field response coefficient in the feedback control parameters are frozen, further adjustments are stopped, and a lock signal is generated to maintain the current stable state. For example, if N = 5, when the absolute value of the instantaneous energy offset is less than the peak energy deviation threshold and the diffusion runaway index is less than the minimum energy gradient rate of change constraint in 5 consecutive iterations, the current adjustment step size, feedback period, and magnetic field response coefficient are fixed and no further adjustments are made, and a lock signal is generated.

[0119] For example, in adaptive compensation algorithms, dynamic error perception and quantization can be calculated using sliding window differential calculation. A fixed time window (e.g., 10ms) is used to continuously sample the instantaneous energy offset, calculating the mean and standard deviation of the offset within the window to generate a dynamic error baseline. The real-time difference between the current offset and the baseline value is used to trigger the compensation threshold. The energy offset is decomposed into directional polarity (positive / negative) and absolute amplitude. The direction is extracted using a sign function, and the amplitude weight is quantized using an absolute value function. The error signal is processed through a sliding window to reduce noise interference, and the polarity-amplitude decomposition provides discrete control input for subsequent compensation. The feedback parameter nonlinear mapping technique can employ a piecewise proportional-integral (PI) controller, dividing the adjustment range according to the error amplitude weight. For example, for the low amplitude range (<30% threshold): linear proportional adjustment is used, with the step size proportional to the amplitude weight; for the high amplitude range (≥30% threshold): an integral accumulation term is introduced, with the step size increasing by the square root of the amplitude to suppress overshoot. Dynamic compression of the feedback cycle can shorten the control cycle based on the reciprocal function of the error amplitude, where k is the compression coefficient and |E| is the normalized error amplitude. The nonlinear PI strategy achieves rapid error convergence, and dynamic cycle compression improves the response speed to high-frequency disturbances. The diffusion trend prediction model can be implemented as a time-series autoregressive (AR) model, establishing a first-order autoregressive equation for the focal point energy diffusion coefficient sequence: D t =αD t-1 +βΔP+∈, where D tLet ΔP be the current diffusion coefficient, α and β be the ambient pressure change, and ∈ be the fitting parameters. Residual trend detection is used to calculate the moving average of continuously predicted residuals. If the mean residual exceeds three times the historical standard deviation, diffusion runaway is determined. The AR model captures the time dependence of the diffusion coefficient, and residual analysis enables real-time alerts for abnormal trends. Magnetic field gradient compensation waveform synthesis technology can employ phase-synchronous pulse modulation to generate a sinusoidal modulated waveform based on the diffusion runaway direction. The waveform frequency is synchronized with the ion beam pulse frequency, and the phase difference is dynamically adjusted according to the diffusion coefficient change rate. Based on magnetic field strength envelope control, trapezoidal wave envelope modulation is used. The rising slope is determined by the diffusion runaway exponent, and the peak intensity is exponentially related to the error amplitude (B0). peak =B0e λ|E| ), where λ is the attenuation factor. Phase synchronization avoids magnetic field-pulse interference, and envelope modulation achieves a nonlinear match between intensity and the degree of diffusion degradation.

[0120] Step S600: After adjusting the feedback control parameters, start the energy distribution sampling cycle of the preset number of times, and continuously collect multiple sets of energy fluctuation data in each sampling cycle.

[0121] The preset number of energy distribution sampling periods is used to further verify the stability of ion beam energy regulation. Multiple sets of energy fluctuation data are continuously collected within each sampling period. Analysis of this data provides a more comprehensive understanding of the ion beam energy distribution. For example, setting the preset number of sampling periods to 10, with each sampling period lasting 10 seconds, and collecting one set of energy fluctuation data per second within each sampling period, results in 100 sets of energy fluctuation data collected over 10 sampling periods.

[0122] Step S700: Calculate the standard deviation of each set of energy fluctuation data, and use the median of all standard deviations as the current energy stability index.

[0123] Standard deviation is a statistic that measures the dispersion of data. Calculating the standard deviation of each set of energy fluctuation data can reflect the fluctuation of that set of data. The median of all standard deviations is used as the current energy stability index. The median can avoid the influence of individual outliers and more accurately reflect the overall stability of ion beam energy. For example, for 100 sets of energy fluctuation data collected, the standard deviation of each set of data is calculated. Then, these 100 standard deviations are sorted from smallest to largest, and the median value is taken as the current energy stability index.

[0124] Step S800: If the current energy stability index is lower than the preset stability threshold, the ion beam energy adjustment is determined to be complete, and the current parameter configuration of the ion beam energy regulator is locked.

[0125] The preset stability threshold is a pre-defined standard for judging whether the ion beam energy is stable. When the current energy stability index is lower than this threshold, it indicates that the ion beam energy has reached a stable state, and the ion beam energy adjustment is considered complete. The current parameter configuration of the ion beam energy regulator is locked to ensure that the ion beam energy remains stable in subsequent operations. For example, if the preset stability threshold is δ, when the current energy stability index is less than δ, the ion beam energy adjustment is considered complete, and the current parameter configuration of the ion beam energy regulator, such as the ion source current step size, magnetic field response coefficient, and pulse synchronization delay, is locked.

[0126] Step S801: Record the instantaneous value of vacuum chamber pressure, stable value of ion source temperature and circulation rate of cooling system at the moment of locking in real time, generate an environmental parameter snapshot, and bind the environmental parameter snapshot with the ion source current step size, magnetic field response coefficient and pulse synchronization delay in the current parameter configuration and store it in the historical adjustment database.

[0127] Environmental parameters such as vacuum chamber pressure, ion source temperature, and cooling system circulation rate affect ion beam energy regulation. These environmental parameters are recorded in real-time at the lock-in moment, generating an environmental parameter snapshot. This snapshot is then bound to the current parameter configuration and stored in the historical regulation database. This historical data can be referenced in subsequent adjustments, improving the efficiency and accuracy of regulation. For example, at the lock-in moment, the vacuum chamber pressure is recorded as P, the ion source temperature as T, and the cooling system circulation rate as v. These parameters are combined into an environmental parameter snapshot (P, T, v), and stored in the historical regulation database along with current parameters such as the ion source current step size ΔI, magnetic field response coefficient s, and pulse synchronization delay t.

[0128] Step S802: Based on the comparison diagram of energy distribution before and after regulation in the energy regulation report, extract the peak displacement of energy density and the edge energy decay rate to generate a regulation performance fingerprint. The regulation performance fingerprint includes the peak displacement direction indicator and the decay rate change ratio.

[0129] The energy distribution comparison chart before and after regulation in the energy regulation report can visually demonstrate the changes in ion beam energy before and after regulation. By extracting the peak energy density shift and the edge energy decay rate, a regulation effectiveness fingerprint is generated, which reflects the effect of this energy regulation. The peak shift direction indicator shows the direction of movement of the energy density peak before and after regulation, and the change in decay rate reflects the change in the edge energy decay rate before and after regulation. For example, if the position of the energy density peak is (x1, y1, z1) before regulation and (x2, y2, z2) after regulation, the peak energy density shift is √((x2-x1)). 2 +(y2-y1) 2 +(z2-z1) 2The peak displacement direction identifier can be determined by calculating the direction of the displacement vector. The energy decay rate at the leading edge is adjusted to r1, and after adjustment to r2, with the decay rate change ratio being r2 / r1. The peak displacement direction identifier and the decay rate change ratio are combined to form the regulation performance fingerprint.

[0130] Step S803: Associate and store the adjustment performance fingerprint with the environmental parameter snapshot, and add an environmental matching degree tag to each record in the historical adjustment database. The tag is generated by weighted calculation of vacuum chamber pressure deviation, ion source temperature fluctuation range and cooling rate consistency coefficient.

[0131] Linking and storing the regulation performance fingerprint with environmental parameter snapshots establishes a relationship between environmental parameters and regulation effects. An environmental matching label is added to each record in the historical regulation database. This label is generated through a weighted calculation of vacuum chamber pressure deviation, ion source temperature fluctuation range, and cooling rate consistency coefficient, reflecting the similarity of regulation effects under different environmental conditions. For example, for each record in the historical regulation database, the deviation between its vacuum chamber pressure and the current pressure, the fluctuation range of the ion source temperature during regulation, and the consistency coefficient of its cooling rate with the current rate are calculated. Different weights are assigned to these parameters, and the environmental matching label is obtained through weighted calculation.

[0132] Step S804: When the initial environmental parameters of a new ion beam processing task are detected, calculate the similarity between them and the environmental matching degree tags of each record in the historical adjustment database. If there is a record with a similarity exceeding the preset matching threshold, directly load the ion source current step size, magnetic field response coefficient and pulse synchronization delay in that record as the initial parameter configuration.

[0133] When a new ion beam processing task is initiated, its initial environmental parameters are detected. By calculating the similarity between these parameters and the environmental matching labels of records in the historical conditioning database, historical records with similar environmental conditions can be found. If a record exists with a similarity exceeding a preset matching threshold, it indicates successful conditioning experience under similar environmental conditions. The ion source current step size, magnetic field response coefficient, and pulse synchronization delay from that record can be directly loaded as the initial parameter configuration, saving conditioning time and improving conditioning efficiency. For example, methods such as cosine similarity can be used to calculate the similarity between the initial environmental parameters of the new task and the environmental matching labels of historical records. A preset matching threshold of θ is used; when the similarity is greater than θ, the parameter configuration from that record is loaded.

[0134] Step S805: After loading the historical parameter configuration, skip the step of generating multi-level energy adjustment instructions and directly enter the step of real-time detection of energy fluctuation data of the adjusted ion beam on the sample surface, and synchronously adjust the signal acquisition timing of the array energy sensor according to the loaded pulse synchronization delay.

[0135] After loading historical parameter configurations, since these parameters have been verified in similar environments, the step of generating multi-level energy regulation commands can be skipped, and the process can directly proceed to the step of real-time detection of energy fluctuation data, thus accelerating the regulation process. The signal acquisition timing of the array-type energy sensor is adjusted synchronously according to the loaded pulse synchronization delay, ensuring that the acquired energy data is synchronized with the ion beam's pulse signal, improving data accuracy. For example, if the loaded pulse synchronization delay is t, the signal acquisition timing of the array-type energy sensor is delayed by t time to synchronize it with the ion beam's pulse signal.

[0136] Step S900: If the current energy stability index is higher than the stability threshold, the adaptive compensation algorithm is re-executed, and the compensation intensity of the magnetic field gradient compensation mechanism is increased in each iteration until the maximum number of iterations is reached or the current energy stability index meets the threshold.

[0137] If the current energy stability index is higher than the stability threshold, it indicates that the ion beam energy is not stable enough and the adaptive compensation algorithm described above needs to be re-executed for adjustment. Increasing the compensation intensity of the magnetic field gradient compensation mechanism in each iteration can accelerate the adjustment speed and allow the ion beam energy to reach a stable state more quickly. Iteration stops when the maximum number of iterations is reached or the current energy stability index meets the threshold. For example, setting the maximum number of iterations to M, the compensation intensity of the magnetic field gradient compensation mechanism increases by a fixed percentage in each iteration; iteration stops when the number of iterations reaches M or the current energy stability index is lower than the stability threshold.

[0138] Step S1000: During the ion beam energy adjustment process, continuously collect the voltage fluctuation curve of the ion source high voltage power supply and the vacuum chamber leakage rate monitoring signal. The voltage fluctuation curve includes the instantaneous voltage drop depth and recovery time, and the leakage rate monitoring signal includes the gas pressure rise rate per unit time.

[0139] During ion beam energy regulation, voltage fluctuations in the ion source's high-voltage power supply and leaks in the vacuum chamber affect the ion beam energy. Continuously acquiring voltage fluctuation curves and leakage rate monitoring signals allows for the timely detection of these anomalies, providing a basis for subsequent processing. The instantaneous voltage drop depth and recovery time reflect the severity and recovery status of the voltage fluctuation, while the pressure rise rate per unit time reflects the degree of leakage in the vacuum chamber. For example, a voltage sensor can be used to acquire the voltage signal from the ion source's high-voltage power supply in real time, plotting the voltage fluctuation curve and recording the instantaneous voltage drop depth and recovery time. A pressure sensor can be used to monitor pressure changes in the vacuum chamber and calculate the pressure rise rate per unit time.

[0140] Step S1100: When it is detected that the depth of the instantaneous voltage dip exceeds the preset voltage tolerance, immediately send a beam current truncation instruction to the ion emitter and activate the overvoltage protection circuit of the high-voltage power supply. The execution delay time of the truncation instruction is inversely proportional to the absolute value of the voltage dip depth.

[0141] The preset voltage tolerance is a preset range of allowable voltage dips. When the depth of the instantaneous voltage dip exceeds this tolerance, it indicates that the voltage fluctuation has seriously affected the normal emission of the ion beam, and immediate measures need to be taken. Sending a beam current truncation instruction to the ion emitter can stop the emission of the ion beam and avoid damage to the equipment. Activating the overvoltage protection circuit of the high-voltage power supply can protect the safety of the power supply equipment. The execution delay time of the truncation instruction is inversely proportional to the absolute value of the voltage dip depth. The greater the voltage dip depth, the more urgent the situation is, and the faster the truncation instruction needs to be executed. For example, the preset voltage tolerance is U0. When the depth of the instantaneous voltage dip |ΔU| > U0, send a beam current truncation instruction to the ion emitter. The execution delay time t of the truncation instruction is t = k / |ΔU|, where k is a constant, and at the same time, activate the overvoltage protection circuit of the high-voltage power supply.

[0142] Step S1200: When it is detected that the air pressure rising rate exceeds the safety threshold, close the electromagnetic valves of all ion transfer channels and start the standby vacuum pump group to operate at the preset maximum power until the leakage rate monitoring signal falls back within the safe range.

[0143] The safety threshold is a preset range of allowable air pressure rising rates. When it is detected that the air pressure rising rate exceeds this threshold, it indicates that there is a leakage problem in the vacuum chamber and timely treatment is required. Closing the electromagnetic valves of all ion transfer channels can prevent the ion beam from being affected by the leaked gas. Starting the standby vacuum pump group to operate at the preset maximum power can accelerate the pumping speed and reduce the air pressure in the vacuum chamber until the leakage rate monitoring signal falls back within the safe range. For example, the preset safety threshold is v0. When the air pressure rising rate v > v0, close the electromagnetic valves of all ion transfer channels, start the standby vacuum pump group to operate at the maximum power P, and monitor the leakage rate monitoring signal in real time. When the air pressure rising rate v < v0, stop the operation of the standby vacuum pump group.

[0144] Step S1300: After completing the voltage protection or leakage control operation, re-initialize the parameter configuration of the ion beam energy regulator. The parameter configuration includes the progress identifier of the adjustment instruction sequence being executed at the interruption moment, the remaining duration of the uncompleted feedback control cycle, and the waveform break point position of the magnetic field gradient control instruction.

[0145] After completing voltage protection or leakage control operations, the ion beam energy regulator's parameter configuration needs to be reinitialized due to the interruption of the regulation process to ensure its continuation. The progress indicator records the execution progress of the regulation command sequence at the moment of interruption, the remaining duration of the incomplete feedback control cycle records the remaining time of the feedback control cycle, and the waveform breakpoint position of the magnetic field gradient control command records the waveform position of the magnetic field gradient control command at the moment of interruption. For example, at the moment of interruption, if the regulation command sequence has reached step n, the progress indicator is n; the feedback control cycle has t1 time remaining; and the waveform of the magnetic field gradient control command is interrupted at the m-th data point, the waveform breakpoint position is m.

[0146] Step S1400: Recalibrate the reference sampling time of the array-type energy sensor according to the progress indicator and the waveform breakpoint position. The calibration offset of the reference sampling time is determined by the difference between the interruption duration and the ion beam recovery stabilization time.

[0147] Recalibrating the reference sampling time of the array-type energy sensor based on progress indicators and waveform breakpoint locations ensures that the acquired energy data is synchronized with the actual state of the ion beam. The calibration offset of the reference sampling time is determined by the difference between the interruption duration and the ion beam stabilization time, allowing for accurate energy data acquisition once the ion beam has stabilized. For example, if the interruption duration is t2 and the ion beam stabilization time is t3, the calibration offset Δt = t2 - t3 shifts the reference sampling time of the array-type energy sensor backward by Δt time.

[0148] Step S1500: Based on the calibrated sampling time, continue to execute the multi-level energy regulation command before the interruption, and temporarily adjust the ion source current step size to the preset ratio of the original step size within the first complete feedback control cycle to accelerate system stabilization.

[0149] The multi-level energy regulation commands executed before the interruption are continued based on the calibrated sampling time, allowing the regulation process to continue. Temporarily adjusting the ion source current step size to a preset ratio of the original step size within the first complete feedback control cycle can accelerate the ion beam energy regulation speed, enabling the system to reach a stable state more quickly. For example, if the original ion source current step size is ΔI0 and the preset ratio is α, the ion source current step size is adjusted to α*ΔI0 within the first complete feedback control cycle.

[0150] Step S1600: Trigger a full-energy range scan command at the preset calibration time point, and control the ion beam energy regulator to output a continuous energy beam from the minimum allowable energy to the maximum allowable energy in a linear increment mode.

[0151] The preset calibration time point is a pre-defined time for energy calibration. At this time, a full-energy-range scan command is triggered, controlling the ion beam energy conditioner to output a continuous energy beam from the minimum allowable energy to the maximum allowable energy in a linear increment mode. This allows for comprehensive testing and calibration of the ion beam's performance across the entire energy range. For example, if the preset calibration time point is 2:00 AM daily, after triggering the full-energy-range scan command, the ion beam energy conditioner starts from the minimum allowable energy Emin and gradually increases the energy output linearly until the maximum allowable energy Emax is reached.

[0152] Step S1700: Synchronously collect actual energy distribution parameters at different energy levels using an array-type energy sensor. The actual energy distribution parameters include the average beam intensity, spatial distribution symmetry, and focus drift corresponding to each energy level.

[0153] By synchronously acquiring actual energy distribution parameters at different energy levels using an array of energy sensors, the performance of the ion beam at various energy levels can be accurately understood. The average beam current intensity reflects the average energy carrying capacity of the ion beam at that energy level, the spatial distribution symmetry reflects the spatial symmetry of the ion beam energy distribution, and the focus drift reflects the change in the ion beam focus position at different energy levels. For example, as the ion beam energy gradually increases from Emin to Emax, beam current intensity data at that energy level is acquired at preset energy intervals using an array of energy sensors, and its average value is calculated; the symmetry of the energy distribution is analyzed to obtain the spatial distribution symmetry; and the change in focus position is recorded to obtain the focus drift.

[0154] Step S1800: Compare the actual energy distribution parameters with the expected parameters in the theoretical distribution model item by item to generate an energy calibration offset set. The offset set includes the beam intensity deviation vector, the symmetry error scalar, and the focus drift direction indicator.

[0155] The theoretical distribution model is a pre-established model that includes the expected parameters of the ion beam at different energy levels. The actual energy distribution parameters are compared item by item with the expected parameters in the theoretical distribution model to identify the differences and generate a set of energy calibration offsets. The beam intensity deviation vector reflects the deviation between the mean beam intensity and the expected value, the symmetry error scalar reflects the error between the spatial distribution symmetry and the expected value, and the focus drift direction indicator indicates the direction of focus drift. For example, for a certain energy level E, the expected mean beam intensity in the theoretical distribution model is I0, while the actual average beam intensity is I, so the beam intensity deviation vector is I-I0; the expected spatial distribution symmetry in the theoretical distribution model is S0, while the actual spatial distribution symmetry is S, so the symmetry error scalar is S-S0; the focus drift direction is recorded to obtain the focus drift direction indicator.

[0156] Step S1900: Correct the reference current intensity mapping table in the ion beam energy reference model according to the beam intensity deviation vector. The correction amount is the integral average value of the deviation vector in each energy range.

[0157] The beam current deviation vector reflects the difference between the actual beam current intensity and the expected value. Correcting the reference current intensity mapping table in the ion beam energy reference model based on this deviation vector allows the reference model to more accurately reflect the actual performance of the ion beam. The correction amount is the integral average of the deviation vector across each energy range, thus comprehensively considering the deviation at different energy levels. For example, integrating the beam current deviation vector over the energy range from Emin to Emax and then dividing by the length of the energy range yields the integral average ΔI. Adding ΔI to the reference current intensity corresponding to each energy level in the reference current intensity mapping table completes the correction.

[0158] Step S2000: Adjust the feedback gain coefficient of the adaptive compensation algorithm based on the symmetry error scalar. The adjustment direction is opposite to the sign of the error scalar, and the adjustment magnitude is non-linearly positively correlated with the absolute value of the error scalar.

[0159] The symmetry error scalar reflects the error between the spatial distribution symmetry and the expected value. Adjusting the feedback gain coefficient of the adaptive compensation algorithm based on this error scalar can improve the effectiveness of adaptive compensation. The adjustment direction is opposite to the sign of the error scalar. When the symmetry error is positive, it indicates that the actual symmetry is larger than the expected value, and the feedback gain coefficient needs to be decreased; when the symmetry error is negative, it indicates that the actual symmetry is smaller than the expected value, and the feedback gain coefficient needs to be increased. The adjustment magnitude is non-linearly positively correlated with the absolute value of the error scalar; the larger the error, the larger the adjustment magnitude. For example, if the symmetry error scalar is ΔS, and the initial feedback gain coefficient of the adaptive compensation algorithm is K0, the adjusted feedback gain coefficient K = K0 - k * f(|ΔS|), where k is a constant, and f(|ΔS|) is a function that is non-linearly positively correlated with |ΔS|.

[0160] Step S2100: When the focus drift direction indicator points to the same quadrant X times consecutively, the ion optics system hardware self-test process is triggered to check the mechanical positioning accuracy of the focusing lens and the impedance value of the electromagnetic coil, and the waveform amplitude compensation coefficient of the magnetic field gradient control command is updated according to the impedance value change, X≥3.

[0161] The focus drift direction indicator is used to indicate the direction of the ion beam focus drift in space, usually represented by quadrants. In a Cartesian coordinate system, the four quadrants represent different directional regions. When the focus drift direction indicator points to the same quadrant X times consecutively, this phenomenon is not accidental and likely indicates a problem with the ion optics system hardware. Normally, focus drift should be random or fluctuate within a small range, while X consecutive points to the same quadrant indicate a systematic deviation.

[0162] After triggering the hardware self-test process of the ion optics system, the first step is to check the mechanical positioning accuracy of the focusing lens. The focusing lens plays a crucial role in the ion optics system; it focuses the ion beam by generating a predetermined magnetic field, ensuring the ion beam accurately reaches the target position. Deviations in mechanical positioning accuracy can alter the magnetic field distribution, thus affecting the focusing effect of the ion beam. To check the mechanical positioning accuracy of the focusing lens, high-precision laser measuring equipment can be used to assess its positioning accuracy by measuring the deviation between the actual and theoretical positions of the focusing lens. For example, a laser interferometer can be used to measure the positional deviation of the focusing lens along three coordinate axes. If the deviation exceeds a preset accuracy threshold, it indicates a potential problem with the mechanical positioning of the focusing lens, requiring adjustment or repair. Simultaneously, the impedance value of the electromagnetic coil should be checked. The electromagnetic coil is a key component in generating the magnetic field, and changes in its impedance directly affect the strength and distribution of the magnetic field. With increased usage time, the impedance value of the electromagnetic coil may change due to temperature variations, aging, etc. The impedance value of the electromagnetic coil can be checked using professional impedance measuring instruments, such as a multimeter. The measured actual impedance value is compared with the initial design value to calculate the change in impedance. For example, if the initial design impedance of the electromagnetic coil is R0, and the actual measured impedance is R1, then the impedance change ΔR = |R1 - R0|.

[0163] The waveform amplitude compensation coefficient of the magnetic field gradient control command is updated based on the change in impedance. The magnetic field gradient control command is used to adjust the magnetic field distribution of the focusing lens to achieve focusing and energy regulation of the ion beam. Changes in impedance lead to changes in the magnetic field response characteristics, therefore, the waveform amplitude of the magnetic field gradient control command needs to be compensated. The waveform amplitude compensation coefficient can be updated linearly or nonlinearly, depending on the relationship between the change in impedance and the magnetic field response characteristics. For example, if the change in impedance is linearly related to the change in magnetic field strength, the waveform amplitude compensation coefficient can be updated using a simple linear formula: K' = K + α * ΔR, where K is the initial waveform amplitude compensation coefficient, K' is the updated coefficient, and α is a proportionality constant related to the system characteristics.

[0164] After completing the hardware self-test and updating the waveform amplitude compensation coefficients, the ion beam energy regulation system needs to be recalibrated and tested. By sending a series of test signals, the focusing effect and energy distribution of the ion beam are observed to verify whether the hardware problem has been resolved and whether the waveform amplitude compensation coefficient update is effective. If the test results are still unsatisfactory, further inspection and adjustment of hardware components or re-optimization of the waveform amplitude compensation coefficient update algorithm may be necessary.

[0165] Figure 2 A hardware entity diagram of a computer system provided as an embodiment of the present invention, such as... Figure 2As shown, the hardware entity of the computer system 1000 includes a processor 1001 and a memory 1002, wherein the memory 1002 stores a computer program that can run on the processor 1001, and the processor 1001 executes the program to implement the steps in the method of any of the above embodiments.

[0166] The above description is merely an embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for adaptive control of ion beam energy in FIB equipment, characterized in that, Includes the following steps: The energy distribution parameters of the ion beam in the focusing region are obtained, including the ion beam current intensity, energy attenuation gradient, and spatial distribution characteristics. Based on the energy distribution parameters and the preset ion beam energy reference model, the target energy adjustment range is determined, which includes the allowable peak energy deviation threshold and the minimum energy gradient change rate. Based on the target energy adjustment range, multi-level energy adjustment commands are generated, and the output parameters of the ion beam energy regulator are controlled by the multi-level energy adjustment commands so that the current energy distribution of the ion beam enters the target energy adjustment range. Real-time detection of energy fluctuation data of the adjusted ion beam on the sample surface, the energy fluctuation data including instantaneous energy offset and energy diffusion coefficient at the focal point; Based on the energy fluctuation data and the preset adaptive compensation algorithm, the feedback control parameters of the ion beam energy regulator are adjusted so that the energy distribution of the ion beam is maintained within the target energy regulation range. The step of obtaining the energy distribution parameters of the ion beam in the focusing region includes the following steps: The original energy signals of the ion beam at different angles in the focusing region are synchronously acquired by a ring-shaped multi-axis energy detection array. Each detection unit of the multi-axis energy detection array corresponds to a preset radial detection angle, and the signal acquisition areas of adjacent detection units partially overlap. The original energy signal is denoised in the time domain to eliminate high-frequency electromagnetic interference components and generate a denoised energy waveform cluster. The denoised energy waveform cluster includes waveform data and spatial location labels corresponding to each detection unit. Select the reference waveform with the largest energy integral value among all detection units; During the steady-state phase of the reference waveform, an energy sustaining interval is extracted, and the average value of the waveform peaks within the energy sustaining interval is calculated as the ion beam intensity. The attenuation phase of the reference waveform is divided into multiple sampling windows with equal time intervals; Calculate the rate of decrease of the energy peak within each sampling window, and take the weighted average of the rate of decrease for all sampling windows as the energy decay gradient; Spatial interpolation calculation is performed between the spatial location labels of each detection unit and the corresponding waveform energy integral value to generate a three-dimensional energy density surface covering the focused area. In the three-dimensional energy density surface, the spatial coordinates of the energy density extreme point are identified, and a density gradient scan is performed from the extreme point to the surrounding area to determine the boundary surface range where the energy density drops to a preset critical value. The spatial coordinates of the extreme point are defined as the center position of the spatial distribution feature, and the maximum radial extension distance of the boundary surface range is defined as the diffusion radius of the spatial distribution feature. The energy distribution parameter is a combination of the ion beam intensity, energy decay gradient, and spatial distribution characteristics including the center position and diffusion radius.

2. The method according to claim 1, characterized in that, The step of determining the target energy adjustment range based on the energy distribution parameters and a preset ion beam energy reference model includes the following steps: The deviation of the ion beam current intensity is compared with the reference current intensity recorded in the ion beam energy reference model to generate a current intensity deviation rate, and an initial peak energy deviation threshold is selected from a preset dynamic adjustment rule table based on the current intensity deviation rate. Based on the energy decay gradient and the gradient history curve stored in the ion beam energy reference model, the gradient deviation is calculated, whereby the gradient deviation is the maximum difference between the current gradient and the historical gradient curve within the same time window. The initial peak energy deviation threshold is nonlinearly corrected based on the gradient deviation. If the gradient deviation exceeds a preset gradient stability range, the initial peak energy deviation threshold is reduced by the square of the gradient deviation to generate a corrected peak energy deviation threshold. Extract the diffusion radius from the spatial distribution features, input it into a pre-configured focusing accuracy compensator, output an accuracy compensation coefficient that is negatively correlated with the diffusion radius, and scale the preset minimum energy gradient change rate in the ion beam energy reference model according to the accuracy compensation coefficient. The modified peak energy deviation threshold and the scaled minimum energy gradient change rate are combined to form the target energy adjustment range, wherein the peak energy deviation threshold is used to constrain the amplitude variation range of the ion source current adjustment command, and the minimum energy gradient change rate is used to define the step accuracy threshold of the focusing lens magnetic field adjustment.

3. The method according to claim 2, characterized in that, The generation of multi-level energy regulation commands based on the target energy regulation range includes: The target energy adjustment range is decomposed into multiple energy adjustment stages, each corresponding to a different energy adjustment priority. Specifically: the first adjustment stage prioritizes adjusting the ion beam current intensity to a preset reference intensity range; the second adjustment stage adjusts the ion beam pulse width according to the energy decay gradient, so that the energy decay rate matches the minimum energy gradient change rate; and the third adjustment stage adjusts the magnetic field strength of the focusing lens based on the diffusion radius of the spatial energy distribution matrix, so that the position coordinates of the maximum energy density region coincide with the preset focusing center point. A corresponding time window is allocated to each energy regulation stage, and a corresponding regulation command sequence is generated within each time window. The regulation command sequence includes: Ion source current adjustment command, used to control the output current of the ion emitter; Pulse timing adjustment commands are used to modify the rise time and duration of the ion beam pulse; Magnetic field gradient control commands are used to change the excitation current of the focusing lens to adjust the magnetic field distribution. During the execution of multi-level energy regulation commands, the completion status of each regulation stage is monitored in real time. If the regulation error of a certain stage exceeds the stage's allowable error threshold, the subsequent stages are paused and the error compensation subprocess is initiated.

4. The method according to claim 3, characterized in that, The process of allocating a corresponding time window for each energy regulation stage and generating a corresponding regulation command sequence within each time window includes: The priority order of each energy adjustment stage is determined based on the adjustment urgency weights of the peak energy deviation threshold and the minimum energy gradient change rate within the target energy adjustment range. The adjustment urgency weights are calculated by multiplying the influence coefficient of the peak energy deviation threshold on system stability and the sensitivity coefficient of the minimum energy gradient change rate on focusing accuracy. An initial time window is allocated to the first adjustment stage based on the priority order. The duration of the initial time window is inversely proportional to the difference between the peak energy deviation threshold and the current ion beam intensity. A dynamic time window is allocated to the second adjustment stage. The start time of the dynamic time window is determined by the sum of the end time of the initial time window and a preset buffer interval. Based on the diffusion radius of the spatial distribution characteristics and the maximum adjustment rate of the focusing lens, the length of the time window for the third adjustment stage is calculated. The length of the time window is proportional to the square root of the diffusion radius, and the start time is immediately after the end time of the dynamic time window. Within the initial time window of the first adjustment phase, an ion source current adjustment command sequence is generated. The step size of the ion source current adjustment command sequence is dynamically adjusted by the difference between the peak energy deviation threshold and the current ion beam intensity, and the interval between adjacent commands is inversely proportional to the rate of change of the difference. Within the dynamic time window of the second adjustment phase, a pulse timing adjustment command sequence is generated, wherein the trigger interval of the pulse timing adjustment command is determined by the ratio of the energy decay gradient to the minimum energy gradient change rate, and the duration of each pulse timing adjustment command is proportional to the length of the dynamic time window. Within the time window of the third adjustment stage, a magnetic field gradient control command group is generated. The amplitude of each control signal in the magnetic field gradient control command group is determined by the diffusion radius and the magnetic field response sensitivity of the focusing lens, and the alternation frequency of the control signals is proportional to the reciprocal of the time window length. The fluctuation amplitude of the ion beam intensity after the execution of the command within each time window is monitored in real time. If the fluctuation amplitude exceeds the preset ratio of the peak energy deviation threshold, the current time window is shortened and a transition buffer command is inserted. The duration of the transition buffer command is inversely proportional to the square root of the fluctuation amplitude exceeding the value. After all time window instructions have been executed, the minimum energy gradient change rate is recalibrated based on the change in diffusion radius of the adjusted spatial distribution characteristics. If the calibrated change rate exceeds the allowable error of the target energy adjustment range, the time window reallocation process is triggered.

5. The method according to claim 3, characterized in that, The real-time detection of energy fluctuation data of the adjusted ion beam on the sample surface includes: An array of energy sensors is arranged on the sample surface. The array of energy sensors consists of multiple micro-sensor units distributed in a grid pattern, and each micro-sensor unit covers a preset detection area. The output signals of all micro-sensor units are collected synchronously to generate an energy distribution heat map. Edge detection is performed on the energy distribution heatmap to identify the location and shape characteristics of energy anomaly regions; The instantaneous energy offset is extracted from the energy anomaly region, where the instantaneous energy offset is the difference between the maximum and minimum energy values ​​within the current detection period; Calculate the variance of the energy distribution heatmap and normalize the variance to the energy diffusion coefficient at the focal point; The shape characteristics of the energy anomaly region are matched with a preset defect pattern library. If a known energy diffusion pattern is matched, a corresponding diffusion type identifier is generated and associated with the energy fluctuation data.

6. The method according to claim 5, characterized in that, The step of adjusting the feedback control parameters of the ion beam energy regulator based on the energy fluctuation data and a preset adaptive compensation algorithm includes: The adaptive compensation algorithm compares the instantaneous energy offset with the peak energy deviation threshold in the target energy adjustment range. When the absolute value of the offset exceeds the preset offset threshold, an error polarity identifier and an amplitude weight are generated. The error polarity identifier indicates the direction of energy offset, and the amplitude weight is the ratio of the absolute value of the offset to the offset threshold. The direction of ion source current adjustment is determined based on the error polarity identifier. At the same time, the current adjustment step size is calculated according to the amplitude weight. The current adjustment step size is proportional to the square root of the amplitude weight. The feedback control cycle is shortened based on the reciprocal of the amplitude weight, so that the current adjustment rate in the feedback control parameters is nonlinearly related to the error intensity. The energy diffusion coefficient at the focal point is input into the diffusion trend prediction model in the adaptive compensation algorithm, and the diffusion runaway index is output. When the diffusion runaway index exceeds the minimum energy gradient change rate constraint of the target energy adjustment range, the magnetic field gradient compensation mechanism is triggered. In the magnetic field gradient compensation mechanism, a gradient control command containing magnetic field intensity increment and phase delay is generated based on the diffusion type identifier matching magnetic field waveform correction strategy, and the magnitude of the magnetic field intensity increment and phase delay is scaled by the magnetic field response coefficient in the feedback control parameter. The instantaneous energy offset and the energy diffusion coefficient at the focal point are collected in real time after adjustment, and a composite convergence index is calculated. The composite convergence index is composed of the product of the rate of decrease of the offset and the rate of decay of the diffusion coefficient fluctuation. When the composite convergence index is lower than the preset index threshold, the current adjustment step size and the magnetic field response coefficient are iteratively updated. The step size is reduced according to the exponential decay rule, and the magnetic field response coefficient is linearly adjusted according to the rate of decrease of the diffusion runaway index. When the instantaneous energy offset is within the peak energy deviation threshold and the diffusion runaway index is lower than the minimum energy gradient change rate constraint in N consecutive iterations, the current adjustment step size, feedback period and magnetic field response coefficient in the feedback control parameters are frozen to generate a lock signal, where N > 1.

7. The method according to claim 6, characterized in that, The method also includes an energy stability verification step: After adjusting the feedback control parameters, a preset number of energy distribution sampling cycles are started, and multiple sets of energy fluctuation data are continuously collected in each sampling cycle. Calculate the standard deviation of each set of energy fluctuation data, and use the median of all standard deviations as the current energy stability index; If the current energy stability index is lower than the preset stability threshold, it is determined that the ion beam energy adjustment is complete, and the current parameter configuration of the ion beam energy regulator is locked. If the current energy stability index is higher than the stability threshold, the adaptive compensation algorithm is re-executed, and the compensation intensity of the magnetic field gradient compensation mechanism is increased in each iteration until the maximum number of iterations is reached or the current energy stability index meets the threshold.

8. The method according to claim 7, characterized in that, After locking the current parameter configuration of the ion beam energy regulator, the following steps are performed: The instantaneous value of vacuum chamber pressure, stable value of ion source temperature and circulation rate of cooling system at the moment of locking are recorded in real time. An environmental parameter snapshot is generated and the environmental parameter snapshot is bound to the ion source current step size, magnetic field response coefficient and pulse synchronization delay in the current parameter configuration and stored in the historical adjustment database. Based on the energy distribution comparison diagram before and after regulation in the energy regulation report, the peak displacement of energy density and the edge energy decay rate are extracted to generate a regulation performance fingerprint. The regulation performance fingerprint includes the peak displacement direction indicator and the decay rate change ratio. The regulation performance fingerprint is associated with the environmental parameter snapshot and stored, and an environmental matching degree tag is added to each record in the historical regulation database. The tag is generated by weighted calculation of vacuum chamber pressure deviation, ion source temperature fluctuation range and cooling rate consistency coefficient. When the initial environmental parameters of a new ion beam processing task are detected, the similarity between them and the environmental matching degree tags of each record in the historical adjustment database is calculated. If there is a record with a similarity exceeding the preset matching threshold, the ion source current step size, magnetic field response coefficient and pulse synchronization delay in that record are directly loaded as the initial parameter configuration. After loading the historical parameter configuration, the step of generating multi-level energy adjustment instructions is skipped, and the process directly enters the step of real-time detection of energy fluctuation data of the adjusted ion beam on the sample surface. The signal acquisition timing of the array energy sensor is also adjusted synchronously according to the loaded pulse synchronization delay.

9. A computer system comprising a memory and a processor, the memory storing a computer program executable on the processor, characterized in that, When the processor executes the program, it implements the steps of the method according to any one of claims 1 to 8.

Images