Method for operating a beam device, computer program for executing the method and beam device, and method for generating a training data set and method for training a machine learning model

Abstract

This invention relates to a method for operating a beam device to obtain information about an object. Furthermore, this invention relates to a computer program product having program code that, when executed, controls the beam device, causing the method for operating the beam device to be performed. Additionally, this invention relates to a method for generating a training dataset for a processing unit and / or a machine learning model. Moreover, this invention relates to a method for training a machine learning model for a beam device. The processing unit determines which of a plurality of machine learning models to use to determine the control values ​​of control parameters. The control values ​​of the control parameters are used to operate the control unit to generate information about the object.

Description

Technical Field

[0001] This invention relates to a method for manipulating a beam device to obtain information about an object. Furthermore, this invention relates to a computer program product having program code that, when executed, controls the beam device, causing the method for manipulating the beam device to be performed. Additionally, this invention relates to a method for generating a training dataset for a processing unit and / or machine learning model of a beam device. Moreover, this invention relates to a method for training a machine learning model for a beam device. Background Technology

[0002] Electron beam devices, particularly scanning electron microscopes (hereinafter referred to as SEM) and / or transmission electron microscopes (hereinafter referred to as TEM), are used to examine objects (hereinafter referred to as samples) to obtain information about the properties and behavior of the objects under certain conditions. In SEM, a beam generator is used to generate an electron beam (hereinafter referred to as a primary electron beam), and a beam guiding system focuses the electron beam onto the object to be examined. Objectives are used for focusing purposes. A deflection device guides the primary electron beam onto the surface of the object to be examined. This is also known as scanning. The area scanned by the primary electron beam is also called the scan area. Here, the electrons of the primary electron beam interact with the object to be examined. Interacting particles and / or interacting radiation are generated due to this interaction. For example, the interacting particles are electrons. In particular, electrons are emitted by the object (so-called secondary electrons), and the electrons of the primary electron beam are backscattered (so-called backscattered electrons). The interacting particles form a so-called secondary particle beam and are detected by at least one particle detector. The particle detector generates a detection signal for generating an image of the object. Thus, an image of the object to be examined is obtained. For example, the interacting radiation is X-ray radiation or cathode ray light. Use at least one radiation detector to detect interacting radiation.

[0003] In the case of TEM, a beam generator is also used to generate a primary electron beam, and a beam guiding system is used to guide the primary electron beam to the object to be inspected. The primary electron beam passes through the object. As the primary electron beam passes through the object, the electrons of the primary electron beam interact with the material of the object. The electrons that have passed through the object are imaged onto a fluorescent screen or onto a detector (e.g., in the form of a camera) by a system including an objective lens. For example, the system may also include a projection lens. Here, imaging can also be performed in the scanning mode of TEM. Typically, this type of TEM is called STEM. Alternatively, it may be possible to use at least one additional detector to detect backscattered electrons at the object and / or secondary electrons emitted by the object to image the object.

[0004] It is known to combine the functions of STEM and SEM in a single particle beam device. Therefore, it is possible to use this particle beam device to perform inspections on objects using SEM and / or STEM functions.

[0005] Furthermore, particle beam devices in the form of ion beam columns are known. Ion beam generators arranged within the ion beam column are used to generate ions for processing objects. For example, materials of the object are ablated or applied to the object during processing. Alternatively or alternatively, ions are used for imaging.

[0006] Furthermore, existing technologies have disclosed practices for analyzing and / or processing objects in particle beam apparatuses using both electrons and ions. For example, electron beam columns with SEM functionality are arranged in the particle beam apparatus. Additionally, ion beam columns, as explained above, are also arranged in the particle beam apparatus. Electron beam columns with SEM functionality are specifically used for further examination of processed or unprocessed objects, and also for processing objects.

[0007] Electron beam apparatuses can be used to image objects with high spatial resolution. Specifically, this is achieved by using a primary electron beam with a very small diameter on the plane of the object. Furthermore, the higher the initial acceleration of the electrons in the primary beam within the electron beam apparatus, and the better the spatial resolution, the better the spatial resolution can be achieved when they are finally decelerated to the desired energy (called the landing energy) in the objective lens or in the region between the objective lens and the object. For example, an accelerating voltage of 2 kV to 30 kV is used to accelerate the electrons in the primary electron beam and guide them through the electron column of the electron beam apparatus. The electrons in the primary electron beam are decelerated to the desired landing energy only in the region between the objective lens and the object, and these electrons are incident on the object at this desired landing energy. For example, the landing energy of the electrons in the primary electron beam is in the range of 10 eV to 30 keV.

[0008] Some objects, due to their structural characteristics, can only be effectively examined in an electron beam apparatus when the electrons in the primary electron beam incident on these objects have only low landing energies (e.g., less than 100 eV). For example, electrons with such low energies ensure that these particular objects are not damaged and / or become uncharged when irradiated by electrons. Furthermore, electrons at such low energies are particularly suitable for obtaining images of the objects to be examined that have high surface sensitivity (i.e., are particularly rich in information about the morphology and / or material of the object's surface).

[0009] When generating an image of an object, the user of the electron beam device always operates with care to obtain the ideal image quality required for the object to be inspected. In other words, the user always hopes to generate an object image with such high quality that it allows the user to effectively analyze the object to be inspected based on the image and the image information it contains. Here, image quality can be determined using, for example, objective criteria. For instance, image quality improves with increased resolution or contrast in the image. Alternatively, image quality can be determined based on subjective criteria. Here, the user determines for themselves whether the obtained image quality is sufficient. However, it is entirely possible in this case that an image quality deemed sufficient by one user may be deemed insufficient by another user. For example, the image quality of the object image can also be determined based on the signal-to-noise ratio (SNR) of the detector signal. An SNR in the range of 0 to 5 indicates insufficient image quality. For example, an SNR in the range of 20 to 40 is considered a good SNR (and therefore a good and sufficient image quality). The direction of the secondary particle beam can also be a measure of image quality. Secondary electrons can be emitted from the object at different solid angles. Furthermore, backscattered electrons can be backscattered at different solid angles at the object. The direction of the secondary particle beam (i.e., the solid angle along which the secondary particle beam extends) can be influenced by tilting the primary electron beam and / or the object relative to the optical axis of the electron beam device. As a result, on the one hand, the direction of the secondary particle beam can be selected such that it is incident on the desired detector. On the other hand, the number of generated secondary electrons and the number of backscattered electrons can be affected by the aforementioned tilt. For example, if the primary electron beam is incident on the object parallel to its lattice, the number of secondary electrons and / or backscattered electrons decreases. The detection signal weakens. This leads to a deterioration in image quality. The number of secondary electrons and the number of backscattered electrons can be increased by setting the tilt of the primary electron beam. Using this setting, crystals with a first orientation and crystals with a second orientation can be distinguished based on the intensity of the detection signal.

[0010] As described above, interacting radiation, such as cathode ray light and X-ray radiation, can also be detected. When interacting radiation is detected, the user of the electron beam device must operate with care to obtain the required quality of the radiation detector's signal representation for the object being inspected, based on the detected interacting radiation. For example, if the radiation detector detects X-ray radiation, the quality of this representation is determined, for example, by a good detection signal from the radiation detector. For example, this radiation detector is implemented as an EDX detector. For example, the quality of this representation is thus affected, on the one hand, by the count rate of the detected X-ray quanta, and on the other hand, by the full width at half maximum (FWHM) of the measured peak in the X-ray spectrum. The quality of the detection signal representation improves as the count rate increases and the FWHM decreases. For example, if the radiation detector detects cathode ray light, the quality of this representation can also be determined, for example, by a good detection signal from the radiation detector. For example, the quality of this representation is determined by the count rate of the photons in the detected cathode ray light. This count rate may be affected by the optical unit applied to the light. Furthermore, the electron beam can be configured to emit as many photons as possible from the object as a whole or to emit as many photons as possible within a specific wavelength range.

[0011] To obtain good image quality and / or a good representation of the detection signal based on the detected interacting radiation (which is generated using an electron beam device), the user of a known electron beam device in the prior art initially selects the desired landing energy of the electrons incident on the object. Subsequently, the user selects the setting of additional control parameters for at least one control unit. For example, control parameters are physical variables, particularly control current or control voltage, and also, for example, ratios of physical variables, particularly the amplification of physical variables. The values ​​of the physical variables can be adjusted at or using the control unit, and these physical variable values ​​control and / or feed the units of the electron beam device in a manner that produces the desired physical effect, such as generating a specific magnetic field and / or electrostatic field.

[0012] The first control parameter of the first control unit sets the contrast in the generated image. In principle, the contrast is the image with the maximum illumination L. 最大 The brightest pixel and the pixel with the lowest illuminance L 最小 The contrast ratio is the difference in brightness (i.e., intensity) between the darkest pixels. A smaller difference in brightness between two pixels means lower contrast. A larger difference in brightness between two pixels means higher contrast. For example, contrast ratio can be specified as Weber contrast or Michelson contrast. Here, the following applies to Weber contrast:

[0013]

[0014] The following applies to Michelson contrast:

[0015]

[0016] Contrast generated primarily by secondary electrons is determined by the surface morphology of the object. Contrast generated primarily by backscattered electrons, on the other hand, is primarily determined by the material of the imaged region of the object. This contrast is also known as material contrast. Material contrast depends on the average atomic number of the imaged region of the object. For example, when a higher gain factor is set at the amplifier of the detector, where the detector is used to detect secondary electrons and / or backscattered electrons, the contrast increases. The amplifier amplifies the detection signal generated by the detector. Similarly, for example, when a smaller gain factor is set at the amplifier of the detector, the contrast decreases.

[0017] The second control parameter of the second control unit sets the brightness in the generated image. In principle, the brightness in the image is related to each pixel in the image. A first pixel with a brightness value higher than a second pixel appears brighter than the second pixel in the image. For example, the brightness is set by setting the gain factor and / or offset and / or signal shift of the detector's detection signal. Here, the brightness of each pixel in the image is increased or decreased by the same amount, which can also be achieved, for example, using a color table stored in a memory unit, where a specific brightness corresponds to a certain color included in the color table.

[0018] The third control parameter of the third control unit is used, for example, to induce an objective lens, which is used to set the focus of a primary electron beam onto an object.

[0019] The fourth control parameter used to actuate the fourth control unit is used to center the primary electron beam in the objective lens. For example, the fourth control unit is used to set the electrostatic and / or magnetic units of the electron beam apparatus, using these units to center the primary electron beam in the objective lens.

[0020] Furthermore, the image quality of the object and / or the quality of the representation of the detection signal based on the detected interactive radiation are affected by a fifth control parameter of the fifth control unit, which controls and sets the electrostatic deflection unit and / or magnetic deflection unit used in the electron beam apparatus for so-called "beam shift". As a result, the position of the scanning area can be set and optionally shifted to a desired position. This can occur without using a sample stage (hereinafter also referred to as an object holder) on which the object is disposed. For example, if the scanning area migrates away from the actual area of ​​the object observed using the electron beam apparatus due to a change in the settings on the electron beam apparatus, the electron beam will be shifted due to a translational movement in the case of "beam shift", in such a way that the scanning area is once again located in the desired observation area.

[0021] Astigmatism correctors used in electron beam apparatuses can also affect the image quality of an object's image and / or the quality of the representation of the detection signal based on the detected interacting radiation. Astigmatism correctors (magnetic multipole elements and / or electrostatic multipole elements) are specifically used to correct astigmatism. The astigmatism corrector can be set by a sixth control unit using a sixth control parameter.

[0022] However, the image quality of the object's image and / or the quality of the representation of the detection signal based on the detected interacting radiation can also be affected by the position of the mechanically displaceable units of the electron beam apparatus. For example, image quality is affected by the position of the aperture used to shape and define the primary electron beam within the electron beam apparatus.

[0023] The image quality of the object and / or the quality of the representation of the detection signal based on the detected interacting radiation may be further affected by so-called scan rotation. Scan rotation is the rotation of the scan area around the optical axis of the electron beam apparatus in the plane of the scan area.

[0024] Therefore, in order to obtain the desired image quality of the object and / or the desired quality of the representation of the detection signal based on the detected interaction radiation, the user should take into account as many of the above-mentioned control parameters and / or other control parameters not specified herein as possible, wherein the physical effects obtained by the various control parameters will in turn affect each other.

[0025] The following methods are known for confirming suitable values ​​for control parameters, and these methods can be used to obtain the desired image quality and / or the desired quality based on the representation of the detected signal of the interacting radiation:

[0026] For example, mathematical models can be used to determine suitable values ​​for various control parameters in order to obtain the desired image quality and / or the desired quality of the representation of the detection signal based on the detected interacting radiation. However, these calculated and theoretical values ​​of the control parameters are often unsuitable for obtaining truly good image quality and / or a truly good representation of the detection signal based on the detected interacting radiation. This may be due to facts such as: for example, not all control parameters are considered in the mathematical model, and / or the mathematical model is based on simplifying assumptions, while reality is much more complex.

[0027] Another known method involves experimentally verifying the values ​​of various control parameters, for example, using reference samples. The verified control parameter values ​​are used to set the control unit of the electron beam apparatus. However, a disadvantage is that the object to be examined and imaged does not always correspond to the reference sample, especially in terms of material composition and morphology. This can lead to optical aberrations, thus degrading the quality of the actually obtained image.

[0028] -Another known method uses a manual search to set the desired image quality and / or the desired representation of the detected interaction radiation based on the object to be imaged. Here, the desired landing energy of the electrons in the primary electron beam is first selected, at which the electrons are incident on the object to be examined. Subsequently, brightness, contrast, focus, centering of the primary electron beam in the objective, beam shift, and / or the position of adjustable units are varied and matched to each other through experimentation until the desired image quality and / or desired representation is obtained. This procedure is very complex because it must be performed for each setting of the landing energy.

[0029] Reference is made to DE 10 2016 208 689 A1, which is prior art. Summary of the Invention

[0030] Therefore, the object upon which this invention is based is specifically to specify a method and a beam device for performing the method, using which control parameter values ​​of the control unit for actuating the beam device can be readily obtained, wherein these control parameter values ​​particularly ensure the desired image quality of the object's image and / or the desired representation of the detection signal based on the detected interactive radiation.

[0031] According to the invention, this object is achieved using a method for operating a beam device to obtain information about an object, wherein the method includes the features of claim 1. The feature of claim 14 provides a computer program product having program code that can be loaded or is loaded into a processor and, when executed, controls the beam device, causing the method of claim 1 to be performed. Furthermore, the feature of claim 15 provides a beam device for generating information about an object. Additionally, the feature of claim 21 provides a computer-implemented method for generating training datasets for (a) the processing unit of the beam device and / or (b) a machine learning model. Moreover, the feature of claim 22 provides a method for training a machine learning model for the beam device. Further features of the invention will become apparent from the following description, the appended claims, and / or the drawings.

[0032] The method according to the invention is used to operate a beam device to obtain information about an object. Examples of beam devices are further discussed below. Furthermore, examples of information about the object are further discussed below.

[0033] The method according to the invention includes the step of obtaining first data from a database connected to the processing unit using a processing unit. The processing unit is associated with a beam assembly. For example, the processing unit may be a processing unit and / or processor of the beam assembly. Alternatively or additionally, the processing unit may be a unit separate from but connected to the beam assembly. The connection between the processing unit and the beam assembly may be wireless or wired. The database may be a database associated with the beam assembly. For example, the database may be a database of the beam assembly. Alternatively or additionally, the database may be a unit separate from but connected to the beam assembly. The connection between the processing unit and the database may be wireless or wired.

[0034] As described above, the first data is obtained from a database. The first data includes object data about the object, data about multiple machine learning models, and first range data about the possible value range of at least one first control parameter for the control unit used to control the beam device.

[0035] Object data may include specific information about the object. For example, object data may include at least one of the following: (a) information about one or more materials contained in the object, (b) the size of the object, and (c) the temperature of the object. This invention is not limited to such object data. Rather, any information suitable for this invention may be used in this invention.

[0036] A machine learning model is a procedure that provides output based on data inputs that the machine learning model has not seen before. Machine learning models are known in this field. Examples of machine learning models are discussed further below.

[0037] As described above, at least one first control parameter is used in the control unit of the control beam device. Examples of the first control parameter are further discussed above and below. The first range data includes all possible value ranges of the first control parameter. Examples of the value ranges of the first control parameter are further discussed below.

[0038] Furthermore, the method according to the invention includes, on one hand, the step of using a processing unit to determine a first machine learning model among a plurality of machine learning models based on first data, and on the other hand, the step of providing object data and a first range data as input data to the first machine learning model. Additionally, the method according to the invention includes the step of using the first machine learning model to determine a control value for a first control parameter, wherein the control value of the first control parameter is the output of the first machine learning model. For example, the machine learning model uses object data and the first range data to determine the control value of the first control parameter. The control value of the first control parameter is used to provide information about the object at a later stage of the method according to the invention. Therefore, the information about the object can meet a given criterion.

[0039] The method according to the invention further includes the step of using a processing unit to obtain second data from a database. The second data includes object data about an object, data about multiple machine learning models, and second range data about the possible value range of at least one second control parameter for a control unit used to control the beam assembly. The object data and machine learning models are further discussed above and below. As described above, at least one second control parameter is used for a control unit to control the beam assembly. Examples of the second control parameter are further discussed above and below. The second range data includes all possible value ranges of the second control parameter. Examples of the value ranges of the second control parameter are further discussed below.

[0040] Furthermore, the method according to the invention includes the step of using a processing unit to determine a second machine learning model among a plurality of machine learning models based on second data. The determination of the second machine learning model may additionally be based on the determined control value of a first control parameter. Additionally, the method according to the invention includes the step of providing object data and second range data as input data to the second machine learning model. The method according to the invention further includes the step of using the second machine learning model to determine the control value of a second control parameter. The control value of the second control parameter is the output of the second machine learning model. For example, the machine learning model uses object data and second range data to determine the control value of the second control parameter. The control value of the second control parameter is used to provide information about the object at a later stage of the method according to the invention. Therefore, the information about the object can meet a given criterion.

[0041] Furthermore, the method according to the invention includes the step of operating the control unit using control values ​​of a determined first control parameter and a determined second control parameter. Additionally, the method according to the invention includes the step of generating information about an object using a beam device and a control unit of the beam device. Examples of information about the object are further discussed below.

[0042] In embodiments of the method according to the invention, the first machine learning model and the second machine learning model can be the same. Therefore, the method according to the invention uses the same machine learning model for several control parameters (e.g., the first control parameter and the second control parameter). The method steps according to the invention regarding several control parameters can be performed simultaneously.

[0043] In another embodiment of the method according to the invention, the first machine learning model is different from the second machine learning model.

[0044] Compared to existing technologies, the present invention provides a processing unit that determines which of a plurality of machine learning models to use to determine the control value of a first control parameter and the control value of a second control parameter. The processing unit makes its decision based on first data on one hand and second data on the other (particularly the determined control value of the first control parameter) on the other. Therefore, the present invention provides that the processing unit selects a specific machine learning model from a plurality of machine learning models that can determine specific control values ​​of the first and second control parameters. Those specific control values ​​of the first and second control parameters are used to operate the control unit to generate information about an object, particularly within a relatively short time and / or with good quality and / or with the desired quality. The method according to the invention also provides that each machine learning model used in the method according to the invention can use different possible ranges of values ​​for the first and second control parameters when determining their control values. Therefore, the processing unit can always select a specific machine learning model from a plurality of machine learning models, wherein, for a specific range of possible values ​​of the first or second control parameter, the performance of the specific machine learning model may be superior to that of the other machine learning models among the plurality of machine learning models.

[0045] In embodiments of the method according to the invention, additionally or alternatively, the step of generating information about the object includes generating an image of the object. For example, an optical microscope, a laser beam device, and / or a particle beam device are used to generate the image of the object. Additionally or alternatively, the step of generating information about the object includes generating a spectral analysis of the object. For example, a detector is used to detect the interaction radiation between the beam of the beam device and the object. The interaction radiation can be X-ray radiation or cathode ray light. For example, a radiation detector is used to detect the interaction radiation. Further examples are discussed above or below.

[0046] In another embodiment of the method according to the invention, it is additionally or alternatively provided that the step of determining a first machine learning model and / or a second machine learning model among a plurality of machine learning models includes determining one of the following machine learning models: a heuristic autotuning algorithm, a supervised learning algorithm, and a reinforcement algorithm. In other words, the plurality of machine learning models may include a heuristic autotuning algorithm, a supervised learning algorithm, and / or a reinforcement algorithm. The above-mentioned algorithms are well known in the art. Therefore, they are only briefly discussed below:

[0047] A heuristic autotuning algorithm can provide control values ​​for a first control parameter and a second control parameter to obtain information about the object based on (a) object data and (b) values ​​included in a first range of data relating to the possible range of values ​​for the first control parameter and / or values ​​included in a second range of data relating to the possible range of values ​​for the second control parameter. Furthermore, the heuristic autotuning algorithm can evaluate the quality of the obtained information about the object. Additionally, the heuristic autotuning algorithm selects optimal values ​​from the control values ​​obtained as a result of the algorithm; these control values ​​are used to control the control unit to provide information about the object that meets, for example, good quality criteria.

[0048] A supervised learning algorithm provides a model that predicts control values ​​for a first control parameter and / or a second control parameter as the result of this model. These control values ​​are used to control the control unit to provide information about the object, which satisfies, for example, good quality standards. Object data and a first range of possible values ​​for the first control parameter and / or a second range of possible values ​​for the second control parameter are used as inputs to this model.

[0049] The reinforcement algorithm is a trained algorithm. In the first step, this algorithm acquires control values ​​for a first control parameter and / or a second control parameter, wherein these control values ​​are used to control the control unit to provide information about the object. Furthermore, in the second step, this algorithm evaluates the acquired information and selects the next value in a first range of possible values ​​for the first control parameter and / or the next value in a second range of possible values ​​for the second control parameter. The first and second steps are repeated until the acquired information about the object is considered sufficiently good and / or the variations in the values ​​of the first and / or second ranges are minimal.

[0050] It should be noted that the present invention is not limited to the machine learning model described above. Rather, any machine learning model suitable for the present invention can be used.

[0051] In another embodiment of the method according to the invention, it is additionally or alternatively provided that the method according to the invention comprises at least one of the following:

[0052] (i) Use the first physical variable as the first control parameter;

[0053] (ii) Use a first control current or a first control voltage as a first control parameter;

[0054] (iii) Use the first ratio of the physical variable as the first control parameter; and

[0055] (iv) Use the first amplification of the physical variable as the first control parameter.

[0056] In embodiments of the method according to the invention, it is additionally or alternatively provided that the method according to the invention comprises at least one of the following:

[0057] (i) Use the second physical variable as the second control parameter;

[0058] (ii) Use a second control current or a second control voltage as a second control parameter;

[0059] (iii) Use the second ratio of the physical variable as the second control parameter; and

[0060] (iv) Use the second amplification of the physical variable as the second control parameter.

[0061] In another embodiment of the method according to the invention, it is additionally or alternatively provided that the method according to the invention comprises at least one of the following:

[0062] (i) One of the following is used as a first control parameter: a first contrast parameter, which sets the contrast of information about an object; a first brightness parameter, which sets the brightness of information about an object; a first actuation parameter, which is used to actuate the objective lens of the beam apparatus; a first setting parameter, which is used to set the electrostatic unit and / or magnetic unit of the beam apparatus; a first controllability parameter, which is used to control and set the electrostatic deflection unit and / or magnetic deflection unit of the beam apparatus to achieve beam shift of the beam apparatus; and a first astigmatism parameter, which sets the astigmatism of the beam apparatus.

[0063] (ii) One of the following is used as a second control parameter: a second contrast parameter, which sets the contrast of information about the object; a second brightness parameter, which sets the brightness of information about the object; a second actuation parameter, which is used to actuate the objective lens of the beam apparatus; a second setting parameter, which is used to set the electrostatic unit and / or magnetic unit of the beam apparatus; a second controllability parameter, which is used to control and set the electrostatic deflection unit and / or magnetic deflection unit of the beam apparatus to achieve beam shift of the beam apparatus; and a second astigmatism parameter, which sets the astigmatism of the beam apparatus.

[0064] This invention is not limited to the examples given with respect to the first control parameter and / or the second control parameter. Rather, any control parameter suitable for this invention can be used. For example, the following control parameters can be used as the first control parameter and / or the second control parameter:

[0065] - Control parameters can set the contrast of an object, an information provided by the generated image. In principle, contrast is the contrast ratio of the image with the maximum illumination L. 最大 The brightest pixel and the pixel with the lowest illuminance L 最小 The contrast ratio is the difference in brightness (i.e., intensity) between the darkest pixels. A smaller difference in brightness between two pixels means lower contrast. A larger difference in brightness between two pixels means higher contrast. For example, contrast ratio can be specified as Weber contrast ratio or Michelson contrast ratio. Here, the following applies to Weber contrast ratio:

[0066]

[0067] The following applies to Michelson contrast:

[0068]

[0069] If an image has already been generated using a particle beam apparatus, the contrast generated primarily by secondary electrons is determined by the surface morphology of the object. Contrast generated primarily by backscattered electrons, on the other hand, is primarily determined by the material of the imaged region of the object. This contrast is also known as material contrast. Material contrast depends on the average atomic number of the imaged region of the object. For example, when a higher gain factor is set at the amplifier of the detector used to detect secondary electrons and / or backscattered electrons, the contrast increases. The amplifier amplifies the detection signal generated by the detector. Similarly, for example, when a smaller gain factor is set at the amplifier of the detector, the contrast decreases;

[0070] - Additional control parameters can be set regarding the brightness of the object, which is provided as part of the generated image. In principle, the brightness in the image is related to each pixel in the image. A first pixel with a brightness value higher than a second pixel appears brighter than the second pixel in the image. For example, brightness can be set by setting the gain factor and / or offset and / or signal shift of the detector signal of the beam device. Here, the brightness of each pixel in the image is increased or decreased by the same amount, which can also be achieved, for example, using a color table stored in a memory unit, where a specific brightness corresponds to a certain color included in the color table;

[0071] - Additional control parameters can be used, for example, for the objective lens of the actuating beam device, wherein the objective lens is used to set the focus of the beam of the beam device on the object;

[0072] - Another control parameter can be used to center the beam of the beam assembly within the objective lens of the beam assembly. For example, this control parameter can be used to set the electrostatic and / or magnetic units of the beam assembly, using these units to center the particle beam within the objective lens;

[0073] The image quality of the object and / or the quality of the representation of the detection signal based on the detected interaction radiation are affected by control parameters used to control and set the electrostatic deflection units and / or magnetic deflection units used in a beam apparatus in the form of a particle beam device for so-called “beam shifting.” As a result, the position of the scanning area can be set and optionally shifted to a desired position. This can occur without using a sample stage (also known as an object holder) on which the object is disposed. For example, if the scanning area migrates away from the actual area of ​​the object observed using the particle beam device due to a change in the settings on the particle beam device, a single particle beam of the particle beam device will be shifted due to a translational movement in the case of “beam shifting,” in such a way that the scanning area is once again located in the desired observation area.

[0074] - Astigmatism correctors used in beam apparatuses in the form of particle beam devices can also affect the image quality of an object's image and / or the quality of the representation of the detection signal based on the detected interacting radiation. Astigmatism correctors (magnetic multipole elements and / or electrostatic multipole elements) are specifically used to correct astigmatism. Astigmatism correctors can be set using astigmatism control parameters;

[0075] Information about an object can also be affected by the position of the mechanically displaceable units of the beam assembly. For example, the quality of information is affected by the position of the aperture used to shape and define the beam within the beam assembly.

[0076] In another embodiment of the method according to the invention, it is additionally or alternatively provided that the method according to the invention comprises at least one of the following:

[0077] (i) The range between the minimum and maximum values ​​of the first control parameter is used as the possible value range of the first control parameter, wherein the minimum and maximum values ​​of the first control parameter depend on the configuration of the control unit;

[0078] (ii) Using the range between the minimum and maximum values ​​of a first contrast parameter as the possible value range of a first control parameter configured as the first contrast parameter, the first contrast parameter setting the contrast of information about an object, wherein the minimum and maximum values ​​of the first contrast parameter depend on the configuration of the control unit;

[0079] (iii) The range between the minimum and maximum values ​​of a first brightness parameter is used as the possible value range of a first control parameter, which is configured as the first brightness parameter, which sets the brightness of information about an object, wherein the minimum and maximum values ​​of the first brightness parameter depend on the configuration of the control unit;

[0080] (iv) Using the range between the minimum and maximum values ​​of a first actuation parameter as the possible value range of a first control parameter, the first control parameter being configured as the first actuation parameter, the first actuation parameter being used to actuate the objective lens of the beam apparatus, wherein the minimum and maximum values ​​of the first actuation parameter depend on the magnification of the beam apparatus;

[0081] (v) Use the range between the minimum and maximum values ​​of the second control parameter as the possible value range of the second control parameter, wherein the minimum and maximum values ​​of the second control parameter depend on the configuration of the control unit;

[0082] (vi) The range between the minimum and maximum values ​​of the second contrast parameter is used as the possible value range of the second control parameter, which is configured as the second contrast parameter, which sets the contrast of information about the object, wherein the minimum and maximum values ​​of the second contrast parameter depend on the configuration of the control unit.

[0083] (vii) The range between the minimum and maximum values ​​of the second brightness parameter is used as the possible value range of the second control parameter, which is configured as the second brightness parameter, which sets the brightness of information about the object, wherein the minimum and maximum values ​​of the second brightness parameter depend on the configuration of the control unit;

[0084] (viii) The range between the minimum and maximum values ​​of the second actuation parameter is used as the possible value range of the second control parameter, which is configured as the second actuation parameter for actuating the objective lens of the beam apparatus, wherein the minimum and maximum values ​​of the second actuation parameter depend on the magnification of the beam apparatus.

[0085] In embodiments of the method according to the invention, additional or alternatively, additional control parameters are used to generate information about the object. Specifically, the information about the object generated above is first information about the object. The method according to the invention further includes the step of using a processing unit to determine whether the first information about the object meets a desired quality standard. The desired quality standard may be given by a user or may be determined as further outlined above or below. If the desired quality standard is not met, the following steps of the method according to the invention are performed:

[0086] (a) The processing unit obtains third data from a database connected to the processing unit. The third data includes object data about the object, data about multiple machine learning models, and third range data about the possible value range of at least one third control parameter for the control unit of the beam control device. The third control parameter can be any control parameter for the control unit of the beam control device, such as at least one of the control parameters mentioned above or below;

[0087] (b) Using a processing unit, a third machine learning model is determined from among multiple machine learning models based on third data. The determination of the third machine learning model may additionally be based on the control values ​​of the determined first control parameter and / or the control values ​​of the determined second control parameter. The third machine learning model may be one of the machine learning models mentioned above or below;

[0088] (c) Provide the object data and the third-range data as input data to the third machine learning model;

[0089] (d) Use a third machine learning model to determine the control value of the third control parameter, wherein the control value of the third control parameter is the output of the third machine learning model;

[0090] (e) Operate the control unit using the control value of the determined third control parameter; and

[0091] (f) Using the beam device and the control unit of the beam device to generate second information about the object.

[0092] Therefore, the second information can be generated by operating the control unit using the control value of the determined first control parameter, the control value of the determined second control parameter, and / or the control value of the determined third control parameter.

[0093] In embodiments of the method according to the invention, additionally or alternatively, the step of generating second information about the object includes generating an image of the object. For example, an optical microscope, a laser beam device, and / or a particle beam device are used to generate the image of the object. Additionally or alternatively, the step of generating second information about the object includes generating a spectral analysis of the object. For example, a detector is used to detect the interaction radiation between the beam of the beam device and the object. The interaction radiation can be X-ray radiation or cathode ray light. For example, a radiation detector is used to detect the interaction radiation. Further examples are discussed below.

[0094] In embodiments of the method according to the invention, it is additionally or alternatively provided that the method comprises at least one of the following:

[0095] (i) Using an optical microscope as a beam device;

[0096] (ii) Using a laser beam device as a beam device;

[0097] (iii) Use a particle beam device as a beam device;

[0098] (iv) Using an electron beam device as a beam device; and

[0099] (v) Use an ion beam device as a beam device.

[0100] A particle beam apparatus may include at least one beam generator for generating a particle beam comprising charged particles. For example, the charged particles are electrons and / or ions. Further, the particle beam apparatus may include at least one guiding unit for guiding the particle beam onto an object. A guiding unit should be understood to refer to any unit for guiding the particle beam onto an object, and also to a unit for shaping the particle beam that is then guided onto the object. For example, the guiding unit is implemented as an objective lens for focusing the particle beam onto the object, an electrostatic and / or magnetic unit for beam shaping or beam guiding, an astigmatism corrector, a focusing lens, or a mechanically adjustable aperture unit by means of which the particle beam is defined. The charged particles may have landing energy when incident on the object. In other words, the landing energy of the charged particles is the energy required to examine and / or image the object. The landing energy of the charged particles may differ from the energy required to guide the charged particles through the beam column of the particle beam apparatus. Specifically, it is provided that the charged particles are initially accelerated very strongly and then decelerated to landing energy only just before incident on the object. For example, the landing energy of charged particles ranges from 1 eV to 30 keV.

[0101] The particle beam device may also include at least one control unit for setting the guiding unit by selecting at least one control parameter of the control unit. For example, the control parameter is a physical variable, particularly a control current or control voltage, and also, for example, a ratio of the physical variable, particularly the amplification of the physical variable. The values ​​of the physical variables can be adjusted at the control unit or adjusted using the control unit, and these physical variable values ​​control and / or feed the guiding unit of the particle beam device in a manner that produces a desired physical effect, such as generating a specific magnetic field and / or electrostatic field.

[0102] Furthermore, the particle beam apparatus may include at least one detector for detecting interacting particles and / or interacting radiation emitted from the interaction between the particle beam and the object when the particle beam is incident on the object. For example, interacting particles are secondary particles emitted by the object (e.g., secondary electrons), particles backscattered at the object (also called backscattered particles), and / or scattered particles, for example, those transmitted through the object in the beam direction. For example, backscattered particles are backscattered electrons. For example, interacting radiation is X-ray radiation or cathode ray light. For example, a radiation detector is used to detect interacting radiation.

[0103] Furthermore, the particle beam device may include at least one display unit for displaying an image of an object and / or a representation of data about the object, wherein the image and / or the representation is generated by means of detection signals generated by detecting interacting particles and / or interacting radiation.

[0104] All the above and following embodiments of the method according to the invention are not limited to the order of the explained method steps. The invention also includes different orders of method steps suitable for achieving the objectives of the invention. Additionally or alternatively, at least two method steps are provided to be performed in parallel in the method according to the invention. Furthermore, the above and following embodiments of the method according to the invention are not limited to all of the method steps described above or below. In particular, it is intended to omit individual or several method steps from the above or below method steps in other embodiments.

[0105] The present invention also relates to a computer program product comprising program code loadable or loadable into a processor of a beam apparatus. When executed in the processor, the program code controls the beam apparatus to perform a method having at least one of the features described above or below, or a combination of at least two of the features described above or below. In other words, the present invention also relates to a non-volatile computer-readable medium comprising software loadable or loadable into a processor of a beam apparatus. When executed in the processor, the software controls the beam apparatus to perform a method having at least one of the features described above or below, or a combination of at least two of the features described above or below. The software includes executable code for performing at least one method step of the method steps explained above or below.

[0106] Therefore, the present invention also relates to a processor configured to perform a method having at least one of the above-described or below features, or a combination of at least two of the above-described or below features.

[0107] The present invention further relates to a beam apparatus for generating information about an object, wherein the beam apparatus has been explained above and will be further described in detail below. This will be briefly summarized below. The beam apparatus according to the invention includes at least one beam generator for generating a beam, at least one objective lens for focusing the beam onto an object, at least one detector unit for detecting interacting particles and / or interacting radiation generated due to the interaction between the beam and the object, and at least one processor loaded with a computer program product having the features mentioned above or below.

[0108] In embodiments of the beam apparatus according to the invention, the beam apparatus additionally or alternatively includes (a) an optical microscope for imaging and / or analyzing an object and / or (b) a laser beam apparatus for imaging, processing and / or analyzing an object.

[0109] In embodiments of the beam apparatus according to the invention, the beam apparatus additionally or alternatively includes a particle beam device for imaging, processing, and / or analyzing an object. For example, the beam generator is a particle beam generator for generating a particle beam containing charged particles. These charged particles may be electrons and / or ions. Furthermore, the objective lens may be an objective lens for focusing the particle beam onto the object, and the detector unit may be a detector unit for detecting interacting particles and / or interacting radiation generated due to the interaction between the particle beam and the object. Additionally, the particle beam apparatus may include at least one scanning device for scanning the particle beam over the object.

[0110] In yet another embodiment of the particle beam apparatus according to the invention, additionally or alternatively provided, the particle beam generator is a first particle beam generator, and the particle beam is a first particle beam having first charged particles. The objective lens is a first objective lens for focusing the first particle beam onto an object. Furthermore, the particle beam apparatus according to the invention includes at least one second beam generator for generating a second particle beam having second charged particles. Additionally, the particle beam apparatus according to the invention includes at least one second objective lens for focusing the second particle beam onto an object.

[0111] Specifically, the beam device according to the invention is provided as an electron beam device and / or an ion beam device.

[0112] The present invention also relates to a computer-implemented method for generating a training dataset for (a) a processing unit associated with a beam assembly and / or (b) a machine learning model, the machine learning model being used to determine control values ​​of a first control parameter to operate a control unit of the beam assembly and to determine control values ​​of a second control parameter to operate the control unit of the beam assembly. This training dataset is used to train the processing unit and / or the machine learning model. The computer-implemented method includes the step of determining the control value of the first control parameter using a method including at least one of the features described above or a combination of at least two of the features described above or a combination ...

[0113] All the above and following embodiments of the computer-implemented method according to the present invention are not limited to the order of the explained method steps. The present invention also includes different orders of method steps suitable for achieving the objectives of the invention. Additionally or alternatively, the computer-implemented method according to the present invention is provided with the parallel execution of at least two method steps. Furthermore, the above and following embodiments of the computer-implemented method according to the present invention are not limited to all of the above or following method steps. In particular, it is intended to omit individual or several method steps from the above or following method steps in other embodiments.

[0114] The present invention also relates to another method for training a machine learning model for a beam device to identify images of objects that meet quality criteria given by a user. This other method may be a computer-implemented method. This other method is used to train the machine learning model. The beam device may be any beam device as described above or below. This other method includes the following method steps:

[0115] A user of the beam apparatus provides a first image of an object, which meets a quality standard. For example, image quality can be determined based on subjective criteria. Here, the user determines for themselves whether the obtained image quality is sufficient. Alternatively, the image quality of the object image can also be determined based on the signal-to-noise ratio (SNR) of the detector signal. An SNR in the range of 0 to 5 indicates insufficient image quality. For example, an SNR in the range of 20 to 40 is considered a good SNR (and therefore a good and sufficient image quality). The direction of the secondary particle beam can also be a measure of image quality. Secondary electrons can be emitted from the object at different solid angles. Furthermore, backscattered electrons can be backscattered at the object at different solid angles. The direction of the secondary particle beam (i.e., the solid angle along which the secondary particle beam extends) can be influenced by tilting the primary electron beam and / or the object relative to the optical axis of the electron beam apparatus. As a result, on the one hand, the direction of the secondary particle beam can be selected such that it is incident on the desired detector. On the other hand, both the number of generated secondary electrons and the number of backscattered electrons can be influenced by the aforementioned tilting. For example, if a primary electron beam is incident on an object parallel to its crystal lattice, the number of secondary electrons and / or backscattered electrons decreases. The detection signal weakens, leading to a degraded image quality. The number of secondary electrons and backscattered electrons can be increased by adjusting the tilt of the primary electron beam. Using this setting, crystals with a first orientation can be distinguished from those with a second orientation based on the intensity of the detection signal. As described above, it is also possible to detect interacting radiation, such as cathode rays and X-rays.

[0116] - Provide at least one first degraded image by degrading the first image using a processing unit associated with the beam device, and provide at least one second degraded image by degrading the first image using a processing unit associated with the beam device;

[0117] - Users evaluate whether at least one first degraded image meets the quality standard, and the process of evaluating at least one first degraded image determines whether at least one first degraded image is marked.

[0118] - Users evaluate whether at least one second degraded image meets the quality standard, and the at least one second degraded image is marked according to the steps of evaluating at least one second degraded image;

[0119] - A machine learning model is used to simulate the steps of evaluating at least one first degraded image and evaluating at least one second degraded image, wherein the first image, at least one first degraded image and at least one second degraded image are used as training data for the machine learning model, and the simulation steps are based on the first image, at least one first degraded image and at least one second degraded image.

[0120] - Optionally, the user of the beam device uses the beam device to provide a second image of the object, wherein the second image meets quality standards;

[0121] - Provide at least one third degraded image by degrading the first image and / or the second image using a processing unit associated with the beam device, and provide at least one fourth degraded image by degrading the first image and / or the second image using a processing unit associated with the beam device;

[0122] - A machine learning model is used to evaluate whether at least one third degraded image meets the quality standard, and at least one third degraded image is labeled according to the step of evaluating at least one third degraded image, wherein the step of evaluating at least one third degraded image and the step of labeling at least one third degraded image are repeated until the labeling is verified to be correct using a processing unit associated with the beam device;

[0123] - A machine learning model is used to evaluate whether at least one fourth degraded image meets the quality standard, and at least one fourth degraded image is labeled according to the step of evaluating at least one fourth degraded image, wherein the step of evaluating at least one fourth degraded image and the step of labeling at least one fourth degraded image are repeated until the labeling is verified to be correct using a processing unit associated with the beam device;

[0124] - The user and / or the processing unit associated with the beam device provides at least one third image of the object; and

[0125] - A machine learning model is used to evaluate whether at least one third image meets the quality criteria, and at least one third image is labeled according to the step of evaluating at least one third image, wherein the step of evaluating at least one third image and the step of labeling at least one third image are repeated until the labeling is verified to be correct using a processing unit.

[0126] After the machine learning model has been trained using this other method, it can identify whether an image of an object meets the provided quality standards without having to input any additional reference images into the machine learning model.

[0127] All the above and following embodiments of this other method according to the invention are not limited to the order of the explained method steps. The invention also includes different orders of method steps suitable for achieving the objectives of the invention. Additionally or alternatively, at least two method steps are provided to be performed in parallel in this other method according to the invention. Furthermore, the above and following embodiments of this other method according to the invention are not limited to all of the method steps described above or below. In particular, it is intended to omit individual or several method steps from the above or below method steps in other embodiments. Attached Figure Description

[0128] Further practical embodiments and advantages of the present invention are described below with reference to the accompanying drawings:

[0129] Figure 1 A first embodiment of the beam device according to the present invention is shown;

[0130] Figure 2 A second embodiment of the beam device according to the present invention is shown;

[0131] Figure 3 A third embodiment of the beam device according to the present invention is shown;

[0132] Figure 4 A schematic diagram of the operating system for the beam manipulation device is shown;

[0133] Figure 5 A schematic diagram of a machine learning model is shown;

[0134] Figure 6 A flowchart illustrating an embodiment of a method for operating a beam apparatus is shown schematically;

[0135] Figure 7 A flowchart illustrating another embodiment of a method for operating a beam apparatus is shown schematically;

[0136] Figure 8 A flowchart illustrating a method for generating training datasets for processing units and / or machine learning models associated with a beam device is schematically shown; and

[0137] Figure 9 A flowchart illustrating a method for training a machine learning model for a beam device is shown schematically. Detailed Implementation

[0138] The invention will now be explained in more detail with the aid of beam devices in the form of SEMs and in the form of combined devices including electron beam columns and ion beam columns. It is clearly stated that the invention can be used in any beam device, particularly any optical microscope, laser beam device, and / or particle beam device.

[0139] Figure 1 A schematic diagram of a first embodiment of a particle beam apparatus according to the present invention is shown, which is in the form of an SEM 100. The SEM 100 has a beam generator 1 with an electron source, an extraction electrode 2, a control electrode 3, and an anode 4. The anode 4 forms the source-side end of the beam guide tube 21 of the SEM 100. The beam generator 1 is configured, for example, as a thermal field emitter. Alternatively, the beam generator 1 is configured, for example, as a thermal tungsten emitter or a LaB6 emitter.

[0140] Electrons emitted from beam generator 1 form a primary electron beam. The electrons are accelerated to the anode potential due to the potential difference between beam generator 1 and anode 4. For example, the potential of anode 4 is positive relative to the potential of beam generator 1, ranging from 1 kV to 30 kV, so that the electrons have kinetic energy in the range of 1 keV to 30 keV.

[0141] Viewed from the anode 4 and along the optical axis 20 in the direction of the objective lens 10, the SEM 100 has a first condenser lens 5 in front and a second condenser lens 6 behind. An aperture unit 7 is arranged in the beam guide tube 21 between the first condenser lens 5 and the second condenser lens 6. Figure 1 In the SEM 100 shown, the objective lens 10 is configured as a magnetic lens with pole shoes 22 having pole shoe gaps 23. A ring coil 11 is arranged in the pole shoes 22 to generate a magnetic field for the objective lens 10.

[0142] Starting from the second focusing lens 6 and viewed in the direction of the objective lens 10, a guide device in the form of a deflection device is arranged along the optical axis 20 of the SEM 100. This guide device has a first guide device in the form of a first deflection device 9 and a second guide device in the form of a second deflection device 12. The first deflection device 9 is arranged on the source side of the objective lens 10. On the other hand, the second deflection device 12 is arranged inside the objective lens 10 on the object side, on the beam guide tube 21. The first deflection device 9 and the second deflection device 12 are interleaved deflection devices. In other words, both the first deflection device 9 and the second deflection device 12 are configured to deflect the primary electron beam in two non-parallel directions aligned perpendicular to the optical axis 20. For example, the first deflection device 9 and / or the second deflection device 12 are magnetic deflection devices. In particular, the first deflection device 9 and / or the second deflection device 12 have, for example, four air coils arranged around the optical axis 20 of the SEM 100. Alternatively or alternatively, the first deflection device 9 and / or the second deflection device 12 are electrostatic deflection devices. In particular, the first deflection device 9 and / or the second deflection device 12 have, for example, four electrodes arranged around the optical axis 20 of the SEM 100, to which different electrostatic potentials can be applied.

[0143] Objective lens 10 is arranged on object chamber 13. Specifically, objective lens 10 protrudes into the interior space of object chamber 13 through an opening in object chamber 13. A movable object stage 19 is arranged in the interior space of object chamber 13. Object 15 can be arranged on object stage 19.

[0144] Using objective lens 10, a primary electron beam generated by beam generator 1 and formed using first condenser lens 5 and / or second condenser lens 6 is focused onto object plane 16. Appropriate excitation of first deflection device 9 and second deflection device 12 ensures that the primary electron beam can be deflected perpendicular to the optical axis 20 of SEM 100 on object plane 16, allowing scanning of the surface of object 15 arranged on object plane 16 by different deflections of the primary electron beam. Electrons from the primary electron beam interact with object 15. As a result of the interaction, electrons are emitted from object 15 (so-called secondary electrons), and electrons from the primary electron beam are backscattered (so-called backscattered electrons). Secondary electrons and backscattered electrons are detected and used for image generation. Thus, an image of the object 15 to be examined is obtained. Furthermore, interaction radiation (e.g., X-rays or cathode rays) is generated during the interaction, which is detected and subsequently evaluated for analysis of object 15.

[0145] To detect the aforementioned interacting particles and / or the aforementioned interacting radiation, a first detector unit 14 is arranged, for example, in the object chamber 13. Alternatively, a second detector unit 8 for detecting the aforementioned interacting particles is arranged, for example, in the beam guide tube 21 in the region between the first deflection device 9 and the second focusing lens 6. A radiation detector 500 is arranged in the object chamber 13 to detect the interacting radiation (e.g., X-rays and / or cathode rays) generated when a primary electron beam strikes the object 15.

[0146] exist Figure 1 In the embodiment of the SEM 100 shown, a pressure-level aperture holder 17 is provided, which can be arranged on the pole shoe 22 of the objective lens 10 protruding into the object chamber 13. The pressure-level aperture holder 17 has a pressure-level aperture with an aperture 18. For example, additional pressure-level aperture units can be arranged within the beam guide tube 21 of the SEM 100. These are in Figure 1 Not shown in the image. Figure 1 Vacuum pumps are not shown in the diagram. These vacuum pumps are intended to generate and maintain the vacuum within the beam guide tube 21 and the object chamber 13 as desired for the operation of the SEM 100.

[0147] If the SEM 100 operates under high vacuum in the object chamber 13, the pressure-stage aperture holder 17 is not necessary and can therefore be removed from the pole shoe 22 of the objective lens 10. On the other hand, if the SEM 100 operates at a relatively high pressure (in the range of about 1 Pa to 3000 Pa) in the object chamber 13, the pressure-stage aperture holder 17 can be mounted on the pole shoe 22 of the objective lens 10 so that a sufficiently good vacuum can be maintained in the beam guide tube 21 by differential pumping despite the high pressure in the object chamber 13.

[0148] Specifically, the first detector unit 14, the second detector unit 8, the first deflection device 9, and the second deflection device 12 are connected to a control device 123, which includes a monitor 124. Specifically, the first deflection device 9 is connected to the control device 123 via a first signal connection 800. The first signal connection 800 can be a physical connection (e.g., a cable) and / or a wireless connection (e.g., a radio communication system and / or a wireless local area network). The second deflection device 12 is connected to the control device 123 via a second signal connection 801. The second signal connection 801 can be a physical connection (e.g., a cable) and / or a wireless connection (e.g., a radio communication system and / or a wireless local area network). The control device 123 processes the detection signals generated by the first detector unit 14 and the second detector unit 8, and displays these detection signals as images on the monitor 124. The control device 123 further includes a database 126, in which data is stored and data is read from. Additionally, the control device 123 is connected to other units of the SEM 100. This is in Figure 1 This was not further shown in the text.

[0149] The control device 123 of the SEM 100 includes a processor 127. A computer program product including program code is loaded into the processor 127. When executed, the program code performs methods for operating the SEM 100. This will be explained in more detail below.

[0150] In SEM 100, distance A can be set using the control device 123 of SEM 100. Distance A is given by: (a) the object distance between the outer boundary of the objective lens 10 of SEM 100 and the object 15; or (b) the focal plane distance between the outer boundary of the objective lens 10 of SEM 100 and the focal plane of the objective lens 10. The aforementioned distance A according to case (a) or case (b) is also referred to as the working distance. For example, distance A in case (a) is set by moving the object stage 19 and / or moving the objective lens 10 using the moving device 25. In particular, distance A in case (b) is adjusted by changing the excitation of the objective lens 10 along the optical axis 20 of SEM 100.

[0151] Figure 2 A schematic diagram of another SEM 100 is shown. This other SEM 100 has a first beam generator in the form of an electron source 101, which is configured as a cathode. Furthermore, the other SEM 100 is provided with an extraction electrode 102 and an anode 103, which is placed at one end of a beam guide tube 104 of the other SEM 100. For example, the electron source 101 is configured as a thermal field emitter. However, the invention is not limited to this electron source 101. Rather, any electron source suitable for the invention can be used.

[0152] Electrons emitted from electron source 101 form a primary electron beam. The electrons are accelerated to the anode potential due to the potential difference between electron source 101 and anode 103. In the embodiment shown here, the anode potential is 100 V to 35 kV, for example 5 kV to 15 kV, particularly 8 kV, relative to the ground potential of the housing 120. However, alternatively, the anode potential can be at ground potential.

[0153] Two focusing lenses, namely a first focusing lens 105 and a second focusing lens 106, are arranged on the beam guide tube 104. Viewed from the electron source 101 and in the direction of the first objective lens 107, the first focusing lens 105 is positioned in front, followed by the second focusing lens 106. It is explicitly noted that another embodiment of the SEM 100 may have only a single focusing lens. A first aperture unit 108 is arranged between the anode 103 and the first focusing lens 105. The first aperture unit 108, together with the anode 103 and the beam guide tube 104, is at a high voltage potential (i.e., the potential of the anode 103) or ground potential. The first aperture unit 108 has a plurality of first apertures 108A, in Figure 2 One of these is shown. For example, there are two first aperture stops 108A. Each of the plurality of first aperture stops 108A has a different aperture diameter. Using an adjustment mechanism (not shown), the desired first aperture stop 108A can be adjusted to the optical axis OA of another SEM 100. It is explicitly noted that in another embodiment, the first aperture stop unit 108 may be provided with only a single first aperture stop 108A. In this embodiment, an adjustment mechanism may not be provided, and the first aperture stop unit 108 is configured in a fixed manner. A fixed second aperture stop unit 109 is arranged between the first condenser lens 105 and the second condenser lens 106. Alternatively, the second aperture stop unit 109 may be movable.

[0154] The first objective lens 107 has pole shoes 110, which have apertures. The beam guide tube 104 is guided through these apertures. Coils 111 are arranged in the pole shoes 110.

[0155] An electrostatic deceleration system is arranged in the lower region of the beam guide tube 104. The electrostatic deceleration system has a single electrode 112 and a tubular electrode 113. The tubular electrode 113 is arranged at one end of the beam guide tube 104, which faces the object 125 arranged at the movable object holder 114.

[0156] The tubular electrode 113, together with the beam guide tube 104, is at the potential of the anode 103, while the single electrode 112 and the object 125 are at a potential lower than that of the anode 103. In this case, this is the ground potential of the shell of the object chamber 120. In this way, the electrons of the primary electron beam can be slowed down to the energy desired for examining the object 125.

[0157] The object 125 and the single electrode 112 can also be at different potentials and potentials different from ground potential. This allows for adjustment of the deceleration position of the primary electron beam relative to the object 125. For example, if deceleration is performed very close to the object 125, the imaging error is reduced.

[0158] Another SEM 100 also includes a guiding system in the form of deflection devices, comprising a first guiding device in the form of a first deflection device 130 and a second guiding device in the form of a second deflection device 115. The first deflection device 130 is arranged within the first objective lens 107 on the source side. On the other hand, the second deflection device 115 is arranged within the first objective lens 107 on the object side, on the beam guide tube 104. The first deflection device 130 and the second deflection device 115 are interleaved deflection devices. In other words, both the first deflection device 130 and the second deflection device 115 are configured to deflect the primary electron beam in two non-parallel directions aligned perpendicular to the optical axis OA of the other SEM 100. For example, the first deflection device 130 and / or the second deflection device 115 are magnetic deflection devices. In particular, the first deflection device 130 and / or the second deflection device 115 have, for example, four air coils arranged around the optical axis OA of the other SEM 100. Alternatively or alternatively, the first deflection device 130 and / or the second deflection device 115 are electrostatic deflection devices. Specifically, the first deflection device 130 and / or the second deflection device 115 have four electrodes arranged, for example, around the optical axis OA of the SEM 100, to which different electrostatic potentials can be applied. By means of the first deflection device 130 and the second deflection device 115, a primary electron beam is deflected and can be scanned over the object 125. Electrons from the primary electron beam interact with the object 125. Due to this interaction, interacting particles are generated, and these interacting particles are detected. Specifically, electrons are emitted from the surface of the object 125 as interacting particles (so-called secondary electrons), or electrons from the primary electron beam are backscattered (so-called backscattered electrons).

[0159] A detector system is arranged in the beam guide tube 104, comprising a first detector 116 and a second detector 117 for detecting secondary electrons and / or backscattered electrons. In the beam guide tube 104, the first detector 116 is arranged along the optical axis OA on the source side, while the second detector 117 is arranged along the optical axis OA on the object side. The first detector 116 and the second detector 117 are offset from each other in the direction of the optical axis OA of the SEM 100. Both the first detector 116 and the second detector 117 have openings through which the primary electron beam can pass. The first detector 116 and the second detector 117 are approximately at the potentials of the anode 103 and the beam guide tube 104, and the optical axis OA of the SEM 100 passes through the corresponding openings.

[0160] The second detector 117 is primarily used to detect secondary electrons. These secondary electrons initially possess low kinetic energy and arbitrary direction of motion upon exiting the object 125. Due to a strong extraction field emanating from the tubular electrode 113, the secondary electrons are accelerated toward the first objective lens 107. The secondary electrons enter the first objective lens 107 in an approximately parallel manner. Within the first objective lens 107, the beam diameter of the secondary electrons remains relatively small. The first objective lens 107 strongly acts on the secondary electrons, thereby generating a relatively short focal point at a sufficiently steep angle relative to the optical axis OA, causing the secondary electrons to diverge after the focal point and collide with the second detector 117 on its effective surface. In contrast, electrons backscattered from the object 125 (i.e., backscattered electrons, which possess relatively higher kinetic energy compared to the secondary electrons upon exiting the object 125) are detected by the second detector 117 to a very small extent. The high kinetic energy of the backscattered electrons upon exiting the object 125 and the angle relative to the optical axis OA cause the beam waist of the backscattered electrons (i.e., the region with the smallest diameter beam) to be located near the second detector 117. Most of the backscattered electrons pass through the opening of the second detector 117. Therefore, the first detector 116 is mainly used to detect backscattered electrons.

[0161] In another embodiment of the SEM 100, the first detector 116 may additionally include a reverse field grid 116A. The reverse field grid 116A is arranged on the side of the first detector 116 facing the object 125. The reverse field grid 116A has a negative potential relative to the potential of the beam guide tube 104, such that only backscattered electrons with high kinetic energy pass through the reverse field grid 116A to reach the first detector 116. Alternatively, the second detector 117 may include another reverse field grid designed in the same manner as the aforementioned reverse field grid 116A of the first detector 116 and have similar functionality.

[0162] In addition, another SEM 100 includes a chamber detector 119 (e.g., an Everhart-Thornley detector or an ion detector) in an object chamber 120, which may include a detection surface coated with a light-blocking metal.

[0163] The detection signals generated by the first detector 116, the second detector 117 and the chamber detector 119 are used to generate one or more images of the surface of the object 125.

[0164] It is explicitly stated that the apertures of the first aperture unit 108 and the second aperture unit 109, as well as the openings of the first detector 116 and the second detector 117, are shown in an exaggerated manner. The openings of the first detector 116 and the second detector 117 have a range of 0.5 mm to 5 mm perpendicular to the optical axis OA. For example, these openings are circular in shape and have a diameter in the range of 1 mm to 3 mm perpendicular to the optical axis OA.

[0165] In the embodiment shown here, the second aperture unit 109 is configured as a pinhole aperture unit and is provided with a second aperture 118 for the passage of the primary electron beam. The extension range of the second aperture 118 is in the range of 5 µm to 500 µm, for example, 35 µm. Alternatively, in another embodiment, the second aperture unit 109 is provided with a plurality of apertures that can be mechanically displaced relative to the primary electron beam, or the primary electron beam can be directed to these apertures using electrical deflection elements and / or magnetic deflection elements. The second aperture unit 109 may be a pressure-level aperture. This separates a first region from a second region, in which the electron source 101 is arranged and an ultra-high vacuum (10) exists. -7 hPa to 10 -12 hPa), this second region has a high vacuum (10 hPa), -3 hPa to 10 -7 The second region is the intermediate pressure region of the bundle guide tube 104, which leads to the object chamber 120.

[0166] The object chamber 120 is under vacuum. A pump (not shown) is arranged in the object chamber 120 to generate the vacuum. Figure 2 In the illustrated embodiment, the object chamber 120 operates within a first pressure range or a second pressure range. The first pressure range includes only pressures less than or equal to 10. -3 The pressure is hPa, and the second pressure range only includes pressures greater than 10. -3 The pressure is hPa. The object chamber 120 is vacuum-sealed to ensure these pressure ranges.

[0167] An object holder 114 is arranged on an object stage 122. The object stage 122 is configured to move in three mutually perpendicular directions, namely, in the x-direction (first stage axis), y-direction (second stage axis), and z-direction (third stage axis). Additionally, the object stage 122 can rotate about two mutually perpendicularly arranged rotation axes (stage rotation axes). The invention is not limited to the object stage 122 described above. Rather, the object stage 122 may have additional translational axes and additional rotational axes along or about which it can move.

[0168] Another SEM 100 further includes a third detector 121 arranged in an object chamber 120. Specifically, as viewed from the electron source 101 along the optical axis OA, the third detector 121 is arranged behind the object stage 122. The object stage 122, and therefore the object holder 114, can rotate such that a primary electron beam can strike an object 125 arranged on the object holder 114. As the primary electron beam passes through the object 125 to be examined, the electrons of the primary electron beam interact with the material of the object 125. The third detector 121 detects the electrons that have passed through the object 125 to be examined.

[0169] A radiation detector 500 is arranged in object chamber 120 to detect the interacting radiation (e.g., X-rays and / or cathode rays) generated when an electron beam strikes object 125. Radiation detector 500, a first detector 116, a second detector 117, and chamber detector 119 are connected to a control unit 123, which includes a monitor 124. A third detector 121 is also connected to the control unit 123. This is not shown for clarity. The control unit 123 processes the detection signals generated by the first detector 116, the second detector 117, the chamber detector 119, the third detector 121, and / or the radiation detector 500, and displays these detection signals as images on the monitor 124.

[0170] The first deflection device 130 and the second deflection device 115 are connected to the control device 123. Specifically, the first deflection device 130 is connected to the control device 123 via a first signal connection 800. The first signal connection 800 can be a physical connection (e.g., a cable) and / or a wireless connection (e.g., a radio communication system and / or a wireless local area network). The second deflection device 115 is connected to the control device 123 via a second signal connection 801. The second signal connection 801 can be a physical connection (e.g., a cable) and / or a wireless connection (e.g., a radio communication system and / or a wireless local area network).

[0171] The control unit 123 further includes a database 126, in which data is stored and read from the database. Additionally, the control unit 123 is connected to another unit of another SEM 100. This is not shown in more detail for clarity.

[0172] The control device 123 of the other SEM 100 includes a processor 127. A computer program product including program code is loaded into the processor 127. When the program code is executed, methods for operating the other SEM 100 are performed. This will be explained in more detail below.

[0173] In another SEM 100, distance A can be set using the control device 123 of the other SEM 100. Distance A is given by: (a) the object distance between the single electrode 112 of the other SEM 100 and the object 125; or (b) the focal plane distance between the single electrode 112 of the other SEM 100 and the focal plane of the first objective lens 107. The aforementioned distance A, depending on case (a) or case (b), is also referred to as the working distance. For example, in case (a), distance A is set by moving the object stage 122 and / or by moving the first objective lens 107 using the moving device 25. For example, in case (b), distance A is set by changing the excitation of the first objective lens 107 along the optical axis OA of the other SEM 100.

[0174] Figure 3 A particle beam device in the form of a combination device 200 is shown. The combination device 200 has two particle beam columns. On one hand, the combination device 200 has as already shown... Figure 2 Another SEM 100 is shown, but without the object chamber 120. Instead, another SEM 100 is arranged in the object chamber 201. The object chamber 201 is under vacuum. A pump (not shown) is arranged in the object chamber 201 to generate the vacuum. Figure 3 In the illustrated embodiment, the object chamber 201 operates within a first pressure range or a second pressure range. The first pressure range includes only pressures less than or equal to 10. -3 The pressure is hPa, and the second pressure range only includes pressures greater than 10. -3 The pressure is hPa. The object chamber 201 is vacuum-sealed to ensure these pressure ranges.

[0175] A chamber detector 119 is arranged in the object chamber 201. The chamber detector 119 is configured, for example, as an Everhart-Thornley detector or an ion detector. The chamber detector 119 may have a detection surface coated with a light-blocking metal. In addition, a third detector 121 is arranged in the object chamber 201.

[0176] Another SEM 100 generates the first particle beam (i.e., the primary electron beam described above) and has the optical axis mentioned above, which is located at... Figure 3 Reference numeral 709 is provided in the accompanying drawings and is referred to hereinafter as the first beam axis. Secondly, the assembly 200 is provided with an ion beam device 300, which is also arranged in the object chamber 201. The ion beam device 300 also has an optical axis, which is located at... Figure 3 The figure is provided with reference numeral 710 and is referred to hereinafter as the second beam axis.

[0177] Another SEM 100 is arranged vertically relative to the object chamber 201. In contrast, the ion beam device 300 is arranged at an angle of approximately 0° to 90° relative to the other SEM 100. Figure 3 The diagram shows an arrangement of approximately 50°. The ion beam apparatus 300 includes a second beam generator in the form of an ion beam generator 301. The ion beam generator 301 generates ions that form a second particle beam in the form of an ion beam. These ions are accelerated by means of an extraction electrode 302 at a predetermined potential. The second particle beam then passes through the ion optics of the ion beam apparatus 300, which includes a focusing lens 303 and a second objective lens 304. Finally, the second objective lens 304 generates an ion probe that is focused onto an object 125 disposed on an object holder 114. The object holder 114 is disposed on an object stage 122.

[0178] An adjustable or selectable aperture unit 306 and a guiding system are arranged above the second objective lens 304 (i.e., in the direction of the ion beam generator 301). The guiding system includes a first guiding device in the form of a first deflection device 307 and a second guiding device in the form of a second deflection device 308. The first deflection device 307 is arranged inside the second objective lens 304 on the source side. On the other hand, the second deflection device 308 is arranged inside the second objective lens 304 on the object side. The first deflection device 307 and the second deflection device 308 are interleaved deflection devices. In other words, both the first deflection device 307 and the second deflection device 308 are configured to deflect the ion beam in two non-parallel directions aligned with the optical axis in the form of a second beam axis 710 of the ion beam device 300. For example, the first deflection device 307 and / or the second deflection device 308 are magnetic deflection devices. Specifically, the first deflection device 307 and / or the second deflection device 308 may have four air coils arranged, for example, around an optical axis in the form of a second beam axis 710 of the ion beam device 300. Alternatively or additionally, the first deflection device 307 and / or the second deflection device 308 are electrostatic deflection devices. Specifically, the first deflection device 307 and / or the second deflection device 308 may have four electrodes arranged, for example, around an optical axis in the form of a second beam axis 710 of the ion beam device 300, to which different electrostatic potentials can be applied. Using the first deflection device 307 and the second deflection device 308, the ion beam is deflected and can be scanned over the object 125.

[0179] As explained above, the object holder 114 is arranged on the object stage 122. Figure 3 In the illustrated embodiment, the object stage 122 is configured to move in three mutually perpendicular directions: the x-direction (first stage axis), the y-direction (second stage axis), and the z-direction (third stage axis). Furthermore, the object stage 122 can rotate about two mutually perpendicular rotation axes (stage rotation axes).

[0180] To better represent the various units of the combined device 200, Figure 3 The distances between the individual units of the combined device 200 shown are exaggerated.

[0181] A radiation detector 500 is arranged in the object chamber 201 for detecting interacting radiation (e.g., X-rays and / or cathode rays). The radiation detector 500 is connected to a control device 123, which includes a monitor 124.

[0182] Control device 123 processes data from first detector 116 ( Figure 3 (not shown in the image), second detector 117 ( Figure 3The detectors 119 (not shown), 121 (room detector), and / or 500 (radiation detector) generate detection signals, and these detection signals are displayed on the monitor 124 in the form of images.

[0183] The control device 123 further includes a database 126, in which data is stored and read from the database. Furthermore, the control device 123 is connected to the first deflection device 130 via a first signal connection 800. Figure 3 (not shown in the image) and connected to the second deflection device 115 via the second signal connection 801. Figure 3 (Not shown in the diagram). Additionally, control device 123 is connected to the first deflection device 307 via a first signal connection and to the second deflection device 308 via a second signal connection. Furthermore, control device 123 is connected to another unit of the combination device 200 (not shown here for clarity).

[0184] The control unit 123 of the combined device 200 includes a processor 127. A computer program product having program code is loaded into the processor 127. When the program code is executed, methods for operating the combined device 200 are executed. This will be explained in more detail below.

[0185] The working distance can also be set in the combination device 200. For example, in another SEM 100, the distance A1 can be set using the control device 123. The distance A1 is given by: (a) the object distance between the outer boundary of the first objective lens 107 of the other SEM 100 and the object 125; or (b) the focal plane distance between the outer boundary of the first objective lens 107 of the other SEM 100 and the focal plane of the first objective lens 107. The aforementioned distance A1 according to case (a) or case (b) is also referred to as the working distance. For example, in case (a), the distance A1 is set by moving the object stage 122 and / or moving the first objective lens 107 using the moving device 25. For example, in case (b), the distance A1 is adjusted by changing the excitation of the first objective lens 107 along the first beam axis 709 of the other SEM 100. Further, the distance A2 can be adjusted using the control device 123. Distance A2 is given by: (a) the object distance between the outer boundary of the second objective lens 304 of the ion beam apparatus 300 and the object 125; or (b) the focal plane distance between the outer boundary of the second objective lens 304 of the ion beam apparatus 300 and the focal plane of the second objective lens 304. The aforementioned distance A2 according to case (a) or case (b) is also referred to as the working distance. For example, in case (a), distance A2 is set by moving the object stage 122 and / or by moving the second objective lens 304 using the moving device 25. For example, in case (b), distance A2 is adjusted by changing the excitation of the second objective lens 304 along the second beam axis 710 of the ion beam apparatus 300.

[0186] Figure 4 A schematic diagram of the operating system 600 for operating the beam device 1000 is shown. The beam device 1000 can be an optical microscope, a laser beam device, and / or a particle beam device, particularly... Figure 1 SEM 100, Figure 2 Another SEM 100 and / or Figure 3 The combined device 200. The operating system 600 includes a processing unit 601. The processing unit 601 is associated with the beam device 1000. For example, the processing unit 601 may be a processor 127 as further discussed above. Alternatively or additionally, the processing unit 601 may be a unit separate from but connected to the beam device 1000. The connection between the processing unit 601 and the beam device 1000 may be wireless or wired.

[0187] The operating system 600 also includes a database 606. Database 606 may be a database associated with the beam device 1000. For example, database 606 may be database 126 as further discussed above. Alternatively, database 606 may be a unit separate from but connected to the beam device 1000. The connection between the processing unit 601 and database 606 may be wireless or wired.

[0188] The operating system 600 also includes a control unit 607. At least one control parameter is used to operate the control unit 607. For example, the control parameter is a physical variable, particularly a control current or control voltage, and also, for example, a ratio of the physical variable, particularly the amplification rate of the physical variable. The values ​​of the physical variables can be adjusted at the control unit 607, and these physical variable values ​​control and / or the units of the feed beam device 1000 in a way that produces a desired physical effect, such as generating a specific magnetic field and / or electrostatic field. For example, the control device 123 can be the control unit 607. For example, regarding... Figure 1 In the SEM 100, control device 123 controls a first deflection device 9 and a second deflection device 12, wherein the first deflection device 9 and the second deflection device 12 are connected to control device 123. Specifically, the first deflection device 9 is connected to control device 123 via a first signal connection 800. The first signal connection 800 can be a physical connection (e.g., a cable) and / or a wireless connection (e.g., a radio communication system and / or a wireless local area network). The second deflection device 12 is connected to control device 123 via a second signal connection 801. The second signal connection 801 can be a physical connection (e.g., a cable) and / or a wireless connection (e.g., a radio communication system and / or a wireless local area network).

[0189] The operating system 600 for the beam device 1000 also includes multiple machine learning models 6021 to 602. n The following condition must be met: 1 ≤ n ≤ N, where N is an integer. In other words, multiple 602 machine learning models 6021 to 602... n This may include heuristic autotuning algorithms, supervised learning algorithms, and / or reinforcement algorithms. These algorithms are known in the art, and have been briefly discussed further above. This discussion also applies here.

[0190] Figure 5 Machine learning model 602 is shown. k The diagram illustrates this, where k is an integer satisfying the condition: 1 ≤ k ≤ n. Other machine learning models 6021 to 602... n They can have similar or identical structures.

[0191] Machine learning model 602k Object data 603 is received as input. Object data 603 may include specific information about objects 15 and 125. For example, object data 603 may include at least one of the following: (a) information about one or more materials contained in objects 15 and 125, (b) the size of objects 15 and 125, and (c) the temperature of objects 15 and 125. The invention is not limited to this object data 603. Rather, any information about objects 15 and 125 applicable to the invention can be used in the invention. Machine Learning Model 602 k The device may also receive first range data 604A and / or second range data 604B as input. First range data 604A includes a range of possible values ​​for at least one first control parameter of the control unit 607 for controlling the beam control device 1000. Second range data 604B includes a range of possible values ​​for at least one second control parameter of the control unit 607 for controlling the beam control device 1000. Examples of the first and second control parameters have been further discussed above or below. This discussion also applies here.

[0192] Machine learning model 602 k The control value 605A of the first control parameter and / or the control value 605B of the second control parameter are provided as the machine learning model 602. k The output of the first control parameter 605A and / or the second control parameter 605B is used to operate the control unit 607.

[0193] Embodiments of the present invention include at least one of the following:

[0194] - Use the first physical variable as the first control parameter;

[0195] - Use the first control current or the first control voltage as the first control parameter;

[0196] - Use the first ratio of the physical variables as the first control parameter;

[0197] - Use the first amplification of the physical variable as the first control parameter;

[0198] - Use a second physical variable as the second control parameter;

[0199] - Use the second control current or the second control voltage as the second control parameter;

[0200] - Use the second ratio of the physical variables as the second control parameter; and

[0201] - Use the second magnification of the physical variable as the second control parameter.

[0202] Another embodiment of the present invention includes using one of the following as a first control parameter: a first contrast parameter, which sets the contrast of information about objects 15 and 125; a first brightness parameter, which sets the brightness of information about objects 15 and 125; and a first actuation parameter, which is used to actuate the beam device 1000 (e.g., in a... Figure 1 The form of SEM 100, present Figure 2 Another SEM 100 form and / or present Figure 3 The objective lenses 10, 107, and 304 (in the form of a combined device 200) are also present. Furthermore, the first control parameter can be a first setting parameter used to set the beam assembly 1000 (e.g., in the form of a combined device 200). Figure 1 The form of SEM 100, present Figure 2 Another SEM 100 form and / or present Figure 3 The electrostatic and / or magnetic units 2, 3, 4, 5, 6, 7, 9, 10, 11, 12, 21, 22, 23, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 115, 116A, 130, 302, 303, 304, 306, 307, 308 of the combined device 200. Additionally, the first control parameter can be a first controllable parameter used to control and set the beam device 1000 (e.g., in the form of...). Figure 1 The form of SEM 100, present Figure 2 Another SEM 100 form and / or present Figure 3 The electrostatic deflection units and / or magnetic deflection units 2, 3, 4, 5, 6, 7, 9, 10, 11, 12, 21, 22, 23, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 115, 116A, 130, 302, 303, 304, 306, 307, 308 of the combined device 200 are used to achieve beam shifting of the beam device 1000. Furthermore, the first control parameter can be a first astigmatism correction parameter, which sets the beam device 1000 (e.g., in the form of a combined device 200) to achieve beam shifting of the beam device 1000. Figure 1 The form of SEM 100, present Figure 2 Another SEM 100 form and / or present Figure 3 The image scattering device (in the form of a combination device 200) is a scattering device.

[0203] Another embodiment of the present invention includes using one of the following as a second control parameter: a second contrast parameter, which sets the contrast of information about objects 15 and 125; a second brightness parameter, which sets the brightness of information about objects 15 and 125; and a second actuation parameter, which is used to actuate the beam device 1000 (e.g., in a... Figure 1 The form of SEM 100, present Figure 2 Another SEM 100 form and / or present Figure 3 The objective lenses 10, 107, and 304 (in the form of a combined device 200) are also present. Furthermore, the second control parameter can be a second setting parameter used to set the beam assembly 1000 (e.g., in the form of a combined device 200). Figure 1 The form of SEM 100, present Figure 2 Another SEM 100 form and / or present Figure 3 The electrostatic and / or magnetic units 2, 3, 4, 5, 6, 7, 9, 10, 11, 12, 21, 22, 23, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 115, 116A, 130, 302, 303, 304, 306, 307, 308 of the combined device 200. Additionally, the second control parameter can be a second controllable parameter used to control and set the beam device 1000 (e.g., in the form of...). Figure 1 The form of SEM 100, present Figure 2 Another SEM 100 form and / or present Figure 3 The electrostatic deflection units and / or magnetic deflection units 2, 3, 4, 5, 6, 7, 9, 10, 11, 12, 21, 22, 23, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 115, 116A, 130, 302, 303, 304, 306, 307, 308 of the combined device 200 are used to realize the beam device 1000 (e.g., in the form of...). Figure 1 The form of SEM 100, present Figure 2 Another SEM 100 form and / or present Figure 3 The beam shifting of the beam assembly 1000 (in the form of a combined device 200). Furthermore, the second control parameter can be a second astigmatism parameter, which sets the beam assembly 1000 (e.g., in the form of a combined device 200) to... Figure 1 The form of SEM 100, present Figure 2 Another SEM 100 form and / or present Figure 3 The image scattering device (in the form of a combination device 200) is a scattering device.

[0204] This invention is not limited to the examples given with respect to the first and / or second control parameters. Rather, any control parameters suitable for this invention can be used. Examples of other control parameters have been further discussed above and below. This discussion also applies here.

[0205] In another embodiment of the invention, it is additionally or alternatively provided that the invention includes at least one of the following:

[0206] - Use the range between the minimum and maximum values ​​of the first control parameter as the possible value range of the first control parameter, wherein the minimum and maximum values ​​of the first control parameter depend on the configuration of the control unit 607;

[0207] - The range between the minimum and maximum values ​​of the first contrast parameter is used as the possible value range of the first control parameter, which is configured as the first contrast parameter, which sets the contrast with respect to the information of objects 15 and 125, wherein the minimum and maximum values ​​of the first contrast parameter depend on the configuration of the control unit 607.

[0208] - The range between the minimum and maximum values ​​of the first brightness parameter is used as the possible value range of the first control parameter, which is configured as the first brightness parameter, which sets the brightness of information about objects 15 and 125, wherein the minimum and maximum values ​​of the first brightness parameter depend on the configuration of the control unit 607;

[0209] - The range between the minimum and maximum values ​​of the first actuation parameter is used as the possible value range of the first control parameter, which is configured as the first actuation parameter, used to actuate the beam device 1000 (e.g., in the form of...). Figure 1 The form of SEM 100, present Figure 2 Another SEM 100 form and / or present Figure 3 The objective lenses 10, 107, and 304 of the combined device 200, wherein the minimum and maximum values ​​of the first actuation parameter depend on the magnification of the beam device 1000;

[0210] - Use the range between the minimum and maximum values ​​of the second control parameter as the possible value range of the second control parameter, wherein the minimum and maximum values ​​of the second control parameter depend on the configuration of the control unit 607;

[0211] - The range between the minimum and maximum values ​​of the second contrast parameter is used as the possible value range of the second control parameter, which is configured as the second contrast parameter, which sets the contrast with respect to the information of objects 15 and 125, wherein the minimum and maximum values ​​of the second contrast parameter depend on the configuration of the control unit 607.

[0212] - The range between the minimum and maximum values ​​of the second brightness parameter is used as the possible value range of the second control parameter, which is configured as the second brightness parameter, which sets the brightness of information about objects 15 and 125, wherein the minimum and maximum values ​​of the second brightness parameter depend on the configuration of the control unit 607;

[0213] - The range between the minimum and maximum values ​​of the second actuation parameter is used as the possible value range of the second control parameter, which is configured as the second actuation parameter, used to actuate the beam device 1000 (e.g., in the form of...). Figure 1 The form of SEM 100, present Figure 2 Another SEM 100 form and / or present Figure 3 The objective lenses 10, 107, and 304 (in the form of a combined device 200) wherein the minimum and maximum values ​​of the second actuation parameter depend on the beam assembly 1000 (e.g., in the form of a combined device 200). Figure 1 The form of SEM 100, present Figure 2 Another SEM 100 form and / or present Figure 3 The magnification of the combined device (in the form of 200).

[0214] Now, the following description is in Figure 1 An exemplary embodiment of the method for operating a beam device 1000 according to the present invention, used in SEM 100. It is explicitly stated that the exemplary embodiment of the method according to the present invention can also be similarly used in any beam device 1000, for example for... Figure 2 Another SEM 100 or Figure 3 The combined device 200.

[0215] Figure 6 The schematic diagram illustrates the operation. Figure 1 A flowchart of an embodiment of the SEM 100 method.

[0216] Method step S1 includes obtaining first data from database 606 (which may be database 126) using processing unit 601 (which may be processor 127). The first data includes object data 603 regarding object 15, and machine learning models 6021 to 602 (602 in total). nThe data includes first range data 604A, which specifies the possible value range of at least one first control parameter used to control control unit 607 (which may be control device 123 of SEM 100). Object data may include specific information about object 15. For example, object data may include at least one of the following: (a) information about one or more materials contained in object 15, (b) the size of object 15, and (c) the temperature of object 15. The invention is not limited to this object data 603. Rather, any information about object 15 applicable to the invention may be used in the invention. First range data 604A includes all possible value ranges of the first control parameter. Examples of the first control parameter have been further discussed above and below. This discussion also applies here.

[0217] Method step S2 includes using processing unit 601 to determine multiple machine learning models 6021 to 602 based on the first data. n The first machine learning model in 602 k The steps are as follows, where k is an integer, satisfying the following condition: 1 ≤ k ≤ n.

[0218] Method step S3 includes providing object data 603 and first range data 604A as input data to the first machine learning model 602. k The steps.

[0219] Method step S4 includes using the first machine learning model 602 k The steps for determining the control value 605A of the first control parameter, wherein the control value 605A of the first control parameter is the first machine learning model 602. k The output. For example, the first machine learning model 602. k The control value 605A of the first control parameter is determined using object data 603 and first range data 604A. The control value 605A of the first control parameter is used to provide information about object 15 at a later stage of the method according to the invention.

[0220] Method step S5 includes obtaining second data from database 606 using processing unit 601. The second data includes object data 603 regarding object 15 and machine learning models 6021 to 602 (602 in total). n The data includes second range data 604B regarding the possible value range of at least one second control parameter used to control control unit 607. Second range data 604B includes all possible value ranges of the second control parameter. Examples of second control parameters have been further discussed above and below. This discussion also applies here.

[0221] Method step S6 includes using processing unit 601 to determine multiple machine learning models 6021 to 602 based on the second data. n The second machine learning model 602 m The steps. Determine the second machine learning model 602. m Alternatively, the control value 605A can be determined based on the first control parameter. m is an integer satisfying the following condition: 1 ≤ m ≤ n.

[0222] Method step S7 includes providing object data 603 and second range data 604B as input data to the second machine learning model 602. m The steps.

[0223] Method step S8 includes using a second machine learning model 602 m The steps for determining the control value 605B of the second control parameter, wherein the control value 605B of the second control parameter is the second machine learning model 602. m The output. For example, the second machine learning model 602. m The control value 605B of the second control parameter is determined using object data 603 and second range data 604B. The control value 605B of the second control parameter is used to provide information about object 15 at a later stage of the method according to the invention.

[0224] Method step S9 includes operating the control unit 607 using the determined control value 605A of the first control parameter and the determined control value 605B of the second control parameter.

[0225] Method step S10 includes the step of generating information about object 15 using a beam device 1000 in the form of SEM 100 and a control unit 607 in the form of control device 123.

[0226] In an embodiment of the method according to the invention, method step S10 includes generating an image of object 15. In other words, an image of object 15 is generated using SEM 100. Alternatively or additionally, method step S10 includes generating a spectral analysis of object 15. For example, a radiation detector 500 is used to detect the interacting radiation between the electron beam of SEM 100 and object 15. The interacting radiation may be X-ray radiation or cathode ray light.

[0227] In an embodiment of the method according to the present invention, the first machine learning model 602 k Second machine learning model 602 mThe same model can be used for several control parameters (e.g., a first control parameter and a second control parameter). The method steps according to the invention regarding several control parameters can be performed simultaneously.

[0228] In another embodiment of the method according to the invention, the first machine learning model 602 k Unlike the second machine learning model 602 m .

[0229] The present invention provides a processing unit 601 that determines to use a plurality of machine learning models 6021 to 602. n Which machine learning model 602 is it? i The processing unit 601 determines a control value 605A for a first control parameter and a control value 605B for a second control parameter. The processing unit 601 makes its decision based on both the first data and the second data (specifically, the determined control value 605A of the first control parameter). Therefore, the present invention provides that the processing unit 601 can process multiple machine learning models 6021 to 602... n Select a specific machine learning model 602 i The machine learning model can determine specific control values ​​for the first and second control parameters. These specific control values ​​are used to operate the control unit 607 to generate information about the object 15, particularly within a relatively short time and / or with good quality. The method according to the invention also provides that each machine learning model 6021 to 602 used in the method according to the invention... n Different possible ranges of values ​​for the first and second control parameters can be used when determining the control values ​​605A and 605B of the first and second control parameters. Therefore, the processing unit 601 can always select a specific machine learning model 602 from multiple machine learning models 602. i Among them, for a specific range of possible values ​​for the first control parameter or the second control parameter, a specific machine learning model 602 i Its performance may outperform multiple 602 machine learning models 6021 to 602. n Other machine learning models 6021 to 602 n .

[0230] Figure 7 The schematic diagram illustrates the operation. Figure 1 A flowchart of another embodiment of the SEM 100 method. This other embodiment uses additional control parameters to generate information about object 15. The method steps of this other embodiment can be... Figure 6The method step S10 of the embodiment of the method is executed after this step. The information about object 15 generated in method step S10 is first information about object 15. Method step S11 includes using processing unit 601 to determine whether the first information about object 15 meets a desired quality standard. The desired quality standard may be given by the user or may be determined as discussed further above or below. This discussion also applies here. If the desired quality standard is met, the method may stop. If the desired quality standard is not met, the following steps of the method according to the invention may be performed.

[0231] Method step S12 includes obtaining third data from database 606 using processing unit 601. The third data includes object data about object 15 and data about multiple machine learning models 6021 to 602. n The data includes a third range of possible values ​​for at least one third control parameter used to control the control unit 607, which is in the form of a control device 123 of SEM 100. The third control parameter can be any control parameter used to control the control unit 607, such as at least one of the control parameters mentioned above or below.

[0232] Method step S13 includes using processing unit 601 to determine multiple machine learning models 6021 to 602 based on third data. n The third machine learning model 602 p The steps. Determine the third machine learning model 602. p Alternatively, the control value 605A of the determined first control parameter and / or the control value 605B of the determined second control parameter can be used. p is an integer satisfying the condition: 1 ≤ p ≤ n. Third machine learning model 602 p It can be one of the machine learning models mentioned above or below.

[0233] Method step S14 includes providing object data 603 and third range data as input data to the third machine learning model 602. p The steps.

[0234] Method step S15 includes using a third machine learning model 602 p The steps for determining the control value of the third control parameter, wherein the control value of the third control parameter is determined by the third machine learning model 602. p The output.

[0235] Method step S16 includes operating the control unit 607 using the control value of the determined third control parameter. As described above, the control value 605A of the determined first control parameter and the control value 605B of the determined second control parameter are also used to operate the control unit 607.

[0236] Method step S17 includes generating second information about object 15 using a beam apparatus 1000 in the form of an SEM 100 and a control unit 607. For example, generating an image of object 15. Alternatively or additionally, generating a spectral analysis of object 15. For example, using a radiation detector 500 to detect the interaction radiation between the electron beam of the SEM 100 and object 15. The interaction radiation can be X-ray radiation or cathode ray light.

[0237] Figure 8 The diagram schematically illustrates the processing unit 601 associated with SEM 100 and / or machine learning models 6021 to 602. n A flowchart illustrating a computer-implemented method for generating a training dataset. This training dataset is used to train processing unit 601 and / or machine learning models 6021 to 602. n .

[0238] The method step S20 of the computer-implemented method includes using... Figure 6 The method involves determining the control value 605A of the first control parameter. Additionally, the method step S21 of the computer-implemented method includes using... Figure 6 The method involves determining the control value 605B of the second control parameter. Method step 22 of the computer-implemented method includes generating information about the object 15 using the SEM 100 and control unit 607, wherein the control unit 607 is controlled using the determined control value 605A of the first control parameter and the determined control value 605B of the second control parameter. This has been further discussed above. This discussion also applies here. Method step S23 of the computer-implemented method includes storing the determined control value 605A of the first control parameter, the determined control value 605B of the second control parameter, and the generated information about the object 15 in a database 606 as a training dataset. The training dataset can now be used to train the processing unit 601 and / or machine learning models 6021 to 602. n Training processing unit 601 and / or machine learning models 6021 to 602 n This is known in the art. Therefore, the training processing unit 601 and / or machine learning models 6021 to 602 will not be discussed in detail. n itself.

[0239] As described above, embodiments of the present invention include determining whether information about object 15 meets desired quality standards. For example, determining whether an image of object 15 meets desired quality standards. Desired quality standards may be provided by the user or may be determined as further discussed above. This discussion also applies here.

[0240] Figure 9 The machine learning models 6021 to 602 used to train SEM 100 are illustrated schematically. n A flowchart illustrating a method for identifying images of object 15 that meet quality standards provided by the user. (This is part of a larger document.) Figure 9 Methods for training machine learning models 6021 to 602 n Subsequently, machine learning models 6021 to 602 n It can identify whether the image of object 15 meets the expected quality standards.

[0241] Figure 9 The method step S30 includes the step of a user of SEM 100 providing a first image of object 15 using SEM 100, wherein the first image meets quality standards. Examples of how to define quality standards have been further discussed above. This discussion also applies here.

[0242] Figure 9 The method step S31 includes the following steps: providing a first degraded image by degrading the first image using the processing unit 601, and providing a second degraded image by degrading the first image using the processing unit 601. In other words, the first image is degraded to obtain the first degraded image and the second degraded image.

[0243] Figure 9 The method step S32 includes a step where the user evaluates whether the first degraded image meets the quality standards. The first degraded image is labeled depending on the step of evaluating it. Figure 9 Method step S33 includes a step where the user evaluates whether the second degraded image meets the quality standards. The second degraded image is labeled depending on the step of evaluating it. Method steps S32 and S33 are performed by the user of SEM 100.

[0244] Now using machine learning models 6021 to 602 n To execute Figure 9 Method step S34. Method step S34 includes using machine learning models 6021 to 602. n The steps for simulating (a) evaluating the first degraded image and (b) evaluating the second degraded image are described. The first image, the first degraded image, and the second degraded image are used as machine learning models 6021 to 602.n The training data. This simulation step is based on the first image, the first degraded image, and the second degraded image.

[0245] Figure 9 Method step S35 is performed by the user of SEM 100. Figure 9 Method step S35 includes an optional step in which the user of SEM 100 provides a second image of object 15 using SEM 100, wherein the second image meets quality standards.

[0246] Figure 9 The method step S36 includes: (a) providing a third degraded image by degrading the first image and / or the second image using the processing unit 601 of the SEM 100; and (b) providing a fourth degraded image by degrading the first image and / or the second image using the processing unit 601 of the SEM 100.

[0247] Now using machine learning models 6021 to 602 n To execute Figure 9 The method steps are as follows: S37. Figure 9 Method step S37 includes using machine learning models 6021 to 602. n The steps for evaluating whether the third degraded image meets the quality standards are as follows. The third degraded image is marked according to the steps for evaluating the third degraded image. The steps for evaluating the third degraded image and marking the third degraded image are repeated until the marking is verified to be correct using the control unit 607 or the processing unit 601.

[0248] Now using machine learning models 6021 to 602 n To execute Figure 9 The method steps are as follows: S38. Figure 9 Method step S38 includes using machine learning models 6021 to 602 n The steps for evaluating whether the fourth degraded image meets the quality standards are as follows. The fourth degraded image is marked according to the steps for evaluating it. The steps for evaluating and marking the fourth degraded image are repeated until the marking is verified to be correct using the control unit 607 or the processing unit 601.

[0249] Figure 9 The method step S39 includes the step of providing a third image of the object to the user and / or the processing unit 601 of the SEM 100. Figure 9 Method step S40 is performed by machine learning models 6021 to 602. n Execution. Machine learning models 6021 to 602 nNow evaluate and label the third image without knowing in advance whether the third image meets the quality standards. Figure 9 Method step S40 includes using machine learning models 6021 to 602. n The steps for evaluating whether the third image meets the quality standards. Depending on the steps for evaluating the third image, machine learning models 6021 to 602... n Mark the third image. Repeat the steps of evaluating the third image and marking the third image until the marking is verified to be completed correctly using the control unit 607 or the processing unit 601.

[0250] In already used Figure 9 Methods for training machine learning models 6021 to 602 n Subsequently, machine learning models 6021 to 602 n It can identify whether the image of object 15 meets the expected quality standards.

[0251] All the above and following embodiments of the method according to the invention are not limited to the order of the explained method steps. The invention also includes different orders of method steps suitable for achieving the objectives of the invention. Additionally or alternatively, at least two method steps are provided to be performed in parallel in the method according to the invention. Furthermore, the above and following embodiments of the method according to the invention are not limited to all of the method steps described above or below. In particular, it is intended to omit individual or several method steps from the above or below method steps in other embodiments.

[0252] The features of the invention disclosed in this specification, drawings, and claims may be essential for implementing the invention in the various embodiments, either individually or in any combination. The invention is not limited to the described embodiments. Modifications can be made within the scope of the claims, taking into account the knowledge of those skilled in the art.

[0253] Figure Labels

[0254] 1 beam generator

[0255] 2 Lead-out electrodes

[0256] 3 control electrodes

[0257] 4 anodes

[0258] 5 First focusing lens

[0259] 6 Second focusing lens

[0260] 7-Aperture Unit

[0261] 8 Second detector unit

[0262] 9 First guiding device / first deflection device

[0263] 10 objectives

[0264] 11-loop coil

[0265] 12 Second guiding device / second deflection device

[0266] 13 Object Room

[0267] 14 First Detector Unit

[0268] 15 objects

[0269] 16 Object Planes

[0270] 17-Pressure Aperture Support

[0271] 18 apertures

[0272] 19 Movable object platform

[0273] 20 optical axes

[0274] 21 bundles of guide tubes

[0275] 22 Extreme Boots

[0276] 23 Pole shoe gap

[0277] 25 mobile devices

[0278] 100SEM

[0279] 101 Electronic Source

[0280] 102 lead-out electrode

[0281] 103 anode

[0282] 104 bundles of guide tubes

[0283] 105 First Converging Lens

[0284] 106 Second Converging Lens

[0285] 107 First Objective

[0286] 108 First Aperture Unit

[0287] 108A First Aperture

[0288] 109 Second Aperture Unit

[0289] 110 Extreme Boots

[0290] 111 coil

[0291] 112 single electrode

[0292] 113 tubular electrode

[0293] 114 Object Holder

[0294] 115 Second guiding device / second deflection device

[0295] 116 First Detector

[0296] 116A reverse field grid

[0297] 117 Second Detector

[0298] 118 Second Aperture

[0299] Detector in Room 119

[0300] 120 Object Room

[0301] 121 Third Detector

[0302] 122 object stage

[0303] 123 control device

[0304] 124 Monitor

[0305] 125 objects

[0306] 126 Database

[0307] 127 processors / processing units

[0308] 130 First guiding device / first deflection device

[0309] 200 combined unit

[0310] 201 Object Room

[0311] 300 Ion Beam Device

[0312] 301 Ion Beam Generator

[0313] Extraction electrodes in the 302 ion beam device

[0314] 303 Converging Lens

[0315] 304 Second Objective

[0316] 306 Adjustable or selectable aperture unit

[0317] 307 First Guiding Device / First Deflection Device

[0318] 308 Second Guiding Device / Second Deflection Device

[0319] 500 radiation detector

[0320] 600 operating system

[0321] 601 Processing Unit

[0322] More than 602 machine learning models

[0323] 6021 to 602 n Machine learning models

[0324] 603 Object Data

[0325] 604A First Range Data

[0326] 604B Second Range Data

[0327] Control value of the first control parameter of 605A

[0328] Control value of the second control parameter of 605B

[0329] 606 Database

[0330] 607 control unit

[0331] 709 First Axis

[0332] 710 Second Beam Axis

[0333] 800 First Signal Connection

[0334] 801 Second Signal Connection

[0335] 1000-beam device

[0336] Distance A

[0337] A1 Distance

[0338] A2 Distance

[0339] OA optical axis

[0340] S1 to S17 Method Steps

[0341] S20 to S23 Method Steps

[0342] S30 to S40 method steps.

Claims

1. A method for operating a beam device (100, 200, 1000) to obtain information about an object (15, 125), the method comprising: - obtaining, using a processing unit (127, 601), first data from a database (126, 606) connected to the processing unit (127, 601), wherein the first data comprise object data (603) about the object (15, 125), data about a plurality (602) of machine learning models (6021 to 602 n N), and first range data (604A) about a possible value range of at least one first control parameter for controlling a control unit (123, 607) of the beam device (100, 200, 1000); - Using the processing unit (127, 601), the plurality of (602) machine learning models (6021 to 602) are determined based on the first data. n The first machine learning model (6021 to 602) in ) n ); - The object data (603) and the first range data (604A) are provided as input data to the first machine learning model (6021 to 602). n ); -Use this first machine learning model (6021 to 602) n The control value (605A) of the first control parameter is determined by the first machine learning model (6021 to 602). n The output of ) - Using the processing unit (127, 601), second data is obtained from the database (126, 606), wherein the second data includes object data (603) about the object (15, 125) and machine learning models (6021 to 602) about the plurality of (602) models. n The data and the second range data (604B) regarding the possible value range of at least one second control parameter of the control unit (123, 607) for controlling the beam device (100, 200, 1000). - Using the processing unit (127, 601), the plurality of (602) machine learning models (6021 to 602) are determined based on the second data. n The second machine learning model (6021 to 602) in ) n ); - The object data (603) and the second range data (604B) are provided as input data to the second machine learning model (6021 to 602). n ); -Use this second machine learning model (6021 to 602) n The control value (605B) of the second control parameter is determined by the second machine learning model (6021 to 602). n The output of ) - The control unit (123, 607) is operated using the determined control value of the first control parameter (605A) and the determined control value of the second control parameter (605B), and - The information about the object (15, 125) is generated using the beam device (100, 200, 1000) and the control unit (123, 607) of the beam device (100, 200, 1000).

2. The method according to claim 1, wherein, The processing unit (127, 601) is used to determine the plurality of (602) machine learning models (6021 to 602). n The second machine learning model (6021 to 602) in ) n Additionally, the control value is based on the determined first control parameter.

3. The method according to claim 1 or 2, wherein, The method includes one of the following: -The first machine learning model (6021 to 602) n ) and the second machine learning model (6021 to 602) n )same; -The first machine learning model (6021 to 602) n Unlike the second machine learning model (6021 to 602) n ).

4. The method according to any one of the preceding claims, wherein, The method includes at least one of the following: - Generating this information about the object (15, 125) includes generating an image of the object (15, 125); - Generating this information about the object (15, 125) includes generating a spectral analysis of the object (15, 125).

5. The method according to any one of the preceding claims, wherein, The first machine learning model (6021 to 602) was determined. n ) and / or the second machine learning model (6021 to 602) n This includes identifying the following machine learning models (6021 to 602). n One of the following: heuristic automatic tuning algorithm, supervised learning algorithm, and reinforcement algorithm.

6. The method according to any one of the preceding claims, comprising at least one of the following: (i) Use the first physical variable as the first control parameter; (ii) Using a first control current or a first control voltage as the first control parameter; (iii) Using a first ratio of the physical variable as the first control parameter; and (iv) Use the first amplification of the physical variable as the first control parameter.

7. The method according to any one of the preceding claims, comprising at least one of the following: (i) Use the second physical variable as the second control parameter; (ii) Use a second control current or a second control voltage as the second control parameter; (iii) Use the second ratio of the physical variable as the second control parameter; and (iv) Use the second amplification of the physical variable as the second control parameter.

8. The method according to any one of the preceding claims, comprising at least one of the following: (i) One of the following is used as the first control parameter: a first contrast parameter, which sets the contrast of the information about the object (15, 125); a first brightness parameter, which sets the brightness of the information about the object (15, 125). The first actuation parameter is used to actuate the objective lenses (10, 107, 304) of the beam assembly (100, 200, 1000); the first setting parameter is used to set the electrostatic and / or magnetic units (2, 3, 4, 5, 6, 7, 9, 10, 11, 12, 21, 22, 23, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 115, 116A, 130, 302, 303, 304, 306, 307, 308) of the beam assembly (100, 200, 1000); the first controllability ... The parameters are used to control and set the electrostatic deflection unit and / or magnetic deflection unit (2, 3, 4, 5, 6, 7, 9, 10, 11, 12, 21, 22, 23, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 115, 116A, 130, 302, 303, 304, 306, 307, 308) of the beam device (100, 200, 1000) to achieve beam shifting of the beam device (100, 200, 1000); and the first astigmatism correction parameter, which sets the astigmatism correction of the beam device (100, 200, 1000); (ii) Use one of the following as the second control parameter: a second contrast parameter that sets the contrast of the information about the object (15, 125); A second brightness parameter sets the brightness of this information about the object (15, 125); The second actuation parameter is used to actuate the objective lenses (10, 107, 304) of the beam apparatus (100, 200, 1000); the second setting parameter is used to set the electrostatic and / or magnetic units (2, 3, 4, 5, 6, 7, 9, 10, 11, 12, 21, 22, 23, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 115, 116A, 130, 302, 303, 304, 306, 307, 308) of the beam apparatus (100, 200, 1000); the second controllability ... The parameters are used to control and set the electrostatic deflection unit and / or magnetic deflection unit (2, 3, 4, 5, 6, 7, 9, 10, 11, 12, 21, 22, 23, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 115, 116A, 130, 302, 303, 304, 306, 307, 308) of the beam device (100, 200, 1000) to achieve beam shifting of the beam device (100, 200, 1000); and the second astigmatism correction parameter, which sets the astigmatism correction of the beam device (100, 200, 1000).

9. The method according to any one of claims 1 to 7, comprising at least one of the following: (i) The range between the minimum and maximum values ​​of the first control parameter is used as the possible value range of the first control parameter, wherein, The minimum and maximum values ​​of this first control parameter depend on the configuration of the control unit (123, 607); (ii) The range between the minimum and maximum values ​​of the first contrast parameter is used as the possible value range of the first control parameter, which is configured as the first contrast parameter, which sets the contrast of the information about the object (15, 125), wherein the minimum and maximum values ​​of the first contrast parameter depend on the configuration of the control unit (123, 607). (iii) The range between the minimum and maximum values ​​of the first brightness parameter is used as the possible value range of the first control parameter, which is configured as the first brightness parameter, which sets the brightness of the information about the object (15, 125), wherein the minimum and maximum values ​​of the first brightness parameter depend on the configuration of the control unit (123, 607). (iv) Using the range between the minimum and maximum values ​​of the first actuation parameter as the possible value range of the first control parameter, the first control parameter being configured as the first actuation parameter, the first actuation parameter being used to actuate the objectives (10, 107, 304) of the beam apparatus (100, 200, 1000), wherein the minimum and maximum values ​​of the first actuation parameter depend on the magnification of the beam apparatus (100, 200, 1000); (v) The range between the minimum and maximum values ​​of the second control parameter is used as the possible range of values ​​for the second control parameter, wherein the minimum and maximum values ​​of the second control parameter depend on the configuration of the control unit (123, 607); (vi) The range between the minimum and maximum values ​​of the second contrast parameter is used as the possible value range of the second control parameter, which is configured as the second contrast parameter, which sets the contrast of the information about the object (15, 125), wherein the minimum and maximum values ​​of the second contrast parameter depend on the configuration of the control unit (123, 607). (vii) The range between the minimum and maximum values ​​of the second brightness parameter is used as the possible value range of the second control parameter, which is configured as the second brightness parameter, which sets the brightness of the information about the object (15, 125), wherein the minimum and maximum values ​​of the second brightness parameter depend on the configuration of the control unit (123, 607). (viii) The range between the minimum and maximum values ​​of the second actuation parameter is used as the possible value range of the second control parameter, which is configured as the second actuation parameter for actuating the objectives (10, 107, 304) of the beam apparatus (100, 200, 1000), wherein the minimum and maximum values ​​of the second actuation parameter depend on the magnification of the beam apparatus (100, 200, 1000).

10. The method according to any one of the preceding claims, wherein, The generated information about the object (15, 125) is the first information about the object (15, 125), and the method further includes: - Using the processing unit (127, 601), it is determined whether the first information about the object (15, 125) meets the expected quality standards; and - If the expected quality standard is not met, perform the following steps: (a) Using the processing unit (127, 601), third data is obtained from the database (126, 606) connected to the processing unit (127, 601), wherein the third data includes object data (603) about the object (15, 125), and machine learning models (6021 to 602) about the plurality of (602) models. n The data and the third range data regarding the possible value range of at least one third control parameter of the control unit (123, 607) used to control the beam device (100, 200, 1000); (b) Using the processing unit (127, 601), the plurality of (602) machine learning models (6021 to 602) are determined based on the third data. n The third machine learning model (6021 to 602) in n ); (c) The object data (603) and the third range data are provided as input data to the third machine learning model (6021 to 602). n ); (d) Using this third machine learning model (6021 to 602) n The control value of the third control parameter is determined by the third machine learning model (6021 to 602). n The output of ) (e) Operate the control unit (123, 607) using the determined control value of the third control parameter; and (f) Using the beam device (100, 200, 1000) and the control unit (123, 607) of the beam device (100, 200, 1000) to generate second information about the object (15, 125).

11. The method according to claim 10, wherein, The third machine learning model (6021 to 602) was determined. n Additionally, based on the determined control value of the first control parameter (605A) and / or the determined control value of the second control parameter (605B).

12. The method according to claim 10 or 11, wherein, The method includes at least one of the following: - Generating the second information about the object (15, 125) includes generating an image of the object (15, 125); - The generation of this second information about the object (15, 125) includes generating a spectral analysis of the object (15, 125).

13. The method according to any one of the preceding claims, wherein, The method includes at least one of the following: (i) Using an optical microscope as the beam apparatus (100, 200, 1000). (ii) Use a laser beam device as the beam device (100, 200, 1000). (iii) Use a particle beam device as the beam device (100, 200, 1000). (iv) Use an electron beam device as the beam device (100, 200, 1000). (v) Use an ion beam device as the beam device (100, 200, 1000).

14. A computer program product having program code that can be loaded into a processor (127, 601) and, when executed, controls a bundle device (100, 200, 1000) to perform the method according to at least one of the preceding claims.

15. A beam device (100, 200, 1000) for generating information about objects (15, 125), the beam device comprising: - At least one beam generator (1, 101, 301) for generating a beam. - At least one objective lens (10, 107, 304) for focusing the beam onto the object (15, 125). - At least one detector unit (8, 14, 106, 107, 119, 121, 500) for detecting interacting particles and / or interacting radiation resulting from the interaction of the beam with the object (15, 125), and - At least one processor (127,601) wherein the at least one processor is loaded with the computer program product according to claim 14.

16. The beam device (100, 200, 1000) according to claim 15, wherein, The beam apparatus (100, 200, 1000) is at least one of the following: an optical microscope for imaging and / or analyzing the object (15, 125); and a laser beam apparatus for imaging, processing and / or analyzing the object (15, 125).

17. The beam device (100, 200, 1000) according to claim 15 or 16, wherein, The beam apparatus (100, 200, 1000) is a particle beam apparatus for imaging, processing and / or analyzing the object (15, 125).

18. The beam device (100, 200, 1000) according to claim 17, wherein, - The beam generator (1, 101, 301) is a particle beam generator (1, 101, 301) used to generate a particle beam with charged particles. - The objective (10, 107, 304) is used to focus the particle beam onto the object (15, 125). - The detector unit (8, 14, 106, 107, 119, 121, 500) is a detector unit (8, 14, 106, 107, 119, 121, 500) used to detect interacting particles and / or interacting radiation generated due to the interaction between the particle beam and the object (15, 125), and The beam assembly (100, 200, 1000) further includes: - At least one scanning device (9, 12, 115, 130) for scanning the particle beam over the object (15, 125).

19. The beam device (200) according to claim 18, wherein, The particle beam generator (101) is implemented as a first particle beam generator, and the particle beam is a first particle beam having first charged particles, wherein the objective lens (107) is a first objective lens for focusing the first particle beam onto the object (15, 125), and wherein the particle beam device (200) further includes: - At least one second particle beam generator (301), the at least one second particle beam generator being used to generate a second particle beam carrying second charged particles; and - At least one second objective lens (304) for focusing the second particle beam onto the object (15, 125).

20. The beam device (100, 200, 1000) according to any one of claims 15 to 19, wherein, The beam apparatus (100, 200, 1000) is an electron beam apparatus and / or an ion beam apparatus.

21. A processing unit (127, 601) for (i) a beam device (100, 200, 1000) and / or (ii) a machine learning model (6021 to 602) n A computer-implemented method for generating a training dataset, the machine learning model being used to determine control values ​​(605A) of a first control parameter to operate a control unit (123, 607) of a beam assembly (100, 200, 1000) and to determine control values ​​(605B) of a second control parameter to operate the control unit (123, 607) of the beam assembly (100, 200, 1000), the method comprising: - The control value of the first control parameter is determined using the method according to any one of claims 1 to 13 (605A); - The control value of the second control parameter is determined using the method according to any one of claims 1 to 13 (605B). - Use the beam device (100, 200, 1000) and the control unit (123, 607) of the beam device (100, 200, 1000) to generate information about the object (15, 125); as well as - The determined control value of the first control parameter (605A), the determined control value of the second control parameter (605B), and the generated information about the object (15, 125) are stored in the database (126, 606) as the training dataset.

22. A machine learning model (6021 to 602) for training beam devices (100, 200, 1000) n A method for identifying images of objects (15, 125) that meet quality criteria given by the user, wherein, This method includes the following steps: - The user of the beam device (100, 200, 1000) uses the beam device (100, 200, 1000) to provide a first image of the object (15, 125), wherein the first image meets the quality standard; -At least one first degraded image is provided by degrading the first image using the processing unit (127, 601) associated with the beam device (100, 200, 1000), and at least one second degraded image is provided by degrading the first image using the processing unit (127, 601) associated with the beam device (100, 200, 1000). - The user evaluates whether the at least one first degraded image meets the quality standard, and marks the at least one first degraded image according to the steps of evaluating the at least one first degraded image; - The user evaluates whether the at least one second degraded image meets the quality standard, and marks the at least one second degraded image according to the steps of evaluating the at least one second degraded image; -Use this machine learning model (6021 to 602) n The steps of evaluating at least one first degraded image and evaluating at least one second degraded image are simulated, wherein the first image, the at least one first degraded image, and the at least one second degraded image are used as inputs for the machine learning model (6021 to 602). n The training data is based on the first image, the at least one first degraded image, and the at least one second degraded image; - The user of the beam device (100, 200, 1000) uses the beam device (100, 200, 1000) to provide a second image of the object (15, 125), wherein the second image meets the quality standard; - Provide at least one third degraded image by degrading the first image and / or the second image using the processing unit associated with the beam device (100, 200, 1000), and provide at least one fourth degraded image by degrading the first image and / or the second image using the processing unit (127, 601) associated with the beam device (100, 200, 1000); -Use this machine learning model (6021 to 602) n The at least one third degraded image is evaluated to determine whether it meets the quality standard, and the at least one third degraded image is marked according to the step of evaluating the at least one third degraded image, wherein the step of evaluating the at least one third degraded image and the step of marking the at least one third degraded image are repeated until the processing unit (127, 601) associated with the beam device (100, 200, 1000) verifies that the marking has been correctly completed; -Use this machine learning model (6021 to 602) n The at least one fourth degraded image is evaluated to determine whether it meets the quality standard, and the at least one fourth degraded image is marked according to the step of evaluating the at least one fourth degraded image, wherein the step of evaluating the at least one fourth degraded image and the step of marking the at least one fourth degraded image are repeated until the processing unit (127, 601) associated with the beam device (100, 200, 1000) verifies that the marking has been correctly completed; - The user and / or the processing unit (127, 601) associated with the beam device (100, 200, 1000) provides at least one third image of the object (15, 125); and -Use this machine learning model (6021 to 602) n The at least one third image is evaluated to determine whether it meets the quality standard, and the at least one third image is marked according to the steps of evaluating the at least one third image, wherein the steps of evaluating the at least one third image and marking the at least one third image are repeated until the processing unit (127, 601) associated with the beam device (100, 200, 1000) verifies that the marking is correctly completed.

Images