Dual velocity acquisition for drift-corrected, high-speed, low-dose, adaptive composition charged particle imaging.

Abstract

To provide methods for drift corrected, fast, low dose, adaptive sample imaging with a charged particle microscopy system.SOLUTION: Methods for adaptive sample imaging with a charged particle microscopy system 104 include scanning a surface region of a sample 102 with a charged particle beam 110 to obtain a first image of the surface region with a first detector modality, and then determining a scan strategy for the surface region. The scan strategy comprises a charged particle beam path, a first beam dwell time associated with at least one region of interest in the first image, and at least a second beam dwell time associated with other regions of the first image, the first beam dwell time being different than the second beam dwell time. The surface region of the sample is then scanned with the determined scan strategy to obtain data from first and second detectors.SELECTED DRAWING: Figure 1

Description

Background Art

[0001] Charged particle microscope systems have been developed to enable scientists to investigate and gather information about the workings of the microscope system. To pursue such knowledge, scientists challenge the limits of what current charged particle microscope systems can investigate. This is particularly true for the evaluation of atoms and nanoparticles. In this case, due to the size of the objects being investigated, many new obstacles to evaluation arise. For example, because atoms and nanoparticles are very small, even using the highest magnification optical systems, most of the area scanned by the charged particle beam during the investigation of such atoms / nanoparticles will necessarily correspond to voids and / or structures other than the atoms / nanoparticles being investigated. This means that current charged particle microscope systems waste time acquiring irrelevant data when evaluating atoms / nanoparticles. Furthermore, as a property of atoms, there is a large void between the atomic nucleus and other atomic structures with respect to the size of the atom, so this problem of scanning the space between atoms cannot be solved simply by increasing the magnification.

[0002] Furthermore, since the size of the structures irradiated during the evaluation of atoms / nanoparticles is reduced, samples containing atoms / nanoparticles are more likely to be damaged by such irradiation. To compensate for such damage, current charged particle microscope systems reduce the beam intensity and the length of time the beam irradiates individual structures. This reduction in beam intensity and dwell time reduces damage to the sample but also reduces the amount of information acquired during each scan. Therefore, in order to obtain statistically meaningful data, current charged particle microscope systems are forced to perform many repeated scans (10 to 1000 times) of the sample before being able to cumulatively acquire statistically significant data on the atoms / nanoparticles being investigated. This means that the current evaluation of such atom / nanoparticle samples can take several hours or days.

[0003] This iteration further complicates the evaluation of atoms and nanoparticles because individual shifts between scanning instances lead to data drift / inaccuracy, which is accumulated over many scans. Therefore, a better method for evaluating atoms and nanoparticles is desired to address the time, damage, and accuracy issues of the current system. [Overview of the project]

[0004] A drift-corrected, fast, low-dose, adaptive sample imaging method using a charged particle microscope system according to this disclosure comprises scanning a surface region of a sample with a charged particle beam to acquire a first image of the surface region by a first detector modality, and then determining a scanning strategy for the surface region. The scanning strategy includes a charged particle beam path and a first beam residence time associated with at least one region of interest in the first image, wherein the first beam residence time is sufficient to acquire statistically significant data from a second detector modality, and at least a second beam residence time associated with other regions of the first image, wherein the first beam residence time is different from the second beam residence time. The method further comprises scanning a surface region of a sample according to the determined scanning strategy to acquire first data from a second detector modality and second data from the first modality.

[0005] A system for investigating a sample using a charged particle microscope system according to the present disclosure comprises a sample holder configured to hold the sample; a charged particle source configured to emit a beam of charged particles toward the sample; an optical column configured to irradiate the beam of charged particles toward the sample; and one or more detectors configured to detect the charged particles of the charged particle beam and / or emissions resulting from the charged particle beam irradiated toward the sample. According to the present disclosure, the one or more detectors include at least a first detector of a first modality and a second detector of a second modality different from the first modality. The system also includes one or more processors and a memory for storing computer-readable instructions that, when executed by the one or more processors, cause the corresponding system to perform one or more steps of the method according to the present disclosure. [Brief explanation of the drawing]

[0006] Detailed explanations are provided in the attached diagram. In the diagram, the leftmost digit of the reference number identifies the diagram in which the reference number first appears. The same reference number in different diagrams indicates similar or identical items.

[0007] [Figure 1] This disclosure provides an exemplary environment for drift-corrected, high-speed, low-dose, adaptive sample imaging using a charged particle microscope system. [Figure 2] This disclosure describes a sample process for drift-corrected, high-speed, low-dose, adaptive sample imaging using a charged particle microscope system. [Figure 3] This is a series of diagrams illustrating the process for determining a customized scanning strategy based on the initial sample image. [Figure 4] This is a series of diagrams illustrating the process for determining a customized scanning strategy based on the initial sample image. [Figure 5] This is a series of diagrams illustrating the process of identifying structures of interest within the initial image. [Figure 6]This is a series of diagrams illustrating the process of identifying different species of structures of interest within the initial image. [Figure 7] This is a series of diagrams illustrating the process for determining the composition of structures of interest identified within initial sample images. [Figure 8] These are a series of figures illustrating an exemplary process for determining the composition of a structure of interest in a sample, as disclosed herein.

[0008] Similar reference numerals refer to corresponding parts throughout several drawings. Generally, in the drawings, elements that are likely to be included in a given example are shown with solid lines, while elements that are optional in a given example are shown with dashed lines. However, elements shown with solid lines are not essential in all examples of this disclosure, and elements shown with solid lines may be omitted from certain examples without departing from the scope of this disclosure. [Modes for carrying out the invention]

[0009] Methods and systems for drift-corrected, fast, low-dose, adaptive sample imaging using a charged particle microscope system are included herein. More specifically, the methods and systems disclosed herein include, and / or are configured to, acquire an initial image of a region of a sample using data from a detector of a first modality, identify a structure of interest in the first image, and determine a scanning strategy including at least two residence times. For example, the scanning strategy may include a first residence time for illuminating a region of the sample associated with a structure of interest, and a second residence time for a region of the sample that does not contain a structure of interest.

[0010] In various embodiments of this disclosure, residence times can be selected based on the type of object of interest, the species of object of interest, the detector modality for the desired information, or a combination thereof. Thus, a first residence time can be selected to allow a statistically significant amount of data from a second modality, and a second residence time can be selected to minimize the time and / or damage induced during scanning of portions of the sample that do not contain the object of interest. These sample types require longer residence times to acquire compositional information (from the relevant detector modality), but this is particularly useful for evaluating atomic samples and / or nanoparticles, as the majority of the scanned area corresponds to the space between atoms / nanoparticles or other structures of no interest. Furthermore, in some embodiments, there may be multiple residence times in the scanning strategy, each corresponding individually to different types of objects of interest, different species of objects of interest, different detector modalities required for the desired information, or combinations thereof. Furthermore, in some embodiments, the beam path can also be determined so that the object of interest is scanned while minimizing the amount of other areas of the sample being scanned.

[0011] In some embodiments, the scanning strategy can be repeated until a sufficient amount of data from the second modality is cumulatively obtained so that the object of interest is compositionally identified, imaged, or otherwise evaluated. Because the scanning strategy allocates a low residence time to areas of the sample that do not contain the object of interest, the time cost and sample damage that would occur when such iterations are performed by conventional methods are significantly reduced (e.g., by more than 600 times). This reduction in scanning time and sample damage allows scientists and researchers to obtain evaluation results that were previously impossible using the systems and methods of this disclosure.

[0012] The methods and systems of this disclosure can also use data portions obtained from scanning a region of the sample that does not contain a structure of interest by a second residence time to compensate for drift between images. For example, if the first modality is TEM dark-field imaging (such as HAADF), dark-field imaging data acquired while the sample is irradiated by a second residence time can be used to positionally align this acquired data (and therefore data acquired by other residence times) with other instances of data acquired during other scans using the scanning strategy. The systems and methods can then adjust the scanning strategy to compensate for determined and / or expected drift, or adjust the detected data results in post-processing.

[0013] Figure 1 illustrates an exemplary environment 100 for drift-corrected, fast, low-dose, adaptive sample imaging using a charged particle microscope system according to the present disclosure. Specifically, Figure 1 shows the exemplary environment 100 including an exemplary charged particle system 104 for the investigation and / or analysis of a sample 102. The exemplary charged particle system 104 may be, or include, one or more different types of optics and / or charged particle microscopes, such as a scanning electron microscope (SEM), scanning transmission electron microscope (STEM), transmission electron microscope (TEM), charged particle microscope (CPM), cryogenic adaptive microscope, focused ion beam (FIB) microscope, dual-beam microscope system, or a combination thereof. Figure 1 shows the exemplary charged particle microscope system 104 as a transmission electron microscope (TEM) 106.

[0014] An exemplary charged particle microscope system 104 includes a charged particle source 108 (e.g., a thermionic electron source, a Schottky emission source, a field emission source, etc.) that emits a charged particle beam 110 along an emission axis 112 toward an accelerating lens 114. The emission axis 112 is a central axis extending from the charged particle source 108 along the length of the exemplary charged particle microscope system 104 and through the sample 102. The accelerating lens 114 accelerates / decelerates, focuses, and / or directs the electron beam 110 toward a focusing column 116. The focusing column 116 focuses the electron beam 110 so that it is incident on at least a portion of the sample 102. In some embodiments, the focusing column 116 may include one or more of an aperture, a scanning coil, and an upper focusing lens. The focusing column focuses electrons from the electron source to a small spot on the sample. Different locations on the sample 102 can be scanned by adjusting the direction of the electron beam via the scanning coil. Furthermore, the focusing column 116 can correct and / or adjust aberrations of the electron beam 110 (e.g., geometric aberration, chromatic aberration). Figure 1 further shows an exemplary charged particle microscope system 104, further including a sample holder 118 configured to hold a sample 102, which can translate, rotate, and / or tilt the sample 102 relative to the exemplary charged particle microscope system 104.

[0015] Figure 1 also shows an exemplary charged particle microscope system 104 as comprising multiple detector systems 120 of at least two different modalities. Possible modalities of the exemplary charged particle microscope system 104 include dark-field imaging, bright-field imaging, diffraction pattern imaging, electron energy loss spectroscopy (EELS), energy-dispersive X-ray spectroscopy (EDS, EDX, or XEDS), and cathode ray emission. oh and backscattered electrons Detection method using one of the following This may include, for example, a charged particle microscope system may include a HAADF detector system as a first detector modality and an EDS detector system as a second detector modality. Multiple detector systems 120 are further shown as being connected to one or more computing devices 122.

[0016] The computing device 122 is configured to control the operation of an exemplary charged particle microscope system 104, generate an image of the sample 102, and / or otherwise determine or interpret data from multiple detector systems 120. According to the present invention, the computing device 122 is configured to cause the charged particle microscope system 104 to scan a surface region of the sample 102 with a charged particle beam 110 (e.g., an electron beam), acquire data from a detector system 122 (e.g., a dark-field imaging detector system), and then generate an initial image of the surface region based on the detector data. The computing device 122 is further configured to identify structures of interest within the initial image.

[0017] Once a structure of interest is identified, the computing device 122 uses the location of the structure of interest in the image to determine a scanning strategy for the surface region of the sample 102. The scanning strategy includes at least a beam path that shows the path of the charged particle beam 110 as it irradiates the surface of the sample 102, where each portion of the beam path has an associated residence time (e.g., the amount of time the charged particle beam 110 irradiates the relevant portion of the sample during the execution of the scanning strategy, the pixel density scanned by the charged particle beam 110, the size of the pixels scanned by the charged particle beam 110, the pulse length of the charged particle beam 110, the pulse frequency of the charged particle beam 110, etc.). Specifically, the scanning strategy includes one or more beam path portions associated with at least one structure of interest to which a first beam residence time is assigned, and at least one additional residence time assigned to the remainder of the beam path. According to the present invention, a first beam residence time is determined by a computing device 122 such that it is sufficient to obtain statistically significant data from a second detector modality different from the first detector modality used to acquire an initial image. Statistically significant data includes a quantity or quality of data that describes the material properties and / or interrelationships of the corresponding structure of interest, which are not statistically random results. Specifically, the first residence time is determined such that the data obtained within a threshold number of scans using a scanning strategy is sufficient to enable compositional determination of the corresponding structure of interest. For example, the first residence time may be determined such that, when the sample is scanned a threshold number of times using a scanning strategy, the first residence time provides sufficient data to enable determination of the species of atoms scanned. In different embodiments, a sufficient amount of data to make such compositional determination is acquired in a single scan or within a threshold number of scans. The computing device 122 is further configured to cause the charged particle microscope system 104 to scan a surface region of the sample according to the determined scanning strategy, to acquire first data from the second detector modality, and to acquire second data from the first modality. Next, the first data is used by the computing device 122 to identify and / or otherwise determine compositional information relating to individual structures of interest.In some embodiments, if the computing device 122 causes the charged particle microscope system 104 to scan a surface area two or more times, the second data can be used to determine and / or correct the drift between images generated from individual iterations of the second data.

[0018] Those skilled in the art will understand that the computing device 122 depicted in Figure 1 is merely illustrative and is not intended to limit the scope of this disclosure. Computing systems and devices may include any combination of hardware or software capable of performing specified functions, including computers, network devices, internet appliances, PDAs, wireless telephones, controllers, oscilloscopes, amplifiers, etc. The computing device 122 may also be connected to other devices not shown, or it may instead operate as a standalone system. In addition, the functions provided by the illustrated components may, in some embodiments, be combined into fewer components or distributed to additional components. Similarly, in some embodiments, some functions of the illustrated components may not be provided, and / or other additional functions may be available.

[0019] The computing device 122 may be a component of the exemplary charged particle microscope system 104, a separate device from the exemplary charged particle microscope system 104, or a combination thereof, communicating with the exemplary charged particle microscope system 104 via a network communication interface. For example, the exemplary charged particle microscope system 104 may include a first computing device 122 which is a component part of the exemplary charged particle microscope system 104 and acts as a controller that drives the operation of the exemplary charged particle microscope system 104 (e.g., adjusting the scanning location on the sample 102 by operating the scanning coil 102, translating the sample 102, etc.). In such an embodiment, the exemplary charged particle microscope system 104 may also include a second computing device 120 which is a desktop computer separate from the exemplary charged particle microscope system 104 and capable of processing data received from one or more detector systems 120 to generate an image of the sample 102, determine a scanning strategy for the sample 102, and / or perform other types of analysis. The computing device 122 may also be configured to receive user selections via a keyboard, mouse, touchpad, touchscreen, or the like.

[0020] FIG. 1 also depicts a visual flowchart 130 that includes a plurality of images depicting, together, an exemplary process that may be performed by a computing device 122 as part of drift-corrected high-speed, low-dose, adaptive imaging of a sample 102 using a charged particle microscope system 104, in accordance with the present disclosure. For example, image 132 shows an image of a surface region of sample 102 generated by computing device 122 based on sensor data from a detector system 120 of a first modality. For example, FIG. 1 shows a lower-resolution image of the surface of sample 102 obtained from HAADF detector data. Image 134 shows an updated version of image 132 in which structures of interest identified by computing system 122 are visually indicated. Specifically, image 134 shows a first plurality of regions 140 of the image identified as corresponding to atoms of a first species and a second plurality of regions 142 identified as corresponding to atoms of a second species. In other embodiments, the structures of interest may correspond to compounds, nanoparticles, and the like. In various embodiments, the plurality of regions may be identified using one or more of an image sharpening algorithm, an image recognition algorithm, a machine learning algorithm, and user selection. Image 134 also includes representative pixels or sets of pixels 144 for the corresponding identified regions. Such representative pixels 144 may correspond to a central pixel, a pixel farthest from an edge, and / or a pixel determined to be likely to return compositional information related to the corresponding structure of interest when irradiated in other manners.

[0021] Image 136 is a mask that identifies positions on the surface of sample 102 that are irradiated to determine compositional information for the corresponding structures of interest. In some embodiments, the mask may indicate dwell times associated with various species of atoms determined by computing device 122 based on the type of structure of interest, the species of the structure of interest, or both. In image 136, the positions to be irradiated correspond to representative pixels 144 determined with respect to image 134. Image 136 shows the mask as optionally differentiating the irradiation positions according to the associated atomic species.

[0022] Image 138 shows a scanning strategy 146 determined by the computing device 122 based on an image 132 of the surface of sample 102 generated by detector data of the first modality. The scanning strategy 146 shows the beam path 148 that charged particles travel through as they irradiate the surface of sample 102. Furthermore, image 138 shows each portion of the beam path 148 as associated with residence time (e.g., the time the charged particle beam 110 irradiates the relevant portion of the sample during the execution of the scanning strategy 146). Specifically, the scanning strategy 146 includes a first residence time for a first beam path portion 150 associated with the centers of first species atoms identified in image 132, and a second residence time for a second beam path portion 152 associated with the centers of second species atoms identified in image 132. In some embodiments, the computing device 122 is configured to determine different residence times for different types of objects of interest, different species of objects of interest, different detector modalities required for the desired information, or combinations thereof. Alternatively, the residence time associated with each structure of interest may be the same for all types and / or species. Scanning strategy 146 is further shown as including a third residence time for portions 154 of the beam path that do not correspond to structures of interest. The third residence time may correspond to the same residence time used to obtain image 132.

[0023] FIG. 1 further includes a schematic diagram showing an exemplary computing architecture 160 of computing device 122. The exemplary computing architecture 160 shows additional details of hardware and software components that can be used to implement the techniques described in this disclosure. One of ordinary skill in the art will understand that the computing architecture 160 may be implemented in a single computing device 122 or may span multiple computing devices. For example, the individual modules and / or data configurations shown in the computing architecture 160 may be executed and / or stored by different computing devices 122. In this way, different process steps of the method of the present invention according to this disclosure may be executed and / or implemented by separate computing devices 122.

[0024] In the exemplary computing architecture 160, the computing device includes one or more processors 162 and a memory 164 communicatively coupled to the one or more processors 162. The exemplary computing architecture 160 can include a structure identification module 166, a scanning strategy determination module 168, a composition determination module 170, an optional position alignment module 172, and a control module 174 stored in the memory 164.

[0025] Where used herein, the term “module” is intended to represent an exemplary division of executable instructions for the purposes of consideration, and not to represent any type of requirement or necessary method, scheme or organization. Thus, although various “modules” are described, their functions and / or similar functions may be arranged differently (e.g., combined into fewer modules, or divided into more modules). Furthermore, while certain functions and modules are described herein as being implemented by software and / or firmware executable on a processor, in other instances, any or all modules may be implemented in whole or in part by hardware (e.g., a specialized processing unit) to perform the described functions. As described above in various embodiments, modules described herein in relation to the exemplary computing architecture 160 can be executed across multiple computing devices 122.

[0026] The control module 174 can be made executable by the processor 162 to cause the computing device 122 and / or the exemplary charged particle microscope system 104 to perform one or more operations. For example, the control module 174 can cause the exemplary charged particle microscope system 104 to scan the surface of a sample 102 by generating a charged particle beam 110. The computing device 112 may then be configured to generate an initial image of the surface of the sample 102 based on detector data from the first modality detector system 120 obtained as the surface of the sample 102 is scanned. In an alternative example, the control module 174 may cause the exemplary charged particle microscope system 104 to scan the surface of the sample 102 according to a scanning strategy determined by the computing device 122 based on the initial image.

[0027] A structure recognition module 166 may be run by the processor 162 for structures of interest in the initial image of the surface of the sample 102 obtained by the detector system 120 of the first modality. In some embodiments, a feature determination module 164 may be run by the processor 162 to first sharpen the initial image using a sharpening algorithm. For example, the structure recognition module 166 can improve the resolution of the initial image by utilizing a neural network-based sharpening algorithm. In this way, according to the disclosure, the initial image can be acquired with reduced dwell time and / or beam intensity so as to reduce time waste and / or damage to the sample.

[0028] The structures of interest identified by the structure identification module 116 may correspond to atoms, molecules, compounds, nanoparticles, and their species or types. For example, the structure identification module 166 may identify regions of the initial image corresponding to individual nanoparticles and / or component parts of nanoparticles, such as boundaries, edges, material transitions, and component substructures. Alternatively or in addition, the structure identification module 166 may identify regions of different types or species of structures of interest within the initial image. For example, the structure identification module 166 may identify a first set of regions of the initial image corresponding to a first compound structure, and a second set of regions of the initial image corresponding to a second compound structure. In some embodiments, the structure identification module 166 is further configured to identify one or more representative pixels associated with individual instances of the identified structures of interest (e.g., the central pixel, the pixel furthest from an edge, and / or pixels otherwise determined to represent the structures of interest).

[0029] In various embodiments of this disclosure, the structure recognition module 166 can identify structures of interest using one or more of the following: a recognition algorithm, a machine learning module trained to identify instances of one or more structures of interest (e.g., an artificial neural network (ANN), a convolutional neural network (CNN), a fully convolutional neural network (FCN), etc.), user selection, and / or a combination thereof. For example, a neural network can be used to segment an initial image according to whether a corresponding pixel contains a structure of interest. This segmented image can then be presented to a user via a display 176, and the user can select individual segmented regions to be investigated. If the structure recognition module 166 segments the initial image into several different types of regions (i.e., regions that are unlikely to contain a structure of interest, regions that contain a first type / kind of structure of interest, regions that contain a second type / kind of structure of interest, etc.), the user can select individual types of regions to investigate. In an alternative example, the structure recognition module 166 could apply an algorithm programmed to select regions to be investigated based on the characteristics / features of the regions (e.g., size, shape, location, proximity to other regions, etc.) to such segmented images.

[0030] A scanning strategy determination module 168 may be executable by the processor 162 to identify a scanning strategy specifically customized for a region of the sample 102 being investigated. Specifically, the scanning strategy determination module 168 may be executable to identify the beam paths that the charged particle beam 110 will follow when scanning the sample 102 and to divide the beam paths into portions based on whether they correspond to portions of the sample determined to have structures of interest, and / or the type or species of structures of interest to which they correspond. The beam paths may correspond to pre-programmed beam paths that are constant for all samples, or they may be individually customized to suit the sample being investigated. For example, the scanning strategy determination module 168 can determine beam paths that illuminate regions of the sample determined to have structures of interest and / or associated representative pixels.

[0031] The scanning strategy determination module 168 may further execute to assign residence times to different portions of the beam path based on whether the portion of the beam path corresponds to a portion of the sample determined to have a structure of interest, and / or based on the type or species of the structure of interest they correspond to. For example, based on whether ... they,

[0032] The portion of the beam path that irradiates areas of sample 102 not associated with the structure of interest can be determined, and then reduced residence times can be assigned to those portions of the beam path. Such reduced residence times may be sufficient to acquire data from the first modality used to generate the initial image, but the reduced residence times reduce the time spent scanning the portions of the sample that do not contain the structure of interest, and at the same time reduce sample damage in general. This is particularly beneficial when the structure being investigated is atoms, nanoparticles, or structures thereof, because in such investigations, the majority of the irradiated area inevitably corresponds to empty space and / or contains structures of no interest to the user. Thus, by assigning reduced residence times to such portions of the beam path, it is possible to save more than 100 times the time compared to conventional systems.

[0033] The scanning strategy determination module 168 further assigns longer residence times to portions of the beam path corresponding to structures of interest (e.g., portions corresponding to pixels including only the centers of atoms or the edge structures of nanoparticles), where the longer residence time is sufficient to obtain statistically significant data from the detector system 120 of the second modality. For example, the scanning strategy determination module 168 may assign a specific residence time to a pixel determined to correspond to the center of an atom of interest, where the specific residence time is sufficient to allow the computing device 122 to obtain enough data from the EDS detector system to make a compositional determination related to that atom when a region of the sample is irradiated for that specific residence time. In some embodiments, the specific residence time may be a preset residence time assigned to all beam path portions corresponding to structures of interest, or it may be determined by the scanning strategy determination module 168 based on the type / sensitivity of the desired detector modality, the type of structure of interest, the species of structure of interest, or a combination thereof. For example, the scanning strategy determination module 168 can assign a first residence time to a portion of the beam path corresponding to a first type of structure of interest, and a second residence time to a portion of the beam path corresponding to a second different type of structure of interest. In this way, the scanning strategy determination module 168 can be customized to acquire statistically significant data related to multiple types of structures of interest, while simultaneously generating a scanning strategy that minimizes the time required to acquire such data and minimizes damage to the sample. In some embodiments, the scanning strategy determination module 168 can be further configured to vary the beam intensity of the charged particle beam 110 based on whether a portion of the beam path corresponds to a portion of the sample that has been determined to have a structure of interest, and / or the type or species of the structure of interest it corresponds to.

[0034] The scanning composition determination module 170 can be made executable by the processor 162 to determine compositional information about a structure of interest within the sample 102 based on detector data from a second modality detector system 120 obtained when the sample 102 is scanned using a customized scanning strategy. For example, if the second modality corresponds to an EDS detector system, the scanning composition determination module 170 can determine the species and / or material composition of atoms in or around the nanoparticles.

[0035] In some embodiments, based on the determination of the scanning composition determination module 170 that the amount of detector data from the second modality is insufficient to determine the composition of a structure of interest, the scanning composition determination module 170 may cause the charged particle microscope system 104 to perform additional scans of the sample 102 using a customized scanning strategy. In some embodiments, this may be repeated until compositional information is determined for all or a threshold number of structures of interest. Alternatively, or in addition, this may be repeated until the scanning composition determination module 170 determines that the additional scans cause sample damage exceeding a threshold, and / or as a result, one or more structures of interest are destroyed (or destroyed).

[0036] An optional positional alignment module 172 may be executable by the processor 162 to positional-align images and / or detector data obtained during repeated scans of a sample using a customized scanning strategy. For example, the positional alignment module 172 can use data portions obtained from scanning regions of the sample that do not contain structures of interest to compensate for drift between image acquisitions. For example, if the first modality is TEM dark-field imaging (such as HAADF), dark-field imaging data acquired while the sample is irradiated with a reduced residence time can be used to positional-align the resulting images, and data from other modalities obtained with increased residence times, with data acquired during other scans using a customized scanning strategy. The optional positional alignment module 172 can then adjust the scanning strategy to compensate for determined and / or expected drift, or perform post-processing on the detected data to adjust for drift. In this way, the positional alignment module 172 can ensure that the charged particle microscope system 104 acquires high-resolution images even when multiple scans are required to obtain sufficient compositional information.

[0037] As described above, the computing device 122 includes one or more processors 162 configured to execute instructions, applications, or programs stored in memory 164 accessible to one or more processors. In some examples, the one or more processors 162 may include, but are not limited to, hardware processors such as a hardware central processing unit (CPU) or a graphics processing unit (GPU). In many cases, the technology is described herein as being executed by one or more processors 162, but in some cases, the technology may be implemented by one or more hardware logic components such as a field-programmable gate array (FPGA), a composite programmable logic unit (CPLD), an application-specific integrated circuit (ASIC), a system-on-a-chip (SoC), or a combination thereof.

[0038] A memory 164 accessible to one or more processors 162 is an example of a computer-readable medium. Computer-readable media may include two types of computer-readable media: computer storage media and communication media. Computer storage media may include volatile and non-volatile, removable and fixed media implemented in any method or technique for storing information such as computer-readable instructions, data structures, program modules, or other data. Computer storage media may include, but are not limited to, random access memory (RAM), read-only memory (ROM), erasable programmable read-only (EEPROM), flash memory or other memory technologies, compact disk read-only memory (CD-ROM), digital multipurpose disk (DVD), or other optical storage, magnetic cassettes, magnetic tapes, magnetic disk storage devices or other magnetic storage devices, or any other non-transmission media that may be used to store desired information and may be accessed by a computing device. Generally, computer storage media may include computer-executable instructions that, when executed by one or more processing units, cause to perform various functions and / or operations described herein. In contrast, communication media embody computer-readable instructions, data structures, program modules, or other data in modulated data signals such as carrier waves, or other transmission mechanisms. Computer storage media, as defined herein, do not include communication media.

[0039] Those skilled in the art will also understand that items or parts thereof may be transferred between memory 164 and other storage devices for the purposes of memory management and data integrity. Alternatively, in other embodiments, some or all of the software components may run in the memory of another device and communicate with the computing device 122. Some or all of the system components or data structures may also be stored in a non-temporary computer-accessible medium or in a portable product that can be read by a suitable drive (e.g., as instructions or structured data), various examples of which are described above. In some embodiments, instructions stored in a computer-accessible medium separate from the computing device 122 may be transmitted to the computing device 122 via a transmission medium or signal such as electrical signals, electromagnetic signals, or digital signals carried over a communication medium such as a wireless link. Various embodiments may further include receiving, transmitting, or storing instructions and / or data implemented in accordance with the foregoing description of computer-accessible mediums.

[0040] Figure 2 is a depiction of a sample process 200 for drift-corrected, fast, low-dose, adaptive sample imaging using a charged particle microscope system, as described herein. Process 200 can be carried out by any of the exemplary charged particle microscope system 104, computing device 122, and computing architecture 160.

[0041] In 202, an initial image of the sample is acquired using data from a first detector modality. Detector data from the first modality can be acquired by scanning the surface of the sample with a charged particle beam, and then an initial image can be acquired based on the detector data from the first modality. In some embodiments, the first modality may correspond to dark-field or bright-field imaging (e.g., HAADF).

[0042] In 204, one or more structures of interest are identified in the initial image. In various embodiments of the Disclosure, structures of interest can be identified using one or more of the following: an identification algorithm, a machine learning module trained to identify instances of one or more structures of interest (e.g., an artificial neural network (ANN), a convolutional neural network (CNN), a fully convolutional neural network (FCN), etc.), user selection, and / or a combination thereof. The identified structures of interest may correspond to atoms, molecules, compounds, nanoparticles, and their species or types. For example, regions in the initial image can be identified that correspond to individual nanoparticles and / or component parts of nanoparticles, such as boundaries, edges, material transitions, component substructures, etc. In various embodiments, a single type / species of structure of interest may be identified in the initial image, or multiple different types / species of structures of interest may be identified in the initial image. In addition, one or more representative pixels associated with individual instances of the identified structures of interest (e.g., the central pixel, the pixel furthest from an edge, and / or pixels otherwise determined to represent the structure of interest) can be determined. Furthermore, in some embodiments, a sharpening algorithm can be used to improve the resolution of the initial image, making it easier to identify structures of interest.

[0043] In step 206, a customized scanning strategy is determined based on the structure of interest identified in step 204. The scanning strategy includes a beam path that the charged particle beam will follow as the sample is scanned, with each portion of the beam path assigned a residence time based on whether or not that portion is associated with a structure of interest. Specifically, portions of the beam path associated with a structure of interest are assigned an increased residence time that allows for the acquisition of a statistically significant amount of second modality data related to the corresponding structure of interest as the charged particle beam passes through that portion of the beam path. For example, the customized scanning strategy may assign longer residence times to portions of the beam path corresponding to a structure of interest (e.g., portions corresponding to pixels, including only the centers of atoms or the edge structures of nanoparticles). The increased residence times may correspond to preset residence times assigned to all beam path portions corresponding to a structure of interest, or they may be determined based on the desired detector modality type / sensitivity, the type of structure of interest, the species of structure of interest, or a combination thereof. Furthermore, according to this disclosure, regions of the beam path not associated with a structure of interest are assigned a reduced residence time. Such reduced residence times may be sufficient to acquire data from the first modality used to generate the initial image, but the reduced residence times also reduce the time spent scanning portions of the sample that do not contain the structure of interest, and at the same time generally reduce damage to the sample.

[0044] In step 208, the sample is scanned again using a scanning strategy. Specifically, the charged particle beam irradiates a portion of the sample using the beam path and residence time specified in the customized scanning strategy. Thus, the longer residence time associated with the structure of interest allows for the acquisition of statistically significant data from the second modality's detector system, while simultaneously reducing the time spent scanning regions of the sample that do not contain the structure of interest. However, detector data from the first modality may be acquired based on the irradiation of regions determined not to contain the structure of interest due to the reduced residence time.

[0045] In 210, data obtained based on scanning is optionally position-aligned with data from previous scans. Specifically, data obtained based on irradiation of a region with reduced residence time can be used to position-align the generated image and / or acquired detector data during iterative scanning of a sample using a customized scanning strategy. For example, between each iteration of scanning a sample using a scanning strategy, changes in the orientation or positional alignment of the sample in the charged particle system may occur. According to some embodiments of the present disclosure, data obtained based on irradiation of a region not of interest can be compared with the same data obtained in previous scans to determine the relative position / orientation between the newly obtained image and the data from previous scans. This relative position / orientation can then be used to position-align the data of a second modality in orientation and / or other manner across multiple iterations. For example, based on drift determination, the positions associated with the data of a second modality can be adjusted to account for drift (e.g., different portions of the data of a second modality originate from the center of an atom).

[0046] In step 212, compositional information is determined based on data from a second detector modality obtained during scanning in step 208. For example, if the second modality corresponds to an EDS detector system, the EDS data can be used to determine the species of atoms in or part of the nanoparticles and / or the material composition.

[0047] In step 214, it is determined whether an additional scan of the sample is necessary. In some embodiments, this decision is based on whether there is enough detector data from the second modality to determine the composition of the structures of interest and / or a threshold number of structures of interest. In this way, the sample can be repeatedly scanned using a customized scanning strategy until compositional information is determined for all or a threshold number of structures of interest, or until it is determined that an additional scan would cause sample damage exceeding a threshold and / or as a result destroy (or have destroyed) one or more structures of interest. If the answer to step 214 is "yes", the process returns to step 208, and the sample is scanned again using the scanning strategy. If the answer to step 214 is "no", the process ends.

[0048] Figure 3 shows a series of diagrams illustrating process 300 for determining a customized scanning strategy based on an initial sample image. Figure 3 shows a depiction of the initial image of sample 302 obtained using detector data from the first modality. Specifically, the initial image 302 corresponds to an image of the sample containing arrays of two types of atoms generated using HAADF detector data. Figure 3 further shows the initial image with multiple overlays 310 and 312 corresponding to regions of the initial image determined to correspond to structures of interest of the first type (i.e., atoms) and structures of interest of the second type (i.e., atoms), respectively. Image 304 further shows representative pixels 314 determined to represent the corresponding regions (i.e., the central pixel, the pixel furthest from the edge, and / or pixels determined to be likely to return compositional information related to the corresponding structures of interest when irradiated in other ways).

[0049] Figure 3 also shows a mask 306 generated using regions of interest 310 and 312 and / or representative pixels 314. The mask 306 identifies regions of the sample to be irradiated to generate detector data for a second modality, which allows compositional information (or other information) to be acquired for the corresponding structures of interest (i.e., atoms). The black regions of the mask 306 correspond to regions scanned with fast residence times (e.g., background raster scanning such as HAADF signals, which can be used for drift compensation and verification of EDS / structure of interest positioning), while the white artifacts correspond to regions scanned with slow residence times (e.g., slow residence times that allow for the acquisition of sufficient EDS data to compositionally identify the structures of interest). The mask 306 distinguishes between representative pixels of a first species of structures of interest and representative pixels of a second species of structures of interest. However, in other embodiments, such a distinction may not be present in the mask.

[0050] Figure 3 also shows a customized scanning strategy 316 for a sample generated based on a mask / representative pixel. The scanning strategy 316 shows the beam path 318 that charged particles travel through as they irradiate the surface of the sample. Furthermore, image 308 shows each portion of the beam path 318 as associated with residence time (e.g., the time the charged particle beam irradiates the relevant portion of the sample during the execution of the scanning strategy). Specifically, the scanning strategy 316 includes a first residence time for a first beam path portion 320 associated with the center of an atom identified in the image, and a second residence time for a second beam path portion 322 not associated with a structure of interest.

[0051] Figure 4 shows a series of diagrams illustrating process 400 for determining a customized scanning strategy based on the initial sample image. Figure 4 shows a depiction of the initial image of sample 402 containing nanoparticles obtained using detector data from the first modality (HAADF detector data). Figure 4 further shows the initial image with multiple overlays 410 corresponding to regions of the initial image determined to correspond to structures of interest (i.e., nanoparticles). Image 404 further shows representative pixels 412 determined to represent the corresponding structures of interest (i.e., the central pixel, the pixel furthest from the edge, and / or pixels determined to be likely to return compositional information related to the corresponding structures of interest when irradiated in other ways).

[0052] Figure 4 also shows a mask 406 generated using the region of interest 410 and / or representative pixel 412. The mask 406 identifies regions of the sample to be irradiated to generate detector data for a second modality, which allows compositional information (or other information) to be acquired about the corresponding structures of interest (i.e., nanoparticles). The black regions of the mask 406 correspond to regions scanned with fast residence times (e.g., background raster scanning such as HAADF signals, which can be used for drift compensation and verification of EDS / structure of interest positioning), while the white artifacts correspond to regions scanned with slow residence times (e.g., slow residence times that allow for the acquisition of sufficient EDS data to compositionally identify the structures of interest).

[0053] Figure 4 also shows a customized scanning strategy 414 for a sample generated based on a mask / representative pixel. The scanning strategy 414 shows the beam path 414 that charged particles traverse as they irradiate the surface of the sample. The beam path 414 is generated so that each representative pixel is scanned while minimizing other areas being scanned. The scanning strategy 414 includes a first residence time for a first beam path portion 418 associated with the center of nanoparticles identified in the image, and a second residence time for a second beam path portion 420 not associated with the structure of interest.

[0054] Figure 5 shows a series of diagrams illustrating the process 500 for identifying structures of interest within the initial image. Figure 5 shows a depiction of the initial image of sample 502 containing nanoparticles obtained using detector data from the first modality (HAADF detector data). Figure 5 also includes a modified version 504 to which one or more sharpening algorithms, one or more image recognition algorithms, or a combination thereof have been applied. Specifically, image 504 corresponds to the result of a neural network sequentially trained to remove noise from charged particle images and segment such images based on whether the region is likely to contain atoms. The resulting image 504 shows the probability that each region of the sample and its neural network contains atoms, with regions determined to be more likely to contain atoms shown as bright and regions determined to be less likely to contain atoms shown as dark.

[0055] Image 506 corresponds to a version of Image 504 to which binary threshold probabilities have been applied. That is, areas determined to potentially contain atoms greater than the threshold are shown in white, and areas that do not meet or exceed this threshold are shown in black. Based on Image 506, an algorithm or user can determine which areas of the image correspond to atoms. Furthermore, in some embodiments, an algorithm or user can further identify areas corresponding to atoms of the same species. Figure 5 further shows an initial image with multiple overlays 510 and 512 corresponding to areas of the initial image determined to correspond to structures of interest of a first species (i.e., atoms) and structures of interest of a second species (i.e., atoms), respectively. For example, these overlays 510 and 512 may be determined by applying a fitting circle to the individual binary objects of Image 506. Image 508 further shows representative pixels 514 determined to represent corresponding areas (i.e., the central pixel, the pixel furthest from the edge, and / or pixels determined to be likely to return compositional information related to the corresponding structure of interest when illuminated in other ways). For example, representative pixel 514 corresponds to the center of the circle used to fit the binary object.

[0056] Figure 6 shows a series of diagrams illustrating the process 500 for identifying different species of structures of interest within the initial image. Figure 6 shows a depiction of the initial image of sample 602 containing nanoparticles obtained using detector data from the first modality (HAADF detector data). Image 504 is generated by applying a sequentially trained neural network and binary threshold stochastic algorithm to image 602 to denoise the charged particle image and segment it into regions likely to contain atoms. Regions determined to potentially contain atoms greater than the threshold are shown in white, and regions that do not meet or exceed this threshold are shown in black. Based on image 604, the algorithm or the user can determine which region of the image corresponds to an atom.

[0057] Figure 606 corresponds to a graph of the radius of the circle used to fit the individual binary objects in Figure 604 into a circle. Graph 606 shows that there appear to be two types of atoms, with the first type having a smaller radius 614 and the second type having a larger radius 616. Figure 608 shows the average radii of the first type atoms 614 and the second type atoms 616. In some embodiments, the algorithm can estimate the possible species of each atom based on the average atomic radius of the species. Similarly, Figure 610 shows the estimated crystal structure of the atoms in the sample.

[0058] Figure 6 further shows the initial image with multiple overlays 618 and 620 corresponding to regions of the initial image that were determined to correspond to the structure of interest 614 of the first species and the structure of interest 616 of the second species, respectively.

[0059] Figure 7 shows a series of diagrams illustrating a process 700 for determining the composition of structures of interest identified in an initial sample image. Figure 7 shows a depiction of an initial image 702 of a sample containing two types of atomic arrays generated using HAADF detector data. Figure 7 also shows a customized scanning strategy 710 for a sample generated using the method and system according to this disclosure. The scanning strategy 710 shows a beam path 712 through which charged particles irradiate the surface of the sample. Furthermore, image 704 shows each portion of the beam path 712 as associated with residence time (e.g., the time the charged particle beam irradiates the relevant portion of the sample during the execution of the scanning strategy). Specifically, the scanning strategy 712 includes a first residence time for a first beam path portion 714 associated with the centers of atoms identified in the image, and a second residence time for a second beam path portion 716 not associated with structures of interest.

[0060] Figure 706 corresponds to a graph of the EDS spectral peaks for each atom identified in Figure 702. In some embodiments, the composition of individual atoms is determined based on the peak EDS spectrum detected when the relevant region of the sample is irradiated with a charged particle beam. However, in other embodiments, it is necessary to obtain multiple sets of EDS data by repeatedly scanning the sample with a customized scanning strategy. For example, Figure 708 corresponds to a graph of the average EDS spectral peaks for each atom over multiple scans using a customized scanning strategy. In some embodiments, the sample may be repeatedly scanned with a customized scanning strategy until all atoms are identified, a threshold number is identified, they are identified above a confidence interval, or a combination thereof. As shown in Figure 708, the average EDS peak associated with atoms of the first species converges to a first value 718, and the average EDS peak associated with atoms of the second species converges to a second value 718. The compositions of the first and second species can then be determined using the first value 716 and the second value 718, respectively.

[0061] Figure 8 shows a series of diagrams illustrating an exemplary process 800 for determining the composition of a structure of interest in a sample according to the present disclosure. Figure 8 shows an initial image 802 of a sample containing two types of atomic arrays generated using HAADF detector data. Figure 8 also includes a modified version 802 to which one or more sharpening algorithms, one or more image recognition algorithms, or a combination thereof are applied. Specifically, image 804 corresponds to the result of a neural network sequentially trained to remove noise from charged particle images and segment such images based on whether the region is likely to contain atoms. The resulting image 804 shows the probability that the individual regions of the sample and the neural network contain atoms, with regions determined to be more likely to contain atoms shown as bright and regions determined to be less likely to contain atoms shown as dark.

[0062] Image 806 corresponds to a version of Image 804 to which binary threshold probabilities have been applied. That is, regions determined to potentially contain atoms greater than the threshold are shown in white, while regions that do not meet or exceed this threshold are shown in black. Based on Image 806, the algorithm or user can determine which regions of the image correspond to atoms. Furthermore, in some embodiments, the algorithm or user can further identify regions corresponding to atoms of the same species.

[0063] Image 808 corresponds to a graph of the radius of the circle used to fit the individual binary objects in Image 806 into a circle. Graph 808 shows that there appear to be two types of atoms, the first type having a smaller radius 820 and the second type having a larger radius 822. In some embodiments, the shape of the binary object can be used to estimate the possible species of each atom (e.g., by comparing the average radius to the atomic species that match that radius).

[0064] Figure 8 further includes image 810 corresponding to the initial image, which has multiple overlays 824 and 826 corresponding to regions of the initial image determined to correspond to structures of interest of a first species (i.e., atoms) and structures of interest of a second species (i.e., atoms), respectively. For example, these overlays 824 and 826 may be determined by applying a fit circle to the individual binary objects of image 806. Image 810 further shows representative pixels 828 determined to represent the corresponding regions (i.e., the central pixel, the pixel furthest from the edge, and / or pixels determined to be likely to return compositional information related to the corresponding structures of interest when illuminated in other ways). For example, the representative pixel 828 corresponds to the center of the circle used to fit the binary objects.

[0065] Figure 8 also shows a customized scanning strategy 830 for a sample generated based on a representative pixel 828. The scanning strategy 830 shows the beam path 832 through which the charged particles irradiate the surface of the sample. Furthermore, image 812 shows each portion of the beam path 830 as associated with residence time (e.g., the time the charged particle beam irradiates the relevant portion of the sample during the execution of the scanning strategy). Specifically, the scanning strategy 830 includes a first residence time for a first beam path portion 834 associated with the center of an atom identified in the image, and a second residence time for a second beam path portion 836 not associated with a structure of interest.

[0066] Figure 814 corresponds to a graph of the EDS spectral peaks for each atom identified in Figure 802. In some embodiments, the composition of individual atoms is determined based on the peak EDS spectrum detected when the relevant region of the sample is irradiated with a charged particle beam. However, in other embodiments, it is necessary to obtain multiple sets of EDS data by repeatedly scanning the sample with a customized scanning strategy. For example, Figure 814 corresponds to a graph of the average EDS spectral peaks for each atom over multiple scans using a customized scanning strategy. In some embodiments, the sample may be repeatedly scanned with a customized scanning strategy until all atoms are identified, a threshold number is identified, atoms are identified above a confidence interval, or a combination thereof.

[0067] Figure 816 is a graph showing the drift between different iterations of a sample scanned using a customized scanning strategy. The drift can be determined based on the data obtained when the sample was scanned using a second residence time (i.e., regions of the sample that do not contain the structure of interest). For example, the drift shown in Graph 816 was determined using HAADF data acquired when scanning regions of the sample that did not contain atoms. By comparing this HAADF data with previous iterations of the data acquired in previous scans, the relative drift of the sample / data between each iteration of the data can be determined. This allows for positional alignment of multiple iterations of the data / image, verification of the position of the EDS data corresponding to indexed atomic centers, and adjustment of the matching of the EDS data so that the correct peak values ​​are assigned to the centers of the corresponding atoms, or this can be used for a combination of them. For example, Figure 818 shows cumulative EDS data from multiple scans of a sample using a customized scanning strategy 830, with drift corrected using HAADF data. As shown in Figure 818, the mean EDS peak associated with atoms of the first species converges to a first value of 838, and the mean EDS peak associated with atoms of the second species converges to a second value of 840. The first value of 838 and the second value of 840 can then be used to determine the composition of the first and second species, respectively.

[0068] Examples of the subject matter of the present invention as described herein are described in the following paragraphs.

[0069] A1. A method for scanning a sample using a charged particle microscope system, the system comprising: scanning a surface region of the sample with a charged particle beam to acquire a first image of the surface region by a first detector modality; determining a scanning strategy for the surface region, the scanning strategy comprising: a charged particle beam path; a first beam dwell time associated with at least one region of interest in the first image, wherein the first beam dwell time is sufficient to acquire statistically significant data from a second detector modality; and at least a second beam dwell time associated with other regions of the first image, wherein the first beam dwell time is different from the second beam dwell time; scanning the surface region of the sample according to the determined scanning strategy; acquiring first data from a second detector modality based on the surface region of the sample being scanned according to the scanning strategy; and acquiring second data from the first modality based on the surface region of the sample being scanned according to the scanning strategy.

[0070] A1.1. The method according to paragraph A1, wherein determining a scanning strategy includes identifying at least one region of a first image containing a structure of interest.

[0071] A1.1.1. The method according to paragraph A1.1, wherein identifying at least one region of a first image is performed by segmenting the first image using a segmentation algorithm.

[0072] A1.1.1.1 The method described in paragraph A1.1.1, comprising: a word identifying at least one region of a first image; presenting a segmented first image to a user; and receiving a selection of one or more segments scanned by a second beam dwell time.

[0073] A1.2. The method according to any one of paragraphs A1 to A1.1, wherein identifying at least one region of a first image includes identifying multiple regions of a first image, each containing an instance of a structure of interest.

[0074] A1.3. The method according to any one of paragraphs A1.1 to A1.2, wherein identifying at least one region of a first image includes identifying one or more pixels associated with each instance of a structure of interest in the first image.

[0075] A1.3.1. The method described in paragraph A1.3, wherein identifying one or more pixels associated with each instance of the structure of interest in the first image corresponds to identifying a representative pixel associated with an instance of the structure of interest.

[0076] A1.3.1.1. The method according to paragraph A1.3.1, wherein the representative pixel is the central pixel of the corresponding instance of the structure of interest in the first image.

[0077] A1.3.1.2. The method according to paragraph A1.3.1, wherein the representative pixel is the pixel furthest from the boundary of the corresponding instance of the structure of interest in the first image.

[0078] A1.4. The method according to any one of paragraphs A1.1 to A1.3.1.2, wherein the structure of interest corresponds to an atom.

[0079] A1.4.1. The method described in paragraph A1.4, wherein the structure of interest corresponds to the species of atom.

[0080] A1.4.2. The method described in any one of paragraphs A1.4 to A1.4.1, wherein the region of interest corresponds to the center of an atom.

[0081] A1.5. The method according to any one of paragraphs A1.1 to A1.3.1.2, wherein the structure of interest corresponds to the compound.

[0082] A1.6. The method according to any one of paragraphs A1.1 to A1.3.1.2, wherein the structure of interest corresponds to a nanoparticle.

[0083] A1.7. The method according to any one of paragraphs A1.1 to A1.3.1.2, wherein the structure of interest corresponds to a portion of a nanoparticle.

[0084] A1.7.1. The method according to paragraph A1.7, wherein the nanoparticle portion is one of the following: boundary, edge, material transition region, and component substructure.

[0085] A1.8. A method according to any one of paragraphs A1 to A1.7.1, wherein identifying at least one region of a first image is performed by sharpening the first image using a sharpening algorithm.

[0086] A1.8.1. The method described in paragraph A1.8, wherein the sharpening algorithm includes a neural network.

[0087] A1.9. The method according to any one of paragraphs A1.1 to A1.8.1, wherein determining a scanning strategy includes identifying at least one additional region of a first image containing different structures of interest.

[0088] A1.9.1. The method according to paragraph A1.9, wherein identifying at least one additional region includes identifying multiple additional regions of a first image, each containing an instance of a different structure of interest.

[0089] A1.9.2. The method according to any one of paragraphs A1.9 to A1.9.1, wherein determining a scanning strategy involves identifying multiple sets of regions of interest of a first image, each set comprising one or more regions of the first image containing a corresponding structure of interest.

[0090] A1.9.3. The method described in any one of paragraphs A1.9 to A1.9.2, wherein the different structure of concern is of a different type from the above-mentioned structure of concern.

[0091] A1.9.4. The method described in any one of paragraphs A1.9 to A1.9.3, wherein the different structures of interest are of a different kind from the structures of interest described above.

[0092] A1.10. The method according to any one of paragraphs A1.1 to A1.9.4, wherein instances of individual structures of interest within a first image are identified by a trained neural network.

[0093] A1.11. The method according to any one of paragraphs A1.1 to A1.9.10, wherein determining the scanning strategy includes determining that a particular region is empty space and allocating a reduced beam residence time to that region.

[0094] A1.12. The method according to any one of paragraphs A1.1 to A1.9.11, wherein determining the scanning strategy includes determining that a particular region is not a region of interest and allocating a reduced beam residence time to that region.

[0095] A2. The method according to any one of paragraphs A1 to A1.12, wherein the scanning strategy includes scanning a portion of the sample corresponding to at least one region of interest in the first image by a second beam residence time.

[0096] A2.1. The method according to paragraph A2, wherein the second beam residence time induces at least a threshold emission required to obtain statistically significant data from the second detector modality.

[0097] A2.2. The method according to any one of paragraphs A2 to A2.1, wherein scanning a portion of the sample corresponding to at least one region of interest in the first image by a second beam residence time corresponds to scanning a portion of the sample corresponding to a representative pixel of the first image corresponding to at least a portion of the region of interest.

[0098] A2.3. The method according to any one of paragraphs A2 to A2.2, wherein the scanning strategy includes scanning a portion of the sample that does not correspond to at least one region of interest in the first image by a first beam residence time.

[0099] A2.4. The method according to any one of paragraphs A2 to A2.3, wherein the scanning strategy further includes a third beam dwell time associated with other regions of the first image associated with instances of different structures of interest, the third beam dwell time being different from the first beam dwell time and the second beam dwell time.

[0100] A2.5. The method according to any one of paragraphs A2 to A2.3, wherein the scanning strategy further includes multiple additional beam dwell times, each additional beam dwell time being associated with a corresponding region of a first image associated with an instance of a corresponding type or species of structure of interest.

[0101] A3. The method according to any one of paragraphs A1 to A2.5, wherein the method comprises determining the composition of at least one region of interest based on first data from a second detector modality.

[0102] A3.1. The method of paragraph A3, further comprising determining that one or more compositions of the region of interest cannot be identified by the first data and obtaining a new scan.

[0103] A3.1.1. The method according to paragraph A3.1, wherein obtaining a new scan involves scanning the surface area of ​​the sample in an additional number of passes according to the scanning strategy.

[0104] A3.1.2. The method according to any one of paragraphs A3.1 to A3.1.1, further comprising repeating the scanning of a surface region of the sample according to the scanning strategy until a threshold region of interest is identified.

[0105] A3.1.3. The method according to any one of paragraphs A3.1 to A3.1.1, further comprising repeating the scanning of the surface area of ​​the sample according to the scanning strategy until it is determined that additional scanning would cause damage to the sample.

[0106] A3.2. The method includes determining the composition of the structure of interest based on the first data from a second detector modality.

[0107] A3.2. The method described in any one of paragraphs A3 to A3.1, wherein the second modality is one of EDS and EELS.

[0108] A3.3. The method according to any one of paragraphs A3 to A3.2, wherein irradiating the sample with a second beam residence time causes an emission insufficient to determine the composition.

[0109] A4. The method according to any one of paragraphs A1 to A3.3, wherein the first detector modality is dark-field imaging.

[0110] A4.1. The method described in paragraph A4, wherein the first detector modality is HAADF.

[0111] A5 The second detector modality is EELS, EDS, Diffraction pattern, Cathode ray emission oh and backscattered electrons Detection method using one of the following One of the methods is the one described in one of paragraphs A1 to A4.1.

[0112] A6. The method according to any one of paragraphs A1-A5, further comprising determining a first beam residence time.

[0113] A6.1. The method according to paragraph A6, wherein the first beam residence time is determined based on the type, size, and / or species of the structure of interest.

[0114] A6.2. The method according to any one of paragraphs A6 to A6.1, wherein the first beam residence time is determined based on the type and / or sensitivity of the second modality.

[0115] A6.3. The method according to any one of paragraphs A6 to A6.2, wherein the first beam residence time is determined based on the material of the structure of interest.

[0116] A7. The method according to any one of paragraphs A1 to A6.3, wherein the strategy further comprises scanning a region of the sample corresponding to at least one region of interest in the first image with different beam intensities.

[0117] A8. The method according to any one of paragraphs A1 to A7, wherein the first image is obtained by scanning a surface region of a sample with a charged particle beam during a second beam residence time.

[0118] A9. The method described in any one of paragraphs A1-A8, wherein statistically significant data corresponds to data describing material properties, multiple material properties, and / or their relationships.

[0119] A9.1. The method described in paragraph A9, wherein statistically significant data include a quantity or quality of data that explains material properties, multiple material properties, and / or their relationships, which are not the result of statistical chance.

[0120] A9.2. The method according to any one of paragraphs A9 to A9.1, including a quantity or quality of data describing material properties, multiple material properties, and / or their relationships, which is sufficient to enable a compositional determination to be made with respect to a corresponding region of the sample within a threshold number of scans performed using the scanning strategy.

[0121] A9.2.1. The method according to paragraph A9.2, wherein the threshold number of scans is 1, 10, 25, or 100.

[0122] A9.2.2. The method described in any one of paragraphs A9.2 to A9.2.1, wherein the statistically significant data includes a quantity or quality of data sufficient to enable the determination of the species of atoms and / or the material composition of the nanoparticles or part thereof.

[0123] A10. The method according to any one of paragraphs A1 to A9.2.2, further comprising generating a second image using the second data and positionally aligning the first data using the second image.

[0124] A10.1. The method according to paragraph A10, wherein positional alignment of first data includes determining drift based on a comparison between a first image and a second image, and correcting the first data to account for the drift.

[0125] A11. The method according to any one of paragraphs A1 to A10.1, further comprising generating a second image using second data and determining the drift between the first image and the second image using the second image.

[0126] A12. The method according to any one of paragraphs A1 to A11, wherein the scanning strategy further includes a third beam dwell time associated with other regions of the first image, the third beam dwell time being different from the first beam dwell time and the second beam dwell time.

[0127] A12.1. The method according to paragraph A12, wherein, when other regions of the first image are scanned by a third dwell time, the third beam dwell time induces at least a threshold emission required to obtain statistically significant data from the third detector modality, and the third detector modality differs from the first and second detector modalities.

[0128] A12.2. The method according to any one of paragraphs A12-A12.1, wherein other areas of the first image are associated with instances of different structures of interest.

[0129] A12.3. The method according to any one of paragraphs A12 to A12.2, wherein the scanning strategy further includes one or more additional beam dwell times associated with other regions of the first image, and the one or more additional beam dwell times are determined to induce a threshold emission required to obtain statistically significant data from at least one or more additional detector modalities when other regions of the first image are scanned by the one or more additional beam dwell times.

[0130] A13. The method according to any one of paragraphs A1 to A12.3, wherein the first residence time and the second residence time have at least one different parameter such that the dose and / or irradiation time experienced by the portion of the sample irradiated by the first residence time and the second residence time are different.

[0131] A13.1. The method according to paragraph A13, wherein the residence time corresponds to the amount of time the charged particle beam irradiates the relevant portion of the sample during the execution of the scanning strategy.

[0132] A13.2. The method according to any one of paragraphs A13 to A13.1, wherein the residence time corresponds to the pixel density scanned by the charged particle beam when the relevant portion of the sample is irradiated.

[0133] A13.3. The method according to any one of paragraphs A13 to A13.2, wherein the residence time corresponds to the size of the pixels scanned by the charged particle beam when the relevant portion of the sample is irradiated.

[0134] A13.4. The method according to any one of paragraphs A13 to A13.3, wherein the residence time corresponds to the pulse length of the charged particle beam when irradiating the relevant portion of the sample.

[0135] A13.5. The method according to any one of paragraphs A13 to A13.4, wherein the residence time corresponds to the pulse frequency of the charged particle beam when irradiating the relevant portion of the sample.

[0136] B1. A system for investigating a sample using a charged particle microscope system, comprising: a sample holder configured to hold the sample; a charged particle source configured to emit a beam of charged particles toward the sample; an optical column configured to irradiate the beam of charged particles toward the sample; one or more detectors configured to detect the charged particles of the charged particle beam and / or emissions arising from the charged particle beam incident on the sample, wherein one or more detectors include at least a first detector of a first modality and a second detector of a second modality; one or more processors; and a memory for storing computer-readable instructions that, when executed by one or more processors, cause the system to perform the method described in any one of paragraphs A1 to A12.3.

[0137] C1. A non-temporary, computer-readable instruction that, when executed on one or more processors of a charged particle microscope system, causes one or more processors to perform an action as described in any one of paragraphs A1 to A12.3.

[0138] D1. Use the smart sample holder described in paragraph B1, performing the method described in any one of paragraphs A1 to A12.3.

Claims

1. A method for scanning a sample using a charged particle microscope system, The process involves scanning the surface region of a sample using a charged particle beam to acquire a first image of the surface region by a first detector of the first modality, The scanning strategy for the surface region is determined, and the scanning strategy is Charged particle beam path, A first beam dwell time associated with at least one region of interest in the first image, wherein the first beam dwell time is sufficient to obtain statistically significant data from a second detector of the second modality, To determine at least a second beam dwell time associated with other regions of the first image, wherein the first beam dwell time includes a second beam dwell time that is longer than the second beam dwell time. Scanning the surface region of the sample according to the scanning strategy, In a single pass, based on the surface region of the sample being scanned according to the scanning strategy, first data including information about the structure of interest is obtained from the second detector of the second modality, In the single pass, based on the surface region of the sample being scanned according to the scanning strategy, second data used for image generation is obtained from the first detector of the first modality, Includes, The method further includes determining and correcting the drift between the first image and the first data, at least partially based on the second data. The second modality is different from the first modality. method.

2. The method according to claim 1, wherein the structure of interest corresponds to one of an atom, a species of atom, a compound, a nanoparticle, and a part of a nanoparticle.

3. The method according to claim 1 or 2, further comprising determining that one or more compositions of the region of interest cannot be identified using the first data, and causing the surface region of the sample to be scanned in an additional number of passes according to the scanning strategy.

4. The method according to any one of claims 1 to 3, further comprising repeating the scanning of the surface region of the sample according to the scanning strategy until a threshold region of interest is identified.

5. The method according to any one of claims 1 to 4, further comprising determining the composition of a structure of interest based on the first data from the second detector of the second modality.

6. The method according to any one of claims 1 to 5, wherein the first modality is HAADF.

7. The method according to any one of claims 1 to 6, wherein the second detector of the second modality detects a diffraction pattern or backscattered electrons, or the second modality is one of EELS, EDS, or cathode ray emission.

8. The first beam residence time is Types of interest structures, Size of the structure of interest, Seeds of interest structures The second modality type, and The method according to any one of claims 1 to 7, further comprising determining based on at least one of the sensitivities of the second modality.

9. The method according to any one of claims 1 to 8, further comprising determining the first beam residence time based on the material of the structure of interest.

10. The method according to any one of claims 1 to 9, wherein the scanning strategy further comprises scanning the region of the sample corresponding to the at least one region of interest in the first image with different beam intensities.

11. The method according to any one of claims 1 to 10, wherein the first image is obtained by scanning the surface region of the sample with the charged particle beam during the second beam residence time.

12. The method described above is Using the second data, generate a second image, The method according to claim 11, further comprising determining the drift between the first image and the second image using the second image.

13. A system for scanning a sample using a charged particle microscope system, wherein the system A sample holder configured to hold a sample, A charged particle source configured to emit a charged particle beam toward the aforementioned sample, An optical column configured to irradiate the sample with the charged particle beam, One or more detectors configured to detect charged particles of the charged particle beam and / or emissions arising from the charged particle beam incident on the sample, wherein the one or more detectors include at least a first detector of a first modality and a second detector of a second modality. One or more processors, When executed by one or more processors, the system The surface region of the sample is scanned using a charged particle beam to obtain a first image of the surface region by the first detector of the first modality, The scanning strategy for the surface region is determined, and the scanning strategy is Charged particle beam path, A first beam dwell time associated with at least one region of interest in the first image, wherein the first beam dwell time is sufficient to obtain statistically significant data from the second detector of the second modality, To determine at least a second beam dwell time associated with other regions of the first image, wherein the first beam dwell time includes a second beam dwell time that is longer than the second beam dwell time. Scanning the surface region of the sample according to the scanning strategy, Based on the surface region of the sample being scanned according to the scanning strategy, first data including information about the structure of interest is obtained from the second detector of the second modality, The system includes a memory that stores computer-readable instructions for obtaining second data used for image generation from the first detector of the first modality based on the surface region of the sample being scanned according to the scanning strategy, Based at least partially on the second data, the drift between the first image and the first data is determined and corrected. The second modality is a system that differs from the first modality.

14. The system according to claim 13, wherein the instruction further causes the system to determine the composition of the structure of interest based on the first data from the second detector of the second modality.

Images