Data acquisition and processing techniques for three-dimensional reconstruction.

Abstract

To provide a method for performing a complete dataset of both of an image and a composition and / or crystal information in the high-speed processing time.SOLUTION: There is provided a generation method for generating data for three-dimensional reconstruction. The generation method at least includes exposing a subsequent surface of a sample, acquiring an image of the subsequent surface, comparing the image of the subsequent surface to an image of a reference surface, based on the comparison exceeding a threshold, acquiring a compositional or crystalline map of the subsequent surface, and based on the comparison not exceeding the threshold, exposing a next surface.SELECTED DRAWING: Figure 3

Description

[Technical Field]

[0001] The present invention relates generally to charged particle microscopes, and more particularly to sample processing for three-dimensional reconstruction in charged particle microscopes. [Background technology]

[0002] Charged particle microscopes can be used for repeated imaging and analysis techniques. One technique, simply called slice-and-view, involves collecting data on successive surfaces of a sample, in which a layer of the sample is removed after data on the exposed surface has been collected. This technique is widely used, for example, in materials research and biological sample analysis, and can provide a reconstructed sample volume after image processing. While this technique may not take very long when acquiring only images, acquiring other types of data, such as compositional or crystallographic data, increases the time required to acquire the desired data. For example, when a sample volume containing hundreds of layers is being processed and the desired data includes both images and elemental information, processing time can increase significantly to days or even weeks.

[0003] There have been previous attempts to improve throughput, but these efforts still present challenges. One prior solution, for example, involves acquiring elemental information for every n surfaces while imaging each surface. This can clearly improve processing time, but the elemental data is limited and may not capture variations in the sample. The choice of n can clearly affect the coverage of the elemental data, but at the expense of processing time and therefore throughput. While other potential solutions may exist, there is a desire to provide a complete data set of both images and compositional and / or crystallographic information with fast processing times. Summary of the Invention

[0004] Disclosed herein are apparatus and processes for generating data for three-dimensional reconstruction. An exemplary method includes at least exposing a subsequent surface of a sample, acquiring an image of the subsequent surface, comparing the image of the subsequent surface with an image of a reference surface, acquiring a compositional or crystallographic map of the subsequent surface based on a comparison that exceeds a threshold, and exposing a next surface based on a comparison that does not exceed a threshold.

[0005] An exemplary apparatus includes at least an electron column coupled to provide an electron beam, a focused ion beam column coupled to provide a focused ion beam at a plurality of detectors arranged to detect at least secondary electrons, x-rays, and backscattered electrons, and a controller coupled to at least the electron column, the focused ion beam column, and each detector of the plurality of detectors, the controller including or coupled to code that, when executed by the controller, causes the apparatus to: expose a subsequent surface of a specimen with the focused ion beam, acquire an image of the subsequent surface, compare the image of the subsequent surface with a reference image, acquire a compositional or crystallographic map of the subsequent surface based on a comparison exceeding a threshold, and expose a next surface based on a comparison not exceeding the threshold. [Brief explanation of the drawings]

[0006] [Figure 1] 1 is an exemplary dual beam system for generating multi-dataset 3D reconstructions, according to one embodiment of the present disclosure. [Figure 2] 1 is an exemplary sample processing workflow according to one embodiment of the present disclosure. [Figure 3] 1 is an exemplary method for acquiring multi-modal data while processing a sample, according to one embodiment of the present disclosure. [Figure 4] 1 is an exemplary image sequence according to one embodiment disclosed herein. [Figure 5]1 illustrates an exemplary workflow for propagating crystal and / or composition maps to an unmapped surface, according to an embodiment of the present disclosure. [Figure 6] 1 is an exemplary image sequence according to one embodiment of the present disclosure. [Figure 7] FIG. 2 is an exemplary functional block diagram according to one embodiment of the present disclosure.

[0007] Like reference numerals refer to corresponding parts throughout the several views of the drawings. DETAILED DESCRIPTION OF THE INVENTION

[0008] Embodiments of the present invention are described below in the context of a charged particle microscope that uses multiple data acquisition modalities to generate a three-dimensional (3D) reconstruction of a sample, such that the 3D reconstruction includes multiple data types. For example, a slice-and-view technique can be employed to image multiple surfaces of a sample, where each image is acquired after a layer of material is removed and further includes periodic acquisition of compositional and / or crystallographic maps of a subset of the surfaces. While not all surfaces are mapped to compositional or crystallographic properties, a neural network is used to determine which surfaces do map based on changes in the image relative to a threshold. Once or while images and maps are acquired, a second neural network is used to propagate the map to surfaces that do not have an associated map, using image pixel values, and mapping them to corresponding compositional or crystallographic map quantities. Once all surfaces have associated compositional and / or crystallographic maps, a 3D reconstruction of all data types is available. However, it should be understood that the methods described herein are generally applicable to a wide range of different tomography methods and devices, including both cone-beam and parallel-beam systems, and are not limited to any particular device type, beam type, object type, length scale, or scan trajectory.

[0009] As used in this application and the claims, the singular forms "a," "an," and "the" include the plural forms unless the context clearly dictates otherwise. Additionally, the term "comprises" means "comprises." Furthermore, the term "coupled" does not exclude the presence of intermediate elements between the coupled items.

[0010] The systems, devices, and methods described herein should not be construed as limiting in any way. Instead, the present disclosure is directed to all novel and non-obvious features and aspects of the various disclosed embodiments, alone and in various combinations and subcombinations with one another. The disclosed systems, methods, and devices are not limited to any particular aspect or feature or combination thereof, nor do the disclosed systems, methods, and devices require that any one or more particular advantages be present or problems be solved. While any theory of operation is for ease of explanation, the disclosed systems, methods, and devices are not limited to such theory of operation.

[0011] Although some operations of the disclosed methods are described in a particular order for convenience of presentation, it is to be understood that this description encompasses reordering unless a specific order is required by specific language described below. For example, operations described sequentially may, in some cases, be reordered or performed simultaneously. Moreover, for simplicity, the accompanying figures may not show the various ways in which the disclosed systems, methods, and apparatuses can be used with other systems, methods, and apparatuses. Additionally, the description sometimes uses terms such as "generate" and "provide" to describe the disclosed methods. These terms are high-level abstractions of the actual operations that are performed. The actual operations that correspond to these terms will vary depending on the particular implementation and will be readily discernible to those skilled in the art.

[0012] In some instances, values, procedures, or devices are referred to as "lowest," "best," "smallest," etc. Such descriptions are intended to indicate that a selection from among many functional options may be made, and it will be understood that no such selection is necessarily better, lesser, or otherwise desirable than other options.

[0013] Conventional techniques for generating 3D reconstruction data of a sample can be time-consuming, especially when multiple data types are desired for the full volume of the sample. For example, when performing a slice-and-view method to acquire the desired data, acquisition times can extend substantially from days to weeks, depending on the size of the sample volume. Of course, if only images of each surface are desired, processing time may not be an issue. However, if multiple types of data are desired for each exposed surface, acquisition times increase significantly beyond acceptable limits. Other types of data besides images can include compositional and / or crystallographic data. For example, compositional maps can be generated using energy-dispersive X-ray spectroscopy (EDS), wavelength-dispersive spectroscopy (WDS), or cathodoluminescence (CL), to name a few, while crystallographic maps can be generated using electron backscatter diffraction (EBSD). While it is possible to acquire multiple data types for each surface, processing times can become excessive. Alternatively, periodic acquisition of slower data types, such as composition or crystallography, can be performed at predetermined intervals while all surfaces are imaged. While this provides some slow data, information may be lost because such a process is performed agnostic to the sample or its modifications. Accordingly, it remains desirable to acquire multiple data points at each surface so that a multi-data 3D reconstruction can be generated.

[0014] One solution may be to monitor images of each exposed surface and determine whether significant changes in the sample result in changes in quantities associated with the reference surface. If the changes are deemed significant, a map (compositional and / or crystalline) of the newly exposed surface is acquired along with the image. However, if the changes are minor, a new surface is exposed and an image is acquired. The determination is then made again. Thus, a reference surface, with an associated reference set of data, is acquired based on the changes in the sample. Once the sample is fully processed, the reference surface dataset is used to generate maps of other surfaces for which only images were acquired. This interpolation can be performed using one or more neural networks that generate interpolated maps using the reference dataset and images of other surfaces. Once all surfaces have associated acquired or interpolated images and maps, a multi-dataset 3D reconstruction can be formed.

[0015] More specifically, the sample is processed using a slice-and-view technique that involves imaging and mapping a reference surface of the sample, removing a layer of the sample to expose a new surface, and imaging the new surface. After the new surface is imaged, a neural network compares the image of the new surface with the image of the reference surface to determine whether there has been a significant change in the sample, such as the appearance of a new feature or the expansion or contraction of a feature, to name a few. If the comparison indicates a significant change, a map of the new surface is also generated, and this new surface becomes the current reference surface. However, if the comparison does not indicate a significant change, the new layer is removed, and the newly exposed surface is imaged. The current reference surface remains the reference surface (rather than being replaced), and a comparison is made between the image of the newly exposed layer and the reference image. Then, based on the comparison, a map of the newly exposed surface is generated, or a subsequent surface is exposed. This process is repeated for the desired volume of the sample, resulting in a number of surfaces that have both an associated image and an associated map, while the remaining surfaces only have an associated image. Again, whether a surface has both an image and a map, or just an image, depends on a comparison of the surface image with that of the most recent reference image, e.g., whether significant changes have occurred.

[0016] Once the volume has been processed, one or more additional neural networks can be used to generate maps of surfaces that do not have associated maps, where the generated maps are based on interpolation from surrounding reference maps and their corresponding images. For example, two adjacent reference surface images and maps can be used in a neural network to associate pixel values ​​in the images with pixel values ​​in the maps. This association of pixel values ​​can then be used to generate maps of surfaces that only have associated images based on the pixel values ​​in those images. Once all surfaces have both images and maps, a 3D reconstruction of the sample can be formed, including both images and maps.

[0017] Using a neural network to determine both when to acquire maps for newly exposed surfaces and when to generate maps for surfaces that do not have associated maps increases the throughput of such slice-and-view processes while providing a complete and robust data set.

[0018] Alternatively, the process can acquire sparse crystal and / or composition maps at each layer in addition to, or instead of, a decision-based map acquisition approach. A sparse map can be a map with fewer than all pixels having data and / or a map with a fast acquisition time that results in noisy data. Then, once the sample volume is fully processed, the layer with the sparse map is processed through a neural network to generate a complete map. Because the sparse map can be acquired simultaneously with image acquisition, the sparse map process may not increase processing time.

[0019] FIG. 1 is an exemplary dual-beam system 100 for generating multi-dataset 3D reconstructions, according to one embodiment of the present disclosure. System 100 can be used to implement the data acquisition techniques discussed herein. In some embodiments, system 100 performs data acquisition and compositional and / or crystallographic map interpolation to provide a 3D reconstruction. However, in other embodiments, system 100 acquires the data, and map interpolation and 3D reconstruction are performed by a separate computing system, such as a user's desk computing system or a cloud-based computing system. While examples of suitable hardware are provided below, the present invention is not limited to being implemented on any particular type of hardware.

[0020] The SEM 141 is included in the dual beam system 100 along with a power supply and control unit 145. An electron beam 143 is emitted from a cathode 152 by applying a voltage between the cathode 152 and an anode 154. The electron beam 143 is focused to a fine spot by a condenser lens 156 and an objective lens 158. The electron beam 143 is scanned two-dimensionally over the specimen by a deflector 160. The operation of the condenser lens 156, the objective lens 158, and the deflector 160 is controlled by the power supply and control unit 145.

[0021] The electron beam 143 can be focused onto a substrate 122 that is located on a stage 125 in the lower chamber 126. The substrate 122 can be located on the surface of the stage 125 or on a TEM sample holder 124 that extends from the surface of the stage 125. When electrons in the electron beam strike the substrate 122, secondary electrons are emitted. These secondary electrons are detected by the detector 140. In some embodiments, a STEM detector 162 located below the TEM sample holder 124 and stage 125 collects electrons that transmit through the sample fixed on the TEM sample holder.

[0022] System 100 also includes a FIB system 111 comprising a vacuum chamber having an ion column 112 within which is located an ion source 114 and a focusing component 116 including an extraction electrode and an electrostatic optical system. The axis of the focusing column 116 may be tilted, for example, 52 degrees from the axis of the electron column 141. The ion column 112 includes an ion source 114, an extraction electrode 115, a focusing element 117, and a deflection element 120, which cooperate to form a focused ion beam 118. The focused ion beam 118 passes from the ion source 114 through the focusing component 116, between electrostatic deflection means, generally indicated at 120, and toward a substrate 122, which may comprise, for example, a semiconductor wafer positioned on a movable stage 125 within a lower chamber 126. In some embodiments, the sample may be placed on a TEM grid holder 124, where the sample may be a chunk extracted from the substrate 122. The chunk may then be further processed in the FIB according to the techniques disclosed herein.

[0023] The stage 125 can move in the horizontal plane (X and Y axes) and vertically (Z axis). The stage 125 can also tilt and rotate about the Z axis. In some embodiments, a separate TEM sample stage 124 can be used. Such a TEM sample stage is also preferably tiltable and rotatable as well as movable in the X, Y, and Z axes. In some embodiments, tilting of the stage 125 / TEM holder 124 can be in or out of the plane of the ion beam 118, and rotation of the stage is about the ion beam 118.

[0024] To insert substrate 122 onto stage 125, door 161 is opened. Depending on the tilt of stages 124 / 125, the Z axis will be in the direction of the optical axis of the associated column. For example, during the data collection phase of the disclosed technique, the Z axis will be in a direction parallel to the FIB optical axis, as indicated by ion beam 118. In such a coordinate system, the X and Y axes will be referenced from the Z axis. For example, the X axis may be in or out of the page depicting FIG. 1, and the Y axis may be in the page, but all three axes will maintain their perpendicular nature to one another throughout.

[0025] An ion pump 168 is used to evacuate the neck portion 112. The chamber 126 is evacuated using a turbomolecular and mechanical pumping system 130 under the control of a vacuum controller 132. The vacuum system pumps approximately 1×10 ions into the chamber 126. -7 Tor ~ 5 × 10 -4 Provide a vacuum of 1000 Torr. If etch-assisting, etch-retarding, or deposition precursor gases are used, the chamber background pressure is typically about 1 x 10 -5 It may rise to torr.

[0026] A high voltage power supply provides an appropriate acceleration voltage to electrodes in the focusing column 116 to energize and focus the ion beam 118. When it strikes the substrate 122, material is sputtered and physically ejected from the sample. Alternatively, the ion beam 118 can decompose a precursor gas to deposit material.

[0027] A high-voltage power supply 134 is connected to appropriate electrodes in the ion source 114 and ion beam focusing component 116 to form and direct an ion beam 118 of approximately 1 keV to 60 keV toward the sample. A deflection controller and amplifier 136, operating according to a predetermined pattern provided by a pattern generator 138, is coupled to the deflection plates 120, so that the ion beam 118 can be manually or automatically controlled to track a corresponding pattern on the top surface of the substrate 122. In some systems, the deflection plates are placed before the final lens, as is known in the art. Beam blanking electrodes (not shown) in the ion beam focusing column 116 cause the ion beam 118 to impinge on a blanking aperture (not shown) instead of the substrate 122 when a blanking controller (not shown) applies a blanking voltage to the blanking electrodes.

[0028] The ion source 114 typically provides an ion beam based on the type of ion source. In some embodiments, the ion source 114 is a liquid metal ion source capable of providing, for example, a gallium ion beam. In other embodiments, the ion source 114 may be a plasma-type ion source capable of delivering a number of different ion species, such as oxygen, xenon, and nitrogen, to name a few. The ion source 114 typically can be focused into a sub-1 / 10 micrometer wide beam at the substrate 122 or TEM grid holder 124 for modifying the substrate 122 by ion milling, ion-induced etching, material deposition, or for imaging the substrate 122.

[0029] Detectors 140 may be included to detect various types of emissions from the sample during processing and / or imaging. Such data may include secondary electrons, backscattered electrons, X-rays, and / or photons. To detect secondary electrons, a detector such as an Everhart Thornley or multichannel plate is connected to a video circuit 142 that provides a drive signal to a video monitor 144 and receives a deflection signal from the system controller 119. EBSD detectors known in the art may be used to detect backscattered electrons. X-rays may be detected with an EDX or WDS detector, and photons may be detected using any known optical detector tuned to various wavelength ranges. The locations of the various detectors 140 within the lower chamber 126 may vary in different embodiments.

[0030] The micromanipulator 147 can precisely move objects within the vacuum chamber. The micromanipulator 147 may include precision electric motors 148 positioned outside the vacuum chamber to provide X, Y, Z, and theta control of a portion 149 positioned within the vacuum chamber. The micromanipulator 147 can be fitted with different end effectors for manipulating small objects. In the embodiment described herein, the end effector is a thin probe 150.

[0031] A gas delivery system 146 extends into the lower chamber 126 for introducing and directing a gas vapor toward the substrate 122. For example, iodine can be delivered to enhance etching, or a metal organic compound can be delivered to deposit a metal.

[0032] A system controller 119 controls the operation of various portions of the system 100. Through the system controller 119, a user can cause the ion beam 118 or the electron beam 143 to scan in a desired manner through commands entered into a conventional user interface (not shown). Alternatively, the system controller 119 may control the system 100 according to programmed instructions stored in memory 121. The memory 121 may further include code for one or more neural networks for implementing the techniques disclosed herein. In some embodiments, the system 100 is coupled to one or more neural networks, either included in the memory 121 or via a network 170, to assist in decision-making and map interpolation / progression, both of which are discussed in more detail below. The controller 119 may also be coupled via the network 170 to a neural network 172, which may be stored on one or more servers. The network 170 may be, for example, an intranet, the Internet, a local area network, or a wide area network.

[0033] In operation in accordance with the techniques disclosed herein, the system 100 performs a slice-and-view process on the sample 122. A basic slice-and-view process involves imaging a surface of the sample using the SEM 141 or FIB 111, where secondary and / or backscattered electrons emitted from the sample form an image, and using the FIB 111 to remove a layer from the sample to expose a new surface. The new surface is then imaged, and this process is repeated until the desired volume of the sample has been imaged. In addition to the basic slice-and-view process, compositional and / or crystallographic maps can be formed for one or more of the surfaces before or after images of the surfaces are captured. For example, EDS and / or EBSD maps can be obtained for one or more surfaces based on a neural network determination of whether the sample has changed in terms of structure, composition, or morphology.

[0034] To further illustrate, a surface of a sample is imaged with the SEM 141, and EDS and / or EBSD maps are acquired from the same sample. This surface may be referred to as a reference surface with associated reference data, since it has both an associated image and an associated map. After imaging and mapping, a layer or slice of the sample is milled using the FIB 111 to remove a layer of the sample. The removal of material exposes a new surface. This new surface is then imaged, for example, using the SEM 141. The neural network then compares the image of the new surface with the image of the reference surface to determine whether there has been a significant change in the visible features in the image of the new surface. In some embodiments, a significant change is based on a threshold criterion for the difference between the two images, where the threshold may be 0.5 on a scale of 0 to 1.0. For example, the appearance of new structures or different morphologies in the new image may be assigned a number greater than 0.5 based on the magnitude of the new features / morphologies, while the absence of such changes is assigned a number less than 0.5. Based on the comparison, the neural network determines the next step the system 100 should perform.

[0035] If the comparison indicates that no significant changes exist (e.g., the difference between the two images is less than 0.5), the system 100 again uses the FIB 111 to remove another layer from the sample and images the newly exposed surface. Note that other forms of layer removal can also be used, such as a microtome to slice a layer, a laser to ablate a layer, or wide ion beam slicing. All such layer removal techniques are discussed herein.

[0036] However, if the comparison indicates a significant change (e.g., the difference between the two images is 0.5 or greater), an EDS and / or EBSD map is acquired from the newly exposed surface. This newly exposed surface then becomes the current reference surface. The newly exposed surface is then imaged and mapped, a new layer is removed, and the process is repeated. This sequence of imaging, determination, layer removal, mapping, etc. is repeated for the desired volume of the sample, resulting in a set of data for the sample volume. Note that the set of data includes a subset of surfaces that have both associated images and maps, while the remaining surfaces only have associated images. At this point, 3D reconstruction using only the images is possible, although there may be significant gaps in the map data for the corresponding 3D reconstruction.

[0037] This lack of a map for each surface is overcome by using one or more additional neural networks to create maps for surfaces that do not have maps. For example, an adjacent reference surface and its associated reference data set can be used to create an EDS / EBSD map for the surface without a map. Specifically, one or more neural networks can associate map pixel values ​​with image pixel values ​​based on the reference data set, and then use this association to generate a map for the surface without a map. Once the data for the entire volume is processed, a map for each surface is obtained, where some surfaces are associated with an acquired map and other surfaces are associated with an interpolated map. Once each surface has a map, a corresponding 3D reconstruction can be obtained.

[0038] Alternatively or additionally, sparse EDS and / or EBSD maps of each non-reference surface can be acquired simultaneously with image acquisition, and such sparse data can then be filled in when the reference map is used to back-propagate the map data to the non-reference surfaces.

[0039] 2 is an exemplary sample processing workflow 201 according to one embodiment of the present disclosure. Workflow 201 illustrates a sample milling, imaging, and data collection technique, which may also be referred to as slice-and-view, as described above. The disclosed technique uses slice-and-view to acquire data on a sample and build a 3D reconstruction including multiple data types. Workflow 201 can be performed, for example, by a dual-beam charged particle microscope, such as system 100, and further addresses data collection aspects of the disclosed technique. 3D reconstruction and data propagation aspects are discussed in more detail below.

[0040] Workflow 201 is performed on sample 222, which is an example of sample 122 and may be fixed on a stage, such as stage 125. Workflow 201 uses ion beam 218 to remove material from sample 222 and then images the newly exposed surface 272 with electron beam 243. After removing each layer or slice of sample 222, a newly exposed surface 272 is formed. For example, ion beam 218 removes slice 274A to expose new surface 272. Removal of slice 274A and subsequent slices can be performed by milling the slice using ion beam 218, or can be performed using ion beam induced etching by flowing a gas precursor onto surface 272, which is then etched through interaction with ion beam 218. Each slice can be approximately 2-5 nanometers thick, although the size of the structures within sample 222 can dictate the desired slice thickness. For example, smaller structures may require thinner slices, while larger structures can tolerate thicker slices. As the workflow progresses, slice 274B is removed and then an image is acquired of the newly exposed surface 272. This two-step process can then be repeated to remove slices 247C and 274D.

[0041] In addition to electron beam imaging, workflow 201 may further include acquiring several compositional and / or crystallographic maps of the surface based on the above determinations. The maps may be acquired in response to irradiating the surface with an electron beam, measuring backscattered electrons in the case of an EBSD map, or measuring the energy of emitted X-rays in the case of an EDS map. While not every surface may have an associated map, workflow 201 outlines a general process for acquiring images and maps of surfaces 272 of sample 222. An alternative or additional approach is to acquire sparse compositional and / or crystallographic maps of each non-reference surface, where such sparse data can be supplemented and combined with backpropagation of the reference map.

[0042] FIG. 3 illustrates an exemplary method 301 for acquiring multi-modal data while processing a sample, according to one embodiment of the present disclosure. Method 301 can be implemented in any type of microscope environment capable of acquiring multiple types of data for a sample. An exemplary microscope includes dual-beam system 200. Regardless of the microscope used, method 301 acquires at least two types of data for a sample on a subset of the surface, and acquires only one type of data for the remaining surface. Method 301 outlines an embodiment of a method for determining when to acquire a second type of data when a sample is being processed in a slice-and-view workflow. Implementing such a technique would result in 3D data for the sample at higher quality and with more efficient throughput than currently possible. Of course, it is possible to acquire all types of data for each surface, but such brute-force techniques would not benefit from the disclosed technology.

[0043] Method 301 may begin at process block 303, which involves acquiring a compositional and / or crystallographic map of a surface of a reference surface of the sample. Acquiring the compositional and / or crystallographic map can be performed using various data acquisition techniques, as described above, and may include at least, for example, EDS, WDS, EBSD, and / or CL, although other compositional or crystallographic techniques are also contemplated herein and within the scope of the present disclosure. Generally, the technique used to collect the map may be referred to as a second modality, distinct from the first modality used in the image acquisition process. The acquired map provides a pixelated map of the sample, relating the composition of the sample at each pixel and / or the crystallographic orientation of the sample at each pixel. Using such data, an associated 3D volume of the sample can be constructed as additional surfaces are exposed and imaged.

[0044] Process block 303 may be followed by process block 305, which involves acquiring an image of the reference surface. The image may be acquired using a first modality, which is traditionally faster than the second modality. The first modality may be an SEM- or FIB-acquired image of the surface, which provides more topographical type information of the surface. This data is also used to generate a 3D volume of the sample. Note that the sequence of process blocks 303 and 305 may be switched, and the sequence shown in FIG. 4 is not required or limiting. Also note that once the initial reference surface is acquired, process block 305 may not be executed after decision block 313, since an image may already have been acquired in process block 309.

[0045] Process block 305 may be followed by process block 307, which includes exposing a subsequent surface of the specimen. Exposing the subsequent surface may include removing a layer of material from the specimen using an available technique. For example, a FIB may be used to mill or etch away a layer to expose the subsequent surface. Alternatively, a laser may be used to ablate a layer of material, or a microtome may be used to slice a layer of material. Regardless of the material removal technique utilized, the subsequent surface is exposed so that at least an image of the subsequent surface may be acquired.

[0046] Process block 307 may be followed by process block 309, which includes acquiring a subsequent image of the surface. As described above, the image is acquired using a first modality. The first modality may be imaging using secondary and / or backscattered electrons emitted from the surface in response to irradiation with an electron beam (e.g., SEM-based imaging) or a FIB beam (e.g., FIB-based imaging). Regardless of the first modality, a grayscale image of the surface is acquired, where pixel-by-pixel grayscale differences indicate differences in morphology and / or material.

[0047] Note that in some examples of method 301, a sparse map of compositional and / or crystallographic data may be acquired simultaneously with image acquisition steps 305 and 309. The sparse map may have data about sparsely located regions of the sample surface. For example, when an image of the surface is acquired with a pixel-by-pixel strategy (e.g., by scanning the electron beam across the surface in a desired pattern, with the beam dwelling on pixels, i.e., dwelling on identifiable locations on the surface), EDS and / or EBSD data may be acquired at a subset of pixels to obtain the sparse map. This sparse data may then be augmented, for example, with data propagation method 501.

[0048] Process block 309 may be followed by process block 311, which involves comparing the image of the subsequent surface with the image of the reference surface. The comparison may be performed by one or more neural networks, such as a Siamese network, and may compare the images pixel by pixel. Generally, the comparison determines whether there has been a significant change in the structure visible on the surface exposed by the removal of the layer, which may be indicated in the image. Thus, significant differences in the images are determined by the comparison, where the significant difference is based on a threshold. For example, the comparison may determine differences between the images, where a difference above a threshold, e.g., a threshold of 0.5 on a scale of 0 to 1, indicates a significant change, while a difference below the threshold indicates no significant change. A difference value equal to the threshold also indicates a significant change.

[0049] Process block 311 may be followed by process block 313, which includes determining the next process to perform based on the comparison. If the comparison exceeds a threshold, a compositional and / or crystallographic map of the subsequent surface is obtained. As a result, the subsequent surface now has an associated image and associated map. In addition, the subsequent surface becomes the next reference surface. However, if the comparison does not exceed the threshold, the method proceeds to process block 307, whereby an additional surface is exposed.

[0050] Alternatively, the reference surface may be referred to as a current reference surface because it has both an image and one or more associated maps. Generally, method 301 involves data acquisition (both maps and images) of a subset of surfaces of a sample, where each surface in the subset of surfaces will have an associated image and crystal / composition map. The remaining surfaces may only have associated images.

[0051] Note that method 301 can be used to expose and image / map multiple surfaces, where a subset of the surfaces become reference surfaces because they have associated maps and images. The remaining surfaces may only have associated images. Each time the comparison exceeds a threshold, a new reference image is identified due to changes in surface morphology. Note that there is no predetermined number or periodicity of reference surfaces; the resulting reference surface is determined by changes in the sample as it is processed. Additionally, once a new reference surface is identified, subsequently exposed surfaces are compared to the new reference surface, which is not one of the previous reference surfaces.

[0052] 4 is an exemplary image sequence 401 according to one embodiment disclosed herein. Image sequence 401 includes data for four consecutive surfaces, where two of the surfaces have associated images and maps, and the intervening surface has only an associated image. The images and maps of image sequence 401 were acquired using method 301 as implemented in a dual-beam charged particle microscope such as DB system 100.

[0053] Image sequence 401 includes data for surfaces 403, 405, 407, and 409. Surfaces 403 and 409 are reference surfaces and have associated images and maps, such as crystal and composition maps. For example, surface 403 has an associated image 411 and two associated composition maps 413 and 415. The two composition maps 413, 415 are, for example, of a single element. The composition map for surface 403 is split into two, one for each of the two elements, although such bifurcation of the maps is not required.

[0054] Surfaces 405 and 407 only have associated images 417 and 419, respectively. Finally, surface 409, like surface 403, has associated image 421 and associated composition maps 423 and 425.

[0055] The data associated with surfaces 403-409 may be part of a larger dataset for the associated sample and can be used to provide a multi-dataset 3D reconstruction of the sample.

[0056] 5 is an exemplary workflow 501 for propagating crystal and / or composition maps to an unmapped surface, according to one embodiment of the present disclosure. Workflow 501 can be performed in real time during the slice-and-view process as data is acquired. Alternatively, workflow 501 can be performed after the sample has been processed and all of the desired data has been acquired. In such an embodiment, map propagation can be performed by the microscope or at a user's desk with access to one or more neural networks. Generally, method 501 results in the generation of a composition map for a surface that does not have an associated acquired map, resulting in a 3D reconstruction with both an image and a composition / crystal map.

[0057] Workflow 501 includes reference images and maps 503, neural networks 505 and 507, images 509, and interpolated map 511. Reference images and maps 503 are images and crystal / composition maps of a subset of surfaces of a sample identified as reference surfaces. As such, each reference surface has an associated image and crystal and / or composition map. Additionally, the reference surfaces and their associated images / maps may be ordered from a first reference surface to an nth reference surface, such that the first and second reference surfaces have one or more intervening surfaces between them. In general, each reference surface other than the first and nth reference surfaces may have preceding and succeeding reference surfaces. For ease of discussion, each reference surface has a set of data, an image, and one or more maps, and such unit data is referred to as RS. i where RS denotes the reference set and i denotes the identification number of the reference surface. Generally, the reference set is RS i-1 , R.S. i , and R.S. i+1 can be identified as:

[0058] Image 509, on the other hand, is an image of an intervening surface that may not have an associated map of either crystallography or composition. Therefore, each intervening surface has two corresponding reference surfaces, one in front of the intervening surface and one behind the intervening surface. Image 509 is similarly represented as I1, I2, I3, ..., I m , or more generally I j , where j is from 1 to m. As previously discussed, in some instances, the intervening surface may also have a sparse map of compositional and / or crystallographic information that can be augmented using the interpolation process disclosed with workflow 501.

[0059] Typically, there will be multiple surfaces that are imaged, such as using method 301, and a subset of these surfaces will also be mapped. Any remaining surfaces will only have associated images. The subset of surfaces that have both associated images and associated maps will become reference images that will be used to form maps using neural networks 505 and 507 for the surfaces that only have images. Neural networks 505 and 507 may both be U-Nets, as known in the art, and have been trained using similar data.

[0060] In operation, the neural network 505 receives as input a reference image and a map 503, e.g., RS i From RS i+1 509. Bidirectional optical flow essentially uses two data sets to map pixel values ​​to associated image pixel values. This mapping is then used as input to neural network 507, which also receives image 509 as input. Neural network 505 then calculates the bidirectional optical flow to RS i and Image I j Between and I j and RS i+1Generate sophisticated bidirectional optical flow between j RS i and R.S. i+1 The neural network 507 further includes a surface between surfaces associated with RS i and R.S. i+1 As a result, and by applying the refined optical flow and visibility maps to the maps, an interpolated map for each intervening surface (having only image 509) can be obtained from workflow 501.

[0061] When viewed from different lenses, workflow 501 associates image grayscale pixel values ​​with composition / crystal map pixel values, and then uses this association to generate a map for each image in images 509. The end result is that each surface of the sample has both an associated map and image, where some maps are acquired and some maps are interpolated.

[0062] 6 is an example image sequence 601 according to one embodiment of the present disclosure. Image sequence 601 may be a continuation of image sequence 401, for example, including interpolated maps. Surfaces 605 and 607 may have interpolated maps generated using workflow 501, for example. For example, images and maps of surfaces 603 and 609 may be provided to workflow 501 along with images of surfaces 605 and 607, such that maps 633 and 635 may be interpolated for surface 605. Similarly, surface 607 may also have an interpolated map generated. In general, images and maps of surfaces 603 and 609 may be used to generate interpolated maps for surfaces 605 and 607.

[0063] FIG. 7 is an exemplary functional block diagram 700 according to one embodiment of the present disclosure. FIG. 7 is a block diagram illustrating a computer system 700 that an embodiment of the present disclosure may include. Computing system 700 may be an example of computing hardware included in system 100, such as controller 119. Computer system 700 includes at least a bus 740 or other communication mechanism for communicating information and a hardware processor 742 coupled with bus 740 for processing information. Hardware processor 742 may be, for example, a general-purpose microprocessor. Computing system 700 may be used to implement the methods and techniques disclosed herein, such as methods 301 and 401, and may also be used to acquire images and segment images into one or more classes.

[0064] Computer system 700 also includes a main memory 744, such as a random access memory (RAM) or other dynamic storage device, coupled to bus 740 for storing information and instructions executed by processor 742. Main memory 744 may also be used for storing temporary variables or other intermediate information during execution of instructions executed by processor 742. Such instructions, when stored on a non-transitory storage medium accessible to processor 742, render computer system 700 a special-purpose machine customized to perform the operations specified in the instructions.

[0065] Computer system 700 further includes a read-only memory (ROM) 746 or other static storage device coupled to bus 740 for storing static information and instructions for processor 742. A storage device 748, such as a magnetic disk or optical disk, is provided and coupled to bus 740 for storing information and instructions.

[0066] Computer system 700 can be coupled via bus 740 to a display 750, such as a cathode ray tube (CRT), for displaying information to a computer user. An input device 752, including alphanumeric and other keys, is coupled to bus 740 for communicating information and command selections to processor 742. Another type of user input device is a cursor control 754, such as a mouse, trackball, or cursor direction keys, for communicating directional information and command selections to processor 742 and for controlling cursor movement on display 750. This input device typically has two degrees of freedom in two axes, a first axis (e.g., x) and a second axis (e.g., y), that allow the device to specify a position in a plane.

[0067] Computer system 700 may implement the techniques described herein using customized hardwired logic, one or more ASICs or FPGAs, firmware and / or program logic in combination with a computer system to make or program computer system 700 a special-purpose machine. According to one embodiment, the techniques described herein are performed by computer system 700 in response to processor 742 executing one or more sequences of one or more instructions contained in main memory 744. Such instructions may be read into main memory 744 from another storage medium, such as storage device 748. Execution of the sequences of instructions contained in main memory 744 causes processor 742 to perform the process steps described herein. In alternative embodiments, hardwired circuitry may be used in place of or in combination with software instructions.

[0068] The term "storage medium," as used herein, refers to any non-transitory medium that stores data and / or instructions that cause a machine to operate in a specific manner. Such storage media may include non-volatile media and / or volatile media. Non-volatile media include, for example, optical or magnetic disks, such as storage device 748. Volatile media include dynamic memory, such as main memory 744. Common forms of storage media include, for example, floppy disks, flexible disks, hard disks, solid-state drives, magnetic tape or any other magnetic data storage medium, CD-ROMs, any other optical data storage medium, any physical medium with a pattern of holes, RAM, PROM, EPROM, FLASH®-EPROM, NVRAM, any other memory chip or cartridge, content addressable memory (CAM), and ternary content addressable memory (TCAM).

[0069] Storage media are distinct from but may be used in conjunction with transmission media. Transmission media involves transferring information between storage media. For example, transmission media include coaxial cables, copper wire and fiber optics, including the wires that comprise bus 740. Transmission media can also take the form of acoustic or light waves, such as those generated during radio wave and infrared data communications.

[0070] Various forms of media may be involved in carrying one or more sequences of one or more instructions to processor 742 for execution. For example, the instructions may initially be carried on a magnetic disk or solid state drive of a remote computer. The remote computer can load the instructions into its dynamic memory and send the instructions over a telephone line using a modem. A modem local to computer system 700 can receive the data on the telephone line and convert the data to an infrared signal using an infrared transmitter. An infrared detector can receive the data carried in the infrared signal and appropriate circuitry can place the data on bus 740. Bus 740 carries the data to main memory 744, from which processor 742 retrieves and executes the instructions. The instructions received by main memory 744 may optionally be stored on storage device 748 either before or after execution by processor 742.

[0071] Computer system 700 also includes a communication interface 756 coupled to bus 740. The communication interface 756 provides a two-way data communication coupling to a network link 758 that is connected to a local network 760. For example, communication interface 756 may be an Integrated Services Digital Network (ISDN) card, cable modem, satellite modem, or a modem that provides a data communication connection to a corresponding type of telephone line. As another example, communication interface 756 may be a local area network (LAN) card to provide a data communication connection to a compatible LAN. A wireless link may also be implemented. In such an implementation, communication interface 756 sends and receives electrical, electromagnetic, or optical signals that carry digital data streams representing various types of information.

[0072] Network link 758 typically provides data communication through one or more networks to other data devices. For example, network link 758 may provide a connection through local network 760 to a host computer 762 or to data equipment operated by an Internet Service Provider (ISP) 764. ISP 764, in turn, provides data communication services through the worldwide packet data communication network now commonly referred to as the "Internet" 766. Local network 760 and Internet 766 both use electrical, electromagnetic, or optical signals that carry digital data streams. The signals through the various networks and the signals on network link 758 and through communication interface 756, which carry the digital data to and from computer system 700, are exemplary forms of transmission media.

[0073] Computer system 700 can send messages and receive data, including program code, through the network(s), network link 758 and communication interface 756. In the Internet example, a server 768 might transmit a requested code for an application program through Internet 766, ISP 764, local network 760 and communication interface 756.

[0074] The received code may be executed by processor 742 as it is received, and / or stored in storage device 748, or other non-volatile storage for later execution.

[0075] The embodiments discussed herein to illustrate the disclosed technology should not be considered limiting and merely provide examples of implementation. For example, instead of or in addition to obtaining a complete map of a subset of surfaces, a sparse map can be obtained from all surfaces, which is then refined using a neural network to complete the map. Those skilled in the art will recognize myriad other ways in which the disclosed technology can be implemented that are contemplated herein and are within the scope of this disclosure.

[0076] An exemplary method for generating data for three-dimensional reconstruction includes at least exposing a subsequent surface of a sample, acquiring an image of the subsequent surface, comparing the image of the subsequent surface with an image of a reference surface, acquiring a compositional or crystallographic map of the subsequent surface based on a comparison that exceeds a threshold, and exposing the next surface based on a comparison that does not exceed a threshold.

[0077] The above exemplary method further includes obtaining an image of the reference surface and obtaining a crystallographic or compositional map of the reference surface.

[0078] The exemplary method above, further including: acquiring an image of a next surface; comparing the image of the next surface to a reference image; acquiring a compositional or crystallographic map of the next surface based on a comparison that exceeds a threshold; and exposing additional surfaces based on a comparison that does not exceed a threshold.

[0079] The exemplary method above, wherein comparing the subsequent image of the surface with the image of the reference image includes determining whether a difference between the images is greater than a threshold value.

[0080] The exemplary method above, wherein exposing the subsequent surface comprises milling a layer of the sample with a focused ion beam, the subsequent surface being exposed upon milling of the layer.

[0081] The above exemplary method, further comprising repeating the steps of exposing, acquiring, and comparing the plurality of surfaces, such that a subset of the surfaces of the plurality of surfaces have associated images and associated crystal or composition maps, and the remaining surfaces of the plurality of surfaces have associated images.

[0082] The exemplary method above, further comprising propagating the crystal or composition map to remaining surfaces of the plurality of surfaces using one or more neural networks.

[0083] The exemplary method above, wherein propagating the crystal or composition map to remaining surfaces of the plurality of surfaces using one or more neural networks includes interpolating crystal or composition maps from a subset of surfaces of the plurality of surfaces using one or more neural networks to form crystal or composition maps for the remaining surfaces of the plurality of surfaces.

[0084] The above example method, wherein propagating the crystals or composition maps to the remaining surfaces of the plurality of surfaces using one or more neural networks includes: using one or more neural networks to associate pixel values ​​of the crystals or composition maps of each surface of the subset of surfaces with pixel values ​​of an image of each surface of the subset of surfaces; and using one or more neural networks to assign pixel values ​​of the crystals or composition maps to pixel values ​​of an image of one of the remaining surfaces based on the associated pixel values.

[0085] The above exemplary method, wherein at least the process of comparing the subsequent surface image with the reference surface image is performed by a neural network.

[0086] The exemplary method above, wherein acquiring the subsequent surface image includes acquiring the subsequent surface image with an electron beam or a focused ion beam.

[0087] The exemplary method above, wherein obtaining a compositional or crystallographic map of the subsequent surface includes obtaining a compositional map of the subsequent surface using energy dispersive X-ray spectroscopy or cathodoluminescence, and obtaining a crystallographic map of the subsequent surface using electron backscatter diffraction.

[0088] The exemplary method above, further comprising obtaining a subsequent sparse compositional or crystallographic map of the surface.

[0089] An exemplary apparatus for generating data for three-dimensional reconstruction includes at least an electron column coupled to provide an electron beam, a focused ion beam column coupled to provide a focused ion beam at a plurality of detectors arranged to detect at least secondary electrons, x-rays, and backscattered electrons, and a controller coupled to at least the electron column, the focused ion beam column, and each detector of the plurality of detectors, the controller including or coupled to code that, when executed by the controller, causes the apparatus to: expose a subsequent surface of a sample with the focused ion beam, acquire an image of the subsequent surface, compare the image of the subsequent surface to a reference image, where the reference image is an image of the reference surface, acquire a compositional or crystallographic map of the subsequent surface based on the comparison exceeding a threshold, and expose a next surface based on the comparison not exceeding the threshold.

[0090] The exemplary apparatus above, wherein the controller further includes or is coupled to code that, when executed by the controller, causes the apparatus to acquire an image of the reference surface and to acquire a crystallographic or compositional map of the reference surface.

[0091] The example apparatus above, wherein the controller further includes or is coupled to code that, when executed by the controller, causes the apparatus to: acquire an image of a next surface; compare the image of the next surface to a reference image; obtain a composition or crystallographic map of the next surface based on a comparison that exceeds a threshold; and expose a subsequent surface based on a comparison that does not exceed the threshold.

[0092] The exemplary apparatus above, wherein comparing the subsequent image of the surface with the reference image includes determining whether a difference between the images is greater than a threshold value.

[0093] The above example apparatus, wherein the controller further includes or is coupled to code that, when executed by the controller, causes the apparatus to repeat the steps of exposing, acquiring, and comparing the plurality of surfaces, such that a subset of the surfaces of the plurality of surfaces have associated images and crystal or composition maps, and the remaining surfaces of the plurality of surfaces have associated images.

[0094] The exemplary apparatus above, wherein the controller further includes or is coupled to code that, when executed by the controller, causes the apparatus to propagate the crystal or composition map to remaining surfaces of the plurality of surfaces using one or more neural networks.

[0095] The example apparatus described above, wherein the controller further includes or is coupled to code that, when executed by the controller, causes the apparatus to use one or more neural networks to interpolate crystal or composition maps from a subset of surfaces of the plurality of surfaces to form crystal or composition maps for remaining surfaces of the plurality of surfaces.

[0096] The example apparatus described above, wherein the controller further includes or is coupled to code that, when executed by the controller, causes the apparatus to: use one or more neural networks to associate pixel values ​​of a crystal or composition map of each surface of the subset of surfaces with pixel values ​​of an image of each surface of the subset of surfaces; and use one or more neural networks to assign pixel values ​​of the crystal or composition map to pixel values ​​of an image of one of the remaining surfaces based on the associated pixel values.

[0097] The above exemplary apparatus, wherein the controller includes or is coupled to one or more neural networks trained to compare the image of the surface following the step with a reference image, the reference image being an image of a reference surface, and provides control signals to the apparatus based on the comparison.

[0098] The exemplary apparatus above, wherein the controller further includes, or is coupled to, code that, when executed by the controller, causes the apparatus to obtain a subsequent compositional map of the surface using energy dispersive X-ray spectroscopy or cathodoluminescence, and to obtain a subsequent crystallographic map of the surface using electron backscatter diffraction.

[0099] The apparatus of any preceding claim, wherein the controller further includes or is coupled to code that, when executed by the controller, causes the apparatus to obtain a subsequent sparse compositional or crystallographic map of the surface.

Claims

1. 1. A method comprising: removing a layer of the surface of the sample from which the image was acquired to expose a subsequent surface of the sample; acquiring an image of the subsequent surface; comparing the image of the subsequent surface with an image of a reference surface, the reference surface being the surface for which a compositional or crystallographic map has been obtained; obtaining a compositional or crystallographic map of the subsequent surface based on the comparison exceeding a threshold; and removing a layer of the subsequent surface to expose a next surface based on the comparison not exceeding the threshold.

2. acquiring an image of the next surface; comparing the image of the next surface with the image of the reference surface; obtaining a compositional or crystallographic map of the next surface based on the comparison exceeding a threshold; The method of claim 1 , further comprising: exposing additional surfaces based on the comparison not exceeding the threshold.

3. 10. The method of claim 1, further comprising repeating the steps of exposing, acquiring, and comparing a plurality of surfaces, such that a subset of surfaces of the plurality of surfaces have associated images and associated crystal or composition maps, and the remaining surfaces of the plurality of surfaces have associated images.

4. The method of claim 3 , further comprising propagating a crystal or composition map to the remaining surfaces of the plurality of surfaces using one or more neural networks.

5. propagating the crystal or composition map to the remaining surfaces of the plurality of surfaces using one or more neural networks; 5. The method of claim 4, comprising using the one or more neural networks to interpolate the crystal or composition maps from a subset of the surfaces of the plurality of surfaces to form the crystal or composition maps of the remaining surfaces of the plurality of surfaces.

6. propagating the crystal or composition map to the remaining surfaces of the plurality of surfaces using one or more neural networks; using the one or more neural networks to associate pixel values ​​of a crystal or composition map of each surface of the subset of surfaces with pixel values ​​of an image of each surface of the subset of surfaces; and using the one or more neural networks to assign pixel values ​​of a crystal or composition map to pixel values ​​of an image of one of the remaining surfaces based on associated pixel values.

7. The method of claim 1 , wherein comparing the image of the subsequent surface with the image of the reference surface is performed by a neural network.

8. The method of claim 1 , further comprising obtaining a sparse compositional or crystallographic map of the subsequent surface.

9. 1. An apparatus comprising: an electron column coupled to provide an electron beam; a focused ion beam column coupled to provide a focused ion beam; a plurality of detectors positioned to detect at least secondary electrons, x-rays, and backscattered electrons; a controller coupled to at least the electron column, the focused ion beam column, and each detector of the plurality of detectors, the controller including or coupled to code that, when executed by the controller, causes the apparatus to: using the focused ion beam to remove a layer of the surface of the sample from which the image was acquired to expose a subsequent surface of the sample; acquiring an image of the subsequent surface; comparing the image of the subsequent surface with an image of a reference surface, the reference surface being the surface for which a compositional or crystallographic map has been obtained; obtaining a compositional or crystallographic map of the subsequent surface based on the comparison exceeding a threshold; and removing a layer of the subsequent surface to expose a next surface based on the comparison not exceeding the threshold.

10. The controller further includes or is coupled to code that, when executed by the controller, causes the device to: acquiring an image of the next surface; comparing the image of the next surface with the image of the reference surface; obtaining a compositional or crystallographic map of the next surface based on the comparison exceeding a threshold; and exposing a subsequent surface based on the comparison not exceeding the threshold.

11. The controller further includes or is coupled to code that, when executed by the controller, causes the device to:

10. The apparatus of claim 9, wherein the apparatus is configured to repeat the steps of exposing, acquiring, and comparing a plurality of surfaces, wherein a subset of surfaces of the plurality of surfaces have associated images and associated crystal or composition maps, and the remaining surfaces of the plurality of surfaces have associated images.

12. The controller further includes or is coupled to code that, when executed by the controller, causes the device to:

12. The apparatus of claim 11, wherein one or more neural networks are used to propagate a crystal or composition map to the remaining surfaces of the plurality of surfaces.

13. The controller further includes or is coupled to code that, when executed by the controller, causes the device to:

13. The apparatus of claim 12, wherein the one or more neural networks are used to interpolate the crystal or composition maps from a subset of the surfaces of the plurality of surfaces to form the crystal or composition maps of the remaining surfaces of the plurality of surfaces.

14. The controller further includes or is coupled to code that, when executed by the controller, causes the device to: using the one or more neural networks to associate pixel values ​​of a crystal or composition map of each surface of the subset of surfaces with pixel values ​​of an image of each surface of the subset of surfaces; and assigning pixel values ​​of a crystal or composition map to pixel values ​​of an image of one of the remaining surfaces based on associated pixel values ​​using the one or more neural networks.

15. The controller further includes or is coupled to code that, when executed by the controller, causes the device to:

10. The apparatus of claim 9, further comprising obtaining a sparse compositional or crystallographic map of the subsequent surface.

Images