Sample preparation method and apparatus
The method of milling only the second surface of a sample to create a trench preserves the region of interest, enabling in-situ high-resolution characterization of polycrystalline thin films, addressing the limitations of existing sample preparation techniques by reducing damage and time.
Patent Information
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- OXFORD INSTR NANOTECHNOLOGY TOOLS LTD
- Filing Date
- 2021-08-26
- Publication Date
- 2026-05-27
AI Technical Summary
Existing sample preparation methods for material characterization, such as electron microscopy and spectroscopy, often result in ion beam-induced damage and are time-consuming, especially for polycrystalline thin films, and fail to preserve the sample surface for comprehensive characterization.
A method and apparatus that mill only the second surface of a sample to create a trench, preserving the region of interest and allowing in-situ characterization with a milling beam system, enabling electron-transparent samples for both structural and surface analysis without transferring the sample between instruments.
This approach minimizes sample damage, reduces preparation time, and allows for high-resolution nanoscale characterization, including TKD and EBSD, within the same instrument, facilitating complete characterization of polycrystalline thin films.
Smart Images

Figure 0007866548000001 
Figure 0007866548000002 
Figure 0007866548000003
Abstract
Description
Technical Field
[0001] The present invention relates to a method and apparatus for preparing a sample for analysis, and more particularly to a method and apparatus using a milling beam system for preparing a sample for further characterization.
Background Art
[0002] There are many material and device characterization techniques that require the preparation of a suitable sample to enable subsequent analysis. Such characterization techniques include, among many others, electron microscope-based analysis such as transmission and scanning electron microscopes (TEM / SEM), electron diffraction techniques such as selected area electron diffraction (SAED) and electron backscatter diffraction (EBSD), spectroscopy such as electron energy loss spectroscopy (EELS) and energy dispersive X-ray spectroscopy (EDS), and surface analysis techniques such as atomic force microscopy (AFM).
[0003] Such characterization techniques often impose requirements on the sample to be met during sample preparation in order to ensure the collection of high-quality data. When multiple characterization techniques are used in a correlation analysis, preparing a sample for characterization by one method may interfere with characterization by other methods. For example, preparing an electron-transparent sample may interfere with subsequent surface characterization, so it is necessary to ensure that these requirements are consistent.
[0004] Such an example of difficult sample preparation for correlation analysis is the characterization of polycrystalline thin films, where in this case, crystal structure information of the thin film is required along with surface-sensitive data such as topography. Generally, such samples must be analyzed by TEM to obtain structural information at a resolution of less than 100 nm. This requires the use of techniques such as selected area electron diffraction and diffraction contrast imaging to obtain nanoscale crystallographic information and prepare an electron-transparent sample for transfer to the TEM for characterization.
[0005] The preparation of site-specific electron-transmission samples is generally performed using FIB-SEM, where a trench is milled around the region of interest, and the thinned sample is lifted out and transferred to TEM. Several problems exist with these known sample preparation techniques. First, these techniques often result in ion beam-induced damage to the sample, hindering subsequent correlation characterization work. Furthermore, the lift-out technique is time-consuming, requires transfer between instruments, and has a significant failure rate. Moreover, simple FIB lift-out techniques cannot create planar samples with preserved surface regions to enable subsequent characterization. Therefore, these methods cannot provide complete characterization for many samples, such as polycrystalline thin films.
[0006] Therefore, there is a need for sample preparation methods that facilitate correlation property evaluation and bring about progress in overcoming some of the above-mentioned problems. In particular, there is a need for sample preparation methods that preserve the sample surface and enable correlation property evaluation of polycrystalline thin films. [Overview of the project]
[0007] A first aspect of the present invention provides a method for preparing a sample for analysis, comprising the steps of: providing a sample including a region of surface interest on a first surface of the sample and a second surface, wherein the second surface is oriented at an angle to the first surface around a common edge between the first and second surfaces, and the second surface extends between the common edge and a second edge of the sample opposite to the second surface; and milling the second surface of the sample to provide a trench on the surface of the second surface, wherein the trench extends from a first position on the second surface between the common edge and the second edge to a second position adjacent to the common edge, and the trench is arranged to provide an electron-transmissive sample layer including the region of surface interest.
[0008] By milling only the second surface of the sample, the region of interest on the first surface of the sample is completely preserved and remains undamaged by the milling beam. This allows for correlation characterization work that requires both an electron-transparent sample and a completely undamaged sample surface to obtain highly surface-sensitive data.
[0009] A further key advantage of the method according to the present invention is that the structural characterization of a sample can be performed in situ with a resolution of <10 nm within a milling beam instrument such as a FIB-SEM. In particular, transmission Kikuchi diffraction (TKD) can be performed on a sample prepared by this method within a FIB-SEM to provide crystallographic information with a resolution generally only possible with TEM. Since the region of interest on the sample surface remains completely undamaged and TKD is sensitive only to the crystallinity of the exit surface, the electron beam can be focused onto the milled back surface of the electron-transparent sample layer, and the transmitted scattered electrons can be collected using, for example, EBSD hardware and software. Thus, the method according to the present invention makes it possible to perform sample preparation and nanoscale characterization within the same instrument, eliminating the need to transfer samples between instruments.
[0010] This sample preparation method uses the surrounding bulk sample to support the electron-transmissive sample layer, eliminating the need for subsequent lift-out and attachment to the TEM grid required in the conventional FIB lift-out method. Even after subtracting the additional time required to transfer the sample to the TEM in the conventional method, this sample preparation method allows for the preparation of suitable electron-transmissive samples in a much shorter time frame.
[0011] The trench is positioned entirely within the second plane of the sample; that is, the trench extends from a position on the second plane of the sample toward the common edge. Specifically, the second plane of the sample extends from the common edge to the second edge opposite the second plane, and the trench extends from a position between the common edge and the second edge toward the common edge. By positioning the trench entirely within the second plane of the sample in this way, it is not necessary to mill through the entire thickness of the sample (i.e., from the second edge to a position adjacent to the region of interest of the sample). This significantly reduces the amount of milling required, minimizes the required preparation time, and limits the risk of damage to the region of interest of the sample, while allowing the region of interest to be characterized by tilting the sample to direct imaging and characterization beams at a certain angle to access the back surface of the region of interest through the trench.
[0012] Furthermore, since a portion of the sample remains intact near the second edge of the second surface, the milled sample remains in a more stable state. Additionally, this sample preparation technique can be applied to bulk samples where milling to the back surface of the region of interest across the entire thickness of the sample is not possible.
[0013] Preferably, the trench is an angled trench extending downward from a position on the second plane (between the common edge and the second edge) toward the common edge. As described above, this allows for the subsequent characterization of a sample of some thickness and significantly reduces the amount of milling required compared to known techniques in which the entire thickness of the sample must be removed to leave an electron-transparent region.
[0014] In some examples of this method, the sample may not have a single, clearly defined “common edge” between the first and second surfaces. The “second surface” of the sample is intended to include a portion of the sample that allows milling of the sample to approach the region of interest on the first surface from below the region of interest, while the region of interest remains unexposed to the milling beam. In particular, the method may include the steps of providing a sample containing the region of interest, and milling the sample to create trenches on the surface of the sample such that the trenches approach the back surface of the region of interest, thereby providing an electron-transparent sample layer containing the region of interest. In particular, the method may include the steps of orienting the sample with respect to a milling beam system such that the milling beam is parallel to the layer of interest, and milling toward the back surface of the region of interest.
[0015] Preferably, the trench on the second surface is positioned such that the electron-transparent sample layer is parallel to the first surface of the sample. The surface region of interest may lie on a substantially planar portion of the first surface, and the trench is positioned such that the electron-transparent sample region includes the planar portion of the first surface.
[0016] Preferably, the step of milling a second surface of the sample is carried out using a milling beam system, and the method further includes the step of oriented the sample so that the milling beam is parallel to the surface layer of interest. This makes it possible to provide a substantially uniform thickness of the electron-transmissive sample layer and to provide a planar electron-transmissive sample layer parallel to the first surface.
[0017] Preferably, the trench on the second surface includes an inclined trench having a deepest side adjacent to a common edge between the first and second surfaces. This minimizes the amount of material milled while providing an electron-permeable surface layer. The inclined trench may include a first internal surface parallel to the surface layer of interest, the bottom of which defines the deepest point of the trench, and an angled bottom surface that slopes upward from the bottom of the first internal surface and intersects with the surface of the second surface.
[0018] By milling an inclined trench with its deepest side adjacent to a common edge, there are no limitations on the sample thickness, and this method can be applied to bulk samples as well. In certain prior art methods, it is necessary to mill the sample from the opposite side (milling from the side opposite to the first surface of the sample to the region of interest on the surface). Therefore, it is necessary to mill and remove the entire thickness of the sample in order to leave the electron-transmitting region, which limits the usable sample thickness (generally up to about 50 μm). In contrast, by using an inclined trench on the surface of the second surface, samples of any thickness can be used, the amount of milling required is greatly reduced, and the region of interest on the surface can be analyzed by orienting the sample so that the electron beam is directed at a certain angle to the second surface of the sample. In particular, the electron beam can be oriented at the same or similar angle to the angle of the inclined surface of the trench relative to the first surface.
[0019] The angled bottom surface can be oriented at an angle of 30 to 85 degrees, preferably 65 to 70 degrees, relative to the first surface of the sample. This provides the most effective balance between allowing the electron beam to be oriented nearly perpendicular to the electron-transparent sample region and reducing the amount of milling required.
[0020] Milling is preferably performed using one or more of the following: a focused ion beam, a broad ion beam, and a laser.
[0021] Preferably, the step of milling a second surface of the sample includes performing a coarse milling step to provide a sample layer having a first thickness, and then performing a fine milling step to reduce the thickness of the sample layer to a second thickness less than the first thickness. In particular, the coarse milling step can be performed with a first beam current or power, and the fine milling step can be performed with a second, lower beam current or power. In this way, the majority of the material can be rapidly milled using the coarse milling step, and then the fine milling step is optimized to reduce damage to the electron-transparent sample layer.
[0022] This method may further include the steps of directing a focused electron beam through an electron-transparent sample layer and collecting the signal generated by the interaction between the focused electron beam and the electron-transparent sample layer using a detector. The signal may include one or more of scattered electrons, transmitted electrons, and X-rays.
[0023] In certain examples, the signal acquisition step includes collecting scattered electrons from an electron-transparent sample layer using an electron detector. The method may include directing a focused electron beam at an angle of 65 to 75 degrees to the plane of the electron-transparent sample layer. The method may include collecting scattered electrons using an electron imaging detector, such as an EBSD detector. The method may include collecting transmission Kikuchi diffraction (TKD) data. The method may include analyzing the collected electron image to identify one or more Kikuchi bands. The method may include using the identified one or more Kikuchi bands to determine the crystallographic orientation of a portion of the sample to which the focused electron beam is directed.
[0024] Preferably, the method includes the step of oriented the sample such that a focused electron beam is directed to a milled surface of the electron-transmissive sample layer and an electron detector is directed to a first surface of the sample opposite to the milled surface. Since transmission Kikuchi diffraction is sensitive only to the crystallinity of the exit surface and the region of interest on the surface is preserved, high-quality data can be obtained even if the quality of the milled surface is poor.
[0025] In some embodiments of the present invention, the step of collecting a signal includes collecting X-rays generated in an electron-transparent sample layer using an X-ray detector. The method may include the step of oriented the sample such that a focused electron beam is directed to a surface region of interest on a first surface of the sample, and the X-ray detector is directed to the surface region of interest on the first surface of the sample. In this way, X-ray spectroscopy can be performed on the surface region of interest.
[0026] In some embodiments of the present invention, the second surface of the sample is oriented at an angle with respect to the first surface, for example, perpendicular to the first surface. The first surface of the sample can include a polycrystalline surface layer. For example, the polycrystalline surface layer can have a thickness less than 100 nm and / or include a nanocrystalline structure having dimensions less than 100 nm. This method is particularly suitable for such samples because it enables the collection of crystallographic information from thin films that was not possible using conventional methods.
[0027] In some embodiments, the method can include depositing a protective layer on the second surface of the sample using electron beam evaporation prior to the milling step. In particular, the protective layer can extend over a region of the second surface of the sample adjacent to the common edge and preferably extend over the common edge to protect the sample during the milling step. The protective layer can include platinum.
[0028] The method can further include preparing the first surface of the sample prior to the milling step to produce a surface without deformation. In this way, standard EBSD analysis can be performed on the surface prior to the milling step, and high-resolution TKD analysis can be performed after the milling step.
[0029] In some embodiments of the present invention, the method can include machining the first surface of the sample to produce a polished surface over the region of interest on the surface before milling a trench in the second surface of the sample. In this way, a uniform thickness of the electron-transparent sample layer can be provided. The milling beam angle used for milling the first surface can be used for polishing the back surface of the electron-transparent sample layer in the trench to further ensure a uniform thickness. In these examples, preferably, the method includes depositing a protective layer on the first and / or second surface of the sample before milling the first surface of the sample.
[0030] Preferably, the method further includes analyzing a surface region of interest on the first surface of the sample using a characterization technique. The characterization technique can include one or more of atomic force microscopy, EBSD, SEM imaging, EDS, and cathodoluminescence spectroscopy. Since the current method leaves the first surface intact, these types of surface characterization techniques can be applied to the first surface of the sample either before or after milling.
[0031] In a further aspect of the invention, there is provided an apparatus for preparing a sample for analysis, the sample including a surface region of interest on a first surface of the sample and a second surface, the second surface being oriented at an angle to the first surface around a common edge between the first surface and the second surface, the second surface extending between the common edge and a second edge on the opposite side of the second surface of the sample, the apparatus including a milling beam system arranged to mill the second surface of the sample to provide a trench on the surface of the second surface, the trench extending from a first position on the second surface between the common edge and the second edge to a second position adjacent to the common edge, the trench being arranged to provide an electron-transparent sample layer including the surface region of interest, an electron beam system arranged to direct a focused electron beam through the electron-transparent sample layer, and a detector arranged to collect signals generated by interaction of the focused electron beam with the electron-transparent sample layer.
[0032] Thus, the apparatus enables the evaluation of the structural characteristics of the sample with a resolution of less than 10 nm when both sample preparation and characterization are performed within the apparatus. In particular, sample preparation and subsequent transmission Kikuchi diffraction (TKD) can be performed using the apparatus, providing crystallographic information at a resolution generally achievable only with TEM. Considering that the surface region of interest on the sample remains completely undamaged and that TKD is sensitive only to the crystallinity of the exit surface, the electron beam can be focused on the milled back surface of the electron-transparent sample layer and transmitted scattered electrons can be collected, for example, using EBSD hardware and software.
[0033] The apparatus preferably includes an electron detector configured to collect scattered electrons from an electron-transparent sample layer to provide an electron intensity image. The apparatus is preferably configured to allow the sample to be oriented such that a focused electron beam is directed towards a milled surface of the electron-transparent surface layer, and the electron detector is directed towards a first surface of the sample opposite to the milled surface.
[0034] The apparatus further preferably includes a processing unit configured to analyze an electron intensity image to identify one or more Kikuchi bands in the signal. In particular, the processing unit may be configured to analyze the collected signal to identify one or more Kikuchi bands and preferably to perform a computer-implemented method for determining the crystallographic orientation of the sample or a portion of the sample. The processing unit may further be configured to determine the distance of the Kikuchi bands from the pattern center and to mitigate the effect of gnomonic strain on the shape of the Kikuchi bands based on the determined distance. More specifically, the processing unit may be configured to calculate a range for a candidate material that includes the maximum and minimum projected bandwidths of the Kikuchi bands and to reject identified Kikuchi bands if the measured bandwidth is outside the calculated range for the candidate material. The processing unit may be configured to calculate a range that includes the maximum and minimum projected bandwidths for the Kikuchi bands based on a plurality of candidate phases stored in memory, the candidate phases may be selected by the user or automatically selected by the processing unit.
[0035] In some embodiments, the apparatus may include an X-ray detector configured to collect X-rays emitted from an electron-transparent surface layer and provide an X-ray energy spectrum. In this way, high-resolution X-ray spectroscopic data can be collected from an electron-transparent sample layer. In particular, the apparatus may be configured to allow sample orientation such that a focused electron beam is directed to a surface region of interest on a first surface of the sample, and the X-ray detector is directed to the same surface region of interest on the first surface of the sample.
[0036] In some embodiments of the present invention, the apparatus comprises an electron detector configured to collect scattered electrons from an electron-transparent sample layer and provide an electron intensity image; an X-ray detector configured to collect X-rays emitted from an electron-transparent surface layer and provide an X-ray energy spectrum; and a sample holder configured to move the sample relative to the electron beam system, electron imaging detector, and X-ray detector between a first orientation in which the focused electron beam is directed towards the milled surface of the electron-transparent surface layer and the electron detector is directed towards a first surface of the sample opposite to the milled surface, and a second orientation in which the focused electron beam and the X-ray detector are directed towards a region of interest on the first surface of the sample. Thus, using this apparatus, it is possible to sequentially mill an electron-transparent sample while preserving the region of interest on the surface, collect TKD data with the electron detector, and collect high-resolution EDS data with the X-ray detector.
[0037] This apparatus may further include a gas injection system, and the apparatus may be configured to perform electron beam-assisted deposition using an electron beam and a gas injection system.
[0038] The apparatus further comprises a memory for holding computer-readable instructions, and when an instruction is executed, the apparatus can be made to perform method steps as defined in the appended claims or as defined in the above or below aspects of the present invention. In a further aspect of the present invention, a computer implementation method for processing an electron image including a Kikuchi diffraction pattern is provided, the method comprising the steps of: identifying a Kikuchi band in the electron image; determining the width of the identified Kikuchi band; calculating a minimum and / or maximum expected Kikuchi bandwidth based on the lattice spacing of one or more candidate phases; and ignoring the identified Kikuchi band if the determined width is less than and / or greater than the calculated minimum expected Kikuchi bandwidth.
[0039] A computer-implemented method for processing an electron image including a Kikuchi diffraction pattern can form part of the method defined above for preparing a sample for analysis. The apparatus defined above for preparing a sample for analysis may include a processing unit configured to perform the computer-implemented method for processing an electron image including a Kikuchi diffraction pattern.
[0040] By ignoring features outside the expected range based on a list of one or more candidate phases, this method reduces the misidentification of image features as Kikuchi bands and enables more accurate indexing of crystal orientation. The maximum and / or minimum lattice spacings of one or more candidate phases can be stored in memory. Candidate phases can be entered by the user in the user interface by selecting them from a list or by user input. In other cases, candidate phases can be determined automatically, for example, using spectroscopy.
[0041] The method preferably includes the steps of calculating minimum and maximum Kikuchi bandwidths and ignoring identified Kikuchi bands if the determined bandwidths fall outside the range defined by the minimum and maximum Kikuchi bandwidths. In this way, detected features are ignored if they are either too wide or too narrow to be consistent with the expected phases of the sample. In some embodiments, the maximum Kikuchi bandwidth can be used to optimize the identification of band edges. In particular, if a Kikuchi band is identified in an electron image, the maximum expected Kikuchi bandwidth (based on the minimum lattice spacing of one or more candidate phases) can be used to determine the distance from the center of the Kikuchi band to be searched to find the Kikuchi band edge.
[0042] Preferably, the step of calculating the minimum and / or maximum Kikuchi band widths additionally takes into account the electron beam acceleration voltage and / or the position of the identified Kikuchi beam relative to the pattern center. In particular, the step of determining the width of the identified Kikuchi bands may include measuring the distance of the identified Kikuchi bands to the pattern center and applying a correction to the determined width of the Kikuchi bands to mitigate the effects of gnomonic distortion based on the measured distance. Alternatively, the method may include measuring the distance of the identified Kikuchi bands to the pattern center and applying a correction to the expected maximum and / or minimum Kikuchi band widths based on one or more candidate phases. The pattern center preferably corresponds to the portion of the detector closest to the source of the Kikuchi pattern on the sample.
[0043] The method may also include a step of using the current accelerating voltage of the electron beam to calculate the minimum and / or maximum Kikuchi bandwidth based on the lattice spacing of one or more candidate phases.
[0044] The step of identifying Kikuchi bands in an electron image may include applying a Hough transform to the electron image to provide a Hough image in which the Kikuchi bands in the electron image are represented by intensity peaks, and identifying the intensity peaks in the Hough image as corresponding to Kikuchi bands. The Hough image may further be processed with a butterfly filter, and the intensity peaks are identified in the filtered Hough image. Subsequent processing may be performed on the unfiltered image. The step of determining the width of the identified Kikuchi bands includes identifying the point of the steepest intensity gradient on the opposing edges of the intensity peaks in the Hough image as an edge position, and determining the width of the Kikuchi bands based on the distance between the edge positions.
[0045] Preferably, the step of identifying the steepest intensity gradient includes searching from a peak position far from the intensity peak to a predetermined maximum distance, where the predetermined maximum distance is based on the maximum grid spacing of one or more candidate phases. In particular, the method may include the step of calculating the corresponding maximum distance in the Huff space image using the maximum grid spacing of one or more candidate phases. The search of the Huff space image for the edge position is preferably performed at subpixel resolution within the Huff image.
[0046] The Kikuchi band center is preferably located at an intermediate position between the Kikuchi band edge positions. The method preferably further includes the steps of identifying a plurality of Kikuchi bands in an electron image and determining the crystallographic orientation based on the relative positions and / or orientations of the Kikuchi bands. [Brief explanation of the drawing]
[0047] [Figure 1A] This figure schematically illustrates a method for preparing a sample for analysis according to the present invention. [Figure 1B] This figure schematically illustrates a method for preparing a sample for analysis according to the present invention. [Figure 1C] This figure schematically illustrates a method for preparing a sample for analysis according to the present invention. [Figure 2] This figure illustrates a method for collecting scattered electron intensity signals from a sample prepared according to the method of the present invention. [Figure 3A] This figure shows a transmitted Kikuchi diffraction pattern collected from a sample prepared according to the present invention. [Figure 3B] This figure shows a diffraction pattern quality map obtained by scanning an electron beam across a sample and collecting transmission Kikuchi diffraction data from the sample region. [Figure 3C] This figure shows a method for processing an electron image including a Kikuchi diffraction pattern according to the present invention. [Figure 4A] This figure shows a method for processing an electron image including a Kikuchi diffraction pattern according to the present invention. [Figure 4B]This figure shows a method for processing an electron image including a Kikuchi diffraction pattern according to the present invention. [Figure 4C] This figure shows a method for processing an electron image including a Kikuchi diffraction pattern according to the present invention. [Figure 5] This figure shows a method for acquiring a high-resolution EDS signal from a sample prepared according to the method of the present invention. [Figure 6A] This figure shows a series of sample preparation and analysis steps of the method according to the present invention. [Figure 6B] This figure shows a series of sample preparation and analysis steps of the method according to the present invention. [Figure 6C] This figure shows a series of sample preparation and analysis steps of the method according to the present invention. [Figure 6D] This figure shows a series of sample preparation and analysis steps of the method according to the present invention. [Figure 6E] This figure shows a series of sample preparation and analysis steps of the method according to the present invention. [Figure 7] This figure shows an optional additional sample preparation step according to an embodiment of the present invention. [Modes for carrying out the invention]
[0048] Figures 1A to 1C schematically illustrate a method for preparing a sample 10 for analysis according to the present invention. The method includes the step of providing a sample 10 as shown in Figure 1A, the sample 10 having a region of surface interest 11 on a first surface 12 of the sample 10 and a second surface 13, the second surface being oriented at an angle a with respect to the first surface 12 around a common edge 14 between the first surface 12 and the second surface 13. The method further includes the step of milling the second surface 13 of the sample 10 to provide an inclined trench 20 on the surface of the second surface 13 adjacent to the common edge 14, as shown in Figure 1B, the trench 20 being positioned to provide an electron-transparent sample layer 15 containing the region of surface interest 11.
[0049] According to this method, since milling is performed only on the second surface 13 so as to approach the back surface of the region of interest 11, an electron-transparent sample including the region of interest 11 can be prepared while preserving the surface of the region of interest 11. In this way, this sample preparation method enables correlation characterization work using characterization methods that require both an electron-transparent sample layer and preservation of the sample surface. Furthermore, the sample preparation step and the sample characterization step can be performed on-site in the same instrument, significantly improving the speed at which data can be acquired. In contrast to prior art methods, the present invention provides a trench that is entirely located within the second surface 13 of the sample rather than extending through the entire thickness of the sample, which significantly reduces the amount of milling required to prepare the electron-transparent sample layer 15. This means that this technique can be applied to samples of any thickness (within the limitations determined by the practicality of sample mounting in the instrument's sample chamber).
[0050] As shown in Figure 1A, the sample in this example has a substantially cubic shape in which the second surface 13 is substantially perpendicular to the first surface 12 around a common edge 14. However, the sample 10 can have any shape, as long as it allows milling from the second surface 13 of the sample, which is at an angle to the first surface, so that milling can be performed from the back surface of the sample region of interest 11. Any suitable position of the sample region of interest 11 can be selected, as long as it is positioned toward the intervening edge 14 between the first surface 12 and the second surface 13 so that milling can be performed toward the back surface of the region of interest 11 from the second surface 13.
[0051] As shown in Figure 1B, the milling can be carried out by any suitable technique that allows the trench on the second surface 13 to be milled to the back surface of the surface region of interest 11, so as to provide an electron-permeable surface layer 15 on which the surface layer of interest 11 is located. The trench 20 on the second surface 13 of the sample 10 is preferably an inclined trench whose deepest side is adjacent to the first surface 12. As shown in Figure 1C, the inclined trench preferably has a first internal side surface 21 that is parallel to the surface layer of interest 11, and is positioned so that the bottom of the first side surface 21 defines the deepest point of the trench 20. The trench 20 is positioned to have an angled bottom surface 22 that is inclined upward away from the bottom of the first internal side surface 21 and conforms to the surface of the second surface 13.
[0052] As described below, this arrangement enables subsequent characterization techniques, including oriented the electron beam 41 on the first internal side surface 21 of the inclined trench 20 and focusing the electron beam 41 onto the electron-transparent layer 15 from the opposite side of the surface region of interest 11. Thus, the angle of the bottom surface 22 of the inclined trench determines the angle at which the electron beam 41 can be oriented onto the milled first internal side surface 21 of the trench 20. To provide maximum flexibility in the orientation of the electron beam 41, the angled bottom surface 22 is preferably between 30 and 85 degrees with respect to the surface of interest, more preferably about 65 to 70 degrees, which provides a good balance between the need to mill and remove a considerable amount of material and also providing an electron beam alignment that is nearly perpendicular to the sample region of interest 11. As shown in Figures 1B and 1C, the trench is preferably arranged such that the electron-transparent sample layer 15 is a layer parallel to the first surface 12. That is, the electron-transparent surface layer 15 is in the plane of the first surface 12 of the sample 10.
[0053] The milling step shown in Figure 1B can be performed using any known milling beam system, such as a focused ion beam, a broad ion beam, or a laser. Preferably, as shown in Figure 1A, the milling step is performed in a dual-beam instrument having a focused ion beam system 30 and an electron beam system 40 to enable simultaneous imaging and milling of the sample 10, similar to conventional FIB-SEM liftout techniques. Specific examples of methods using FIB-SEM are as follows.
[0054] First, the region of interest 11 used for sample preparation is identified using plan view SEM imaging, as shown in Figure 1A, and the sample 10 is oriented in the chamber of a dual-beam FIB-SEM so that the electron beam 41 is nearly perpendicular to the first surface 12 of the sample 10. This step may not be important in the case of a homogeneous thin film on the first surface 12 of the sample 10 being investigated, but if the region of interest is close enough to the edge 14 of the sample 10 and milling is possible from the second surface 13 of the sample to the back surface of the region of interest 11, this step can be used to identify a more specific local region of interest.
[0055] As shown in Figure 1B, the sample 10 is then rotated so that it can be FIB milled on an adjacent second surface 13 of the sample. Preferably, as shown in Figure 1B, the sample 10 is rotated so that the ion beam is substantially parallel to the first surface 12 of the sample 10. If the conventional orientation of the FIB 30 relative to the electron beam system 40 is about 50-60 degrees, this step therefore generally involves a sample tilt of about 140-150 degrees to ensure that the FIB beam system 30 is correctly oriented with respect to the region of interest 11 on the first sample surface 12. Once the correct orientation for milling is achieved as shown in Figure 1B, the inclined trench can be milled using the FIB, similar to conventional FIB sample liftout. In particular, the trench 20 on the second surface can be milled with a sample edge 14 of up to 1-2 micrometers using the initial rough milling step. The rough milling step can be performed by increasing the beam current so that the material can be removed from the second sample surface 13 at a faster rate. Next, the beam current can be reduced to perform a fine milling step in which the sample layer containing the region of interest 11 is milled from a thickness of 1-2 micrometers to electron-transparent. This includes a step of milling the sample layer 15 to a thickness of less than 100 nanometers to ensure electron-transparentness for further characterization.
[0056] As shown in Figure 1C, the final prepared sample has an electron-transparent sample region 15 with a thickness of less than 100 nanometers, which is supported by the surrounding bulk sample. Unlike conventional FIB liftout, no further milling steps are required to free the electron-transparent region 15 from the bulk sample 10 and remove it to the sample grid. Rather, as shown in Figure 1C, further characterization can be performed with the electron-transparent region 15 supported by the surrounding sample 10. Since the refined electron-transparent sample region 15 remains supported by the bulk sample 10, the number of milling steps is reduced compared to the steps required to attach the sample to a micromanipulator probe, free it from the bulk sample 10, and attach it to the sample grid for further analysis. This results in a significant reduction in the failure rate of sample preparation and a reduction in damage to the sample by the first ion beam.
[0057] In the sample preparation stage shown in Figure 1C, the sample 10 can then be transferred to further instruments for additional characterization. In particular, the prepared sample 10 is electron-transparent but also retains the surface 12 of the region of interest 11, enabling further correlated surface and structural characterization techniques such as electron imaging and diffraction in different SEMs, along with surface morphology analysis that requires the surface to be intact, such as AFM. Furthermore, the milled surface 21 of the electron-transparent sample layer can be illuminated with the electron beam 41 shown in Figure 1B, enabling techniques that do not depend on the incident surface 21 being intact but require the exit surface to be preserved in order to obtain high-quality data.
[0058] In a preferred embodiment of the present invention, the sample preparation steps are performed within an instrument further including hardware that enables these steps to be carried out in-situ, eliminating the need to transfer the sample to a second instrument for characterization. Examples of characterization techniques that benefit from the sample preparation method of the present invention will be described with reference to Figures 2 to 4.
[0059] (Transmission Kikuchi diffraction analysis) Figure 2 shows the setup for obtaining transmission Kikuchi diffraction (TKD) data from a sample prepared according to the present invention. TKD is particularly interesting because, when combined with the sample preparation method of the present invention, it offers the possibility of obtaining nanoscale crystallographic information (down to less than 10 nm) from a sample in a scanning electron microscope (SEM), which is normally only possible in a TEM.
[0060] TKD data can be obtained using conventional electron backscatter diffraction (EBSD) hardware within a scanning electron microscope (SEM). While EBSD can provide rich information, particularly regarding crystalline structures such as polycrystalline films, conventional EBSD has resolution limitations and is not suitable for analyzing nanocrystalline structures (dimensions less than 100 nanometers). Typically, a TEM is required to analyze such nanocrystalline samples. However, using a TEM requires complex and time-consuming sample preparation and the transfer of the prepared sample to the TEM.
[0061] TKD uses the same hardware and software as EBSD to collect intensity images of electrons scattered through an electron-transmissive sample. TKD enables the analysis of thin film structures, such as nanocrystalline films with a thickness of 5 to 100 nanometers, which was not possible with conventional EBSD. Since TKD is sensitive only to the crystallinity of the exit surface, the sample preparation technique according to the present invention is ideally suited because it can completely preserve the surface region of interest 11. By orienting the sample 10 as shown in Figure 2, the milled back surface 21 of the sample region of interest 15 is oriented toward the electron beam 41, and transmitted and scattered electrons 42 are collected from the preserved surface region of interest.
[0062] Therefore, combining this sample preparation method with TKD characterization offers advantages over TEM.
[0063] TEM imaging and diffraction project the entire thickness of the sample, thus placing far higher demands on the quality of both the inlet and outlet surfaces. In contrast, with TKD, the inlet surface is not critical and may even be made of a different material. Therefore, compared to TEM sample preparation, which requires meticulous attention during sample preparation and necessitates low-current / low-energy ion polishing to ensure crystallinity at both the inlet and outlet surfaces, the milling process can be performed much more quickly.
[0064] Figure 2 shows the instrumentation and orientation of the sample 10, including the hardware for TKD analysis. In particular, the instrumentation includes an EBSD detector 43, i.e., an electron intensity imaging detector such as a CCD, CMOS, or direct electron detection camera. As shown in Figure 2, the sample 10 is oriented with respect to the electron beam system 40 such that the focused electron beam 41 is oriented through the electron-transparent surface layer 15 and scattered electrons 42 appearing from the first surface 12 of the sample 10 are collected by the electron imaging detector 43.
[0065] Since the electron beam 41 is oriented to the milled back surface 21 of the electron-transparent surface layer, and considering that TKD only needs to remain undamaged at the exit surface, the reduction in crystallinity due to ion implantation will not affect the quality of the data obtained. The irradiation angle of the focused electron beam 41 is limited by the geometry of the trench 20; shallower inclined trenches allow for irradiation angles closer to perpendicular to the electron-transparent sample 15, but require more milling and more material to be removed. Generally, a preferred compromise that allows for the collection of a strong signal at the detector 43 without requiring excessive milling is to provide an inclined trench with an angled bottom surface at an angle of 65 to 75 degrees with respect to the first surface 12 of the sample 10. This allows for a corresponding angle of the focused electron beam of 65 to 75 degrees with respect to the plane of the electron-transparent sample 15. As shown in the figure, scattered electrons 42 are collected by the imaging detector 43 to collect a transmission Kikuchi diffraction (TKD) pattern, as shown in Figure 3A.
[0066] At each incident point of the electron beam from the milled surface 21 of the sample 15, a TKD pattern 50 can be collected, as shown in Figure 3A. The TKD pattern contains multiple Kikuchi bands 51 that indicate orientation relative to the region of the sample irradiated by the electron beam 41 and provide crystallographic information. By scanning the electron beam 41 across the surface of the electron-transparent sample layer 15 shown in Figure 2, a map of the orientation of the crystal grains forming the polycrystalline film can be formed, as shown in Figure 3B. As described above, crystal grain imaging at this resolution is generally only possible in a TEM, but the sample preparation technique of the present invention makes it possible to prepare and characterize the sample in the same FIB SEM instrument.
[0067] (Processing of TKD data) The apparatus preferably further includes software for processing TKD data collected from the TKD analysis process illustrated in Figure 2. The apparatus may further include a processing unit containing software for performing a number of computer implementation method steps. In particular, the software may be configured to perform a method step that provides detection of the Kikuchi band 51 in order to determine the orientation of the crystal grains on which the beam is incident. The software may further include features that provide improved accuracy in identifying the Kikuchi band than known automated techniques. This process, called "Primary Band Detection (PBD)" or "TKD Optimized Band Detection," is performed within the software as follows:
[0068] Once one or more electron intensity images of the TKD pattern 50 are obtained using the procedure illustrated in Figure 2, the TKD pattern 50 is first processed using the Hough transform, as is well known from existing Kikuchi diffraction pattern processing techniques. As shown in the Hough image in Figure 3C, the Kikuchi band in the original TKD pattern corresponds to peak 53. As shown in the enlarged portion of Figure 3C, the peak maximum value is identified around one of the peaks 53 in the Hough image, and two minimum values, p-min1 and p-min2, are identified on either side of peak 53. Between the two minimums, the steepest intensity gradient is identified as edges p-edge1 and p-edge2, corresponding to the edges of the original Kikuchi band 51. The distance between the two edges p-edge1 and p-edge2 determines the bandwidth of the original Kikuchi band 51. The midpoint between the edges is determined as the center of the Kikuchi band. By identifying the position of the Kikuchi band in the image in this way, the Kikuchi pattern 50 can be indexed and the crystal orientation can be determined. By scanning the beam across the sample surface, an orientation map 56 can be formed, as shown in Figure 3B, providing nanoscale crystallographic information.
[0069] This method may have the problem of misidentifying features in the image as Kikuchi bands. This can hinder accurate orientation indexing. This method can address these issues by further optimizing the automatic identification of Kikuchi bands 51 through several additional steps.
[0070] First, the algorithm can enable the selection of one or more expected phases (i.e., material and corresponding crystal structures) present in the sample. These phases can be automatically identified, for example, by spectroscopic analysis within the instrument, or they can be entered by the user. The algorithm can determine the maximum and minimum bandwidths of the Kikuchi bands 51 in the pattern using the maximum and minimum lattice spacings of the phases in the expected phase list, so that the software can automatically discard features identified as Kikuchi bands if they do not match the expected crystallographic phases. The maximum lattice spacing present in the sample determines the minimum bandwidth of the Kikuchi bands 51 present in the TKD pattern, and similarly, the minimum lattice spacing present in the material of the sample determines the maximum bandwidth of the Kikuchi bands expected in the Kikuchi diffraction pattern 50.
[0071] The software can also consider several other parameters to more accurately identify the Kikuchi band 51. In particular, the algorithm takes into account the position of the Kikuchi band 51 of interest relative to the pattern center (i.e., the part of the detector 43 closest to the sample 15). The relative position of the Kikuchi band to the pattern center results in gnomonic distortion of the Kikuchi diffraction pattern 50. By taking into account the distance of the Kikuchi band 51 measured relative to the pattern center, the algorithm can automatically mitigate the effect of gnomonic distortion on the shape of the diffraction pattern 50. The algorithm can further be configured to take into account the acceleration voltage of the electron beam 41 used to form the pattern 50. A higher acceleration voltage reduces the electron wavelength, which also reduces the width of the Kikuchi band 51 in the diffraction pattern 50.
[0072] By calculating the minimum bandwidth predicted for a particular Kikuchi band 51, taking into account the maximum lattice spacing of the phases in the expected phase list, the position of the Kikuchi band 51 relative to the pattern center, and the acceleration voltage of the electron beam, the lower limit of the Kikuchi bandwidth can be accurately predicted. This reduces the likelihood of misidentifying narrow features in the electron image 50 as Kikuchi bands, which degrades the accuracy of orientation measurements. A method for optimizing this Kikuchi band detection is shown in Figures 4A to 4C.
[0073] As described above, the TKD pattern 50 is obtained from the sample and detector orientation illustrated in Figure 2. Conventional Kikuchi band detection methods often mistakenly identify contrast variations at the edges of broad Kikuchi bands as individual Kikuchi bands themselves. As shown in Figure 4B, conventional Kikuchi band detection methods detect narrow contrast variations 54 as Kikuchi bands. Therefore, when indexing the TKD pattern, these contrast variations are incorrectly considered, making it impossible to accurately determine the crystal orientation. Using the above-described primary band detection method to predict the minimum width of the Kikuchi band 51, the narrow contrast variations 54 shown in Figure 4B are ignored based on the calculated minimum predicted width. In this way, broad Kikuchi bands are correctly identified along the centers of these Kikuchi bands 55, as shown in Figure 4C.
[0074] These method steps for identifying Kikuchi bands within the collected TKD pattern significantly improve the accuracy of band detection and subsequent crystal orientation measurements. These methods have the greatest impact when working with typical TKD geometry shown in Figure 2, where the TKD pattern center can be located above the upper edge of the diffraction pattern. Thus, the method steps described above, which can be implemented in software running on a computer as part of the instrument, can be synergistically combined with the sample preparation and analysis steps described in Figure 2 to obtain high-quality nanoscale structural information from samples in the SEM. The software steps can be applied to data obtained from conventionally prepared samples, but even greater advantages are obtained when applied in conjunction with the method steps of the present invention described above.
[0075] The software can also incorporate knowledge of the minimum grid spacing (and therefore the maximum Kikuchi bandwidth). In particular, knowledge of the minimum grid spacing from the expected phase list, along with the position of the Kikuchi band 51 relative to the pattern center and the energy of the electron beam (acceleration voltage), can be used to predict the maximum Kikuchi bandwidth and, consequently, determine where to begin searching for the band edge, as shown in the enlarged portion of Figure 3C. Knowledge of the minimum grid spacing allows the system to know how far it is from the first detected band center, to search for the band edge in the Huff space image, and to provide an upper limit on the detected bandwidth. This allows for more accurate searching of the band edges p-edge1 and p-edge2 around peak 53 in the Huff image, for example, at sub-pixel resolution, and more accurately identification of the band edge and the band center assigned as the midpoint between the two precise band edge positions.
[0076] By combining a sample preparation method that allows for image acquisition by simply reorienting the sample into a geometric shape as shown in Figure 2 after saving a surface region of interest 11 on the surface, with the above-described method steps for analyzing the obtained TKD pattern, it is possible to obtain structural data from a sample in SEM with a resolution far higher than what was previously possible using known methods.
[0077] (X-ray spectroscopy characteristics evaluation) The sample preparation method according to the present invention also allows for further characterization techniques to be performed on the sample 10. Figure 5 shows the geometric shape of an electron-transparent sample 15 after it has been prepared using the method described above, in which case high-resolution X-ray spectroscopy can be performed on the region of interest 11 on the first surface 12 of the sample before the milling step of this method. In particular, the interaction volume between the electron beam 41 and the sample 10 is limited in data resolution due to the generation of X-rays 45 from a relatively large volume of the sample. Therefore, the data resolution is limited in conventional EDS on the sample surface and cannot be used to collect high-resolution data of polycrystalline thin films where the dimensions of specific crystals may be less than 100 nanometers unless the electron beam energy (acceleration voltage) is significantly reduced. As shown in Figure 5, by oriented the sample relative to the electron beam system 40 and the X-ray detector 44, high-resolution EDS data can be collected from electron-transparent thin films without the need for lift-out from or transfer to the TEM. Instead, as shown in Figure 1B, once the milling step is performed, the sample can be tilted back to the orientation shown in Figure 5, with the electron beam 41 perpendicular to the first surface 12 of the sample 10 containing the region of interest 11.
[0078] Depending on the geometry of the sample, the focused electron beam 41 can be focused to a small spot size to provide high-resolution X-ray data, causing X-rays 45 to fluoresce from an extremely small area of the electron-transparent sample 15. As shown in Figure 3, by positioning the X-ray detector 44 facing the first surface 12 of the sample, the X-ray detector 44 collects only X-rays from the electron-transparent sample layer 15 and does not collect X-rays generated on the underlying bulk sample 10 as the electron beam passes through the electron-transparent surface layer. Therefore, by facing both the electron beam 41 and the detector 44 to the first surface 12 and the surface layer of interest 11 of the sample, high-resolution X-ray data can be obtained. With beam voltages up to 30kV available in the SEM, this type of high-resolution EDS analysis can be performed within the same SEM used for sample preparation, eliminating the need for transfer to a TEM, which is usually required for this type of analysis. Similar to conventional techniques, an X-ray spectroscopic image can be obtained by scanning the surface of the electron-transparent sample layer 15 with a focused electron beam 41, and correlation analysis can be performed with TKD data obtained from the orientation shown in Figure 2 and any other data obtained from the sample surface before milling, as shown in Figure 1A.
[0079] (Multi-stage sample preparation and analysis sequencing) As described above, one of the main advantages of the sample preparation and analysis techniques according to the present invention is that a sample can be prepared and multiple characterization techniques can be performed all within the same apparatus, such as a FIB-SEM equipped with appropriate hardware for detecting relevant signals. Figures 6A to 6E show some exemplary steps in the process of characterizing a sample, all of which can be performed within the same apparatus. The apparatus for preparing and analyzing a sample includes a milling beam system 40 suitable for milling a second surface of the sample 13 to create a trench adjacent to the first surface 12, so that an electron-transparent sample layer 15 can be prepared. The apparatus further includes an electron beam system 40 for imaging and characterization, and one or more detectors 43, 44 for collecting various signals. In the specific example in Figure 6, the apparatus includes a FIB system 30 for generating a focused ion beam for milling the sample, as shown in Figure 6C, an electron imaging detector (i.e., an EBSD detector) 43 for collecting scattered electrons, and an X-ray detector 44 for collecting fluorescent X-rays from the sample. Figures 6A to 6E illustrate the various steps in the characterization process, all of which can be easily obtained by simply aligning the sample 10 correctly with respect to the electron beam system 40, the milling beam system 30, and the detectors 43 and 44.
[0080] First, as shown in Figure 6A, data can be collected from a region of interest 11 on the first surface 12 of the sample 10 prior to the milling step. For example, the sample can be oriented so that the first surface 12 is perpendicular to the electron beam 41 in order to enable various characterization techniques performed on the sample surface, including secondary electron and backscattered electron imaging, cathode luminescence, and EDS. Next, the sample can be tilted relative to the electron beam for conventional EBSD, as shown in Figure 6B. In particular, the first surface of the sample can be tilted through about 70 degrees to provide an optimal geometry for conventional EBSD, and the resulting scattered electron signal is collected by the EBSD detector 43.
[0081] Figure 6B also shows optional additional steps that may be included in the method before performing the milling step. In particular, the apparatus may further include a gas injection system, and the method may include a step of depositing a protective layer 19 using an electron beam with the gas injection system and electron beam-assisted deposition. The protective layer 19 can serve to protect the first surface 12 of the sample from unintended ion beam damage during milling and can also improve the uniformity of the milled surface of the electron-transparent sample layer 15. The protective layer can be deposited so as to extend over a portion of the second surface 13 of the sample adjacent to the common edge 14, preferably extending to cover the common edge 14. The protective layer 19 is not shown in the subsequent Figures 6C-6E, but obviously, if this optional step is included, a portion of the protective layer will remain at the edge of the electron-transparent region 15.
[0082] After these conventional imaging and characterization techniques of the sample surface have been performed, the sample preparation steps described above can be carried out with reference to Figure 1, as shown in Figure 6C. In particular, the sample 10 can be tilted 140 to 150 degrees with respect to the planar orientation of Figure 6A such that the focused ion beam 31 is substantially parallel to the first surface 12 of the sample 10. The milling step can proceed as described above to create a trench containing a region of interest 11 and an electron-transparent sample layer on the first surface 12 of the sample 10. Next, the sample can be tilted to about 150 to 160 degrees with respect to the planar orientation of Figure 6A so that the focused electron beam can be directed to the milled surface of the electron-transparent sample layer 15, and electrons scattered through the sample layer 15 are collected by the electron imaging detector 43 in order to perform TKD analysis of the thinned region 15. Finally, as explained with reference to Figure 5, the sample can be returned to a planar orientation as shown in Figure 6E so that high-resolution EDS can be performed by scanning the focused electron beam across the electron-transmissive sample region 15 in order to collect high-resolution EDS data.
[0083] To provide each of the desired orientations shown in Figure 6, the apparatus preferably includes a sample holder that allows the sample to be tilted between each of the orientations shown in Figures 6A to 6E, i.e., allows for tilting from 0 to 160 degrees.
[0084] The method and apparatus according to the present invention may include one or more further features.
[0085] Figure 7 shows an optional additional method step that can be used in a particular embodiment of the present invention. This step includes milling a region on the first surface 12 of the sample 10 with a milling beam system 30 to create a region of surface interest for one or more of the subsequent analysis steps described above. In particular, the method may include the step of milling the first surface 12 of the sample 10 with a milling beam system 30 to create a trench 16 in the first surface 12, thereby providing a flat polished surface within the region of surface interest 11. This provides a uniform sample region for analysis, especially when the region of surface interest is not sufficiently created or when some topography is retained from the initial sample polishing. It also minimizes problems associated with rounding the sample edges.
[0086] Another significant advantage is that it facilitates the preparation of a final sample with uniform thickness. In particular, since the milling beam angle for this initial milling of the trench 16 on the first surface 12 is known, the milling beam angle for subsequent milling of the inclined trench 20 and thinning of the electron-transparent sample layer 15 can be selected more precisely to provide uniform thickness. By selecting the same milling beam inclination angle for the initial milling of the trench 16 on the first surface of the sample and for subsequent thinning of the electron-transparent region 15 in the inclined trench 20, the plane of the electron-transparent sample layer 15 becomes parallel. This provides a high-quality sample for subsequent analysis (such as TKD) and yields improved results.
[0087] This top milling step in Figure 7 is preferably performed before milling the inclined trenches 20 on the second surface 13 of the sample 10. Referring to Figure 1, this step is therefore performed before the milling step in Figure 1B to polish the first surface of the sample 12 and prepare the sample region of interest 11 before milling the inclined trenches 20. Preferably, when this pre-preparation step is used, a protective layer 19 (typically Pt) is first deposited on the second surface of the sample to ensure a better quality milled surface on the first surface of the sample 12.
[0088] Therefore, the preferred sequence of the sample preparation steps is as follows: First, a protective layer is deposited on the second surface 13 of the sample 10 using electron beam-assisted deposition, as described with reference to Figure 6B. Next, a trench 16 is milled into the first surface of the sample using FIB 30 to provide a flat polished surface in the region of interest 11. The method then proceeds by milling the inclined trench 20, as described with reference to Figures 1B and 6C, where the milling beam inclination angle of the inclined trench 20 is selected using the milling beam inclination angle from the previous step to ensure a uniform thickness in the electron-transparent region 15.
[0089] The optional step in Figure 7 is not necessary when the method of the present invention is used to prepare a sample including a thin film deposited on the first surface 12 (except when it is desirable to expose the underlying layer), or when the first surface 12 of the sample 10 has already been completely polished, but in other cases it is useful for preparing a high-quality electron-transmissive region 15 for subsequent analysis.
[0090] In some embodiments of this method, one or more method steps are configured to be performed automatically by the apparatus. In particular, the apparatus may include software configured to perform automated routines for carrying out the sample preparation and / or characterization steps described above.
[0091] In the preferred embodiments described above, the analysis steps are performed within the same instrument, but it will be understood that some steps can be performed in different instruments. For example, the sample preparation and milling steps can be performed in a FIB SEM, and then the sample can be transferred to another SEM for further characterization. [Explanation of Symbols]
[0092] 10 samples 11. Surface Region of Interest 12. First side 13. The second side 14 Common Edge 40 Electron Beam Systems 41 Electron beam
Claims
1. A method for preparing samples for analysis, A step of providing a sample comprising a region of surface interest on a first surface of the sample and a second surface, wherein the second surface is oriented at an angle to the first surface around a common edge between the first surface and the second surface, and the second surface extends between the common edge and the opposing second edge of the second surface of the sample. A step of milling the second surface of the sample such that a trench is provided on the surface of the second surface, wherein the trench extends from a first position on the second surface between the common edge and the second edge to a second position adjacent to the common edge, and the trench is arranged to provide an electron-transmissive sample layer including the surface region of interest, Methods that include...
2. The method according to claim 1, wherein the trench in the second surface includes an inclined trench having a deepest side portion adjacent to the common edge between the first surface and the second surface.
3. The aforementioned sloped trench, A first internal surface parallel to the surface region of interest, the bottom of which defines the deepest point of the trench, An angled bottom surface that is angled upward away from the bottom of the first inner side surface and in contact with the surface of the second surface, The method according to claim 2, including the method described in claim 2.
4. The method according to any one of claims 1 to 3, wherein the trench is arranged such that the electron-permeable sample layer is parallel to the first surface of the sample.
5. The method according to any one of claims 1 to 4, wherein the step of milling the second surface of the sample is performed using a milling beam system, and the method further includes the step of oriented the sample such that the milling beam is parallel to a region of interest.
6. The method according to any one of claims 1 to 5, wherein the milling step is performed using one or more of a focused ion beam, a broad ion beam, and a laser.
7. The steps include directing a focused electron beam through the electron-transparent sample layer, The steps include: collecting the signal generated by the interaction between the focused electron beam and the electron-transparent sample layer using a detector; The method according to any one of claims 1 to 6, further comprising:
8. The step of collecting the aforementioned signal is, The method according to claim 7, comprising the step of collecting scattered electrons from the electron-transparent sample layer using an electron detector.
9. The method according to claim 7 or 8, comprising the step of directing the focused electron beam towards the milled surface of the electron-transparent sample layer and oriented the sample so that the electron detector is directed towards the first surface of the sample opposite to the milled surface.
10. The step of collecting the aforementioned signal is, The method according to any one of claims 7 to 9, further comprising a step of collecting X-rays generated in the electron-transparent sample layer using an X-ray detector.
11. The method according to claim 10, comprising the steps of directing the focused electron beam to the surface region of interest on the first surface of the sample and oriented the sample so that the X-ray detector is directed to the surface region of interest on the first surface of the sample.
12. The method according to any one of claims 1 to 11, wherein the first surface of the sample includes a polycrystalline surface layer.
13. The method according to claim 12, wherein the polycrystalline surface layer includes a thickness of less than 100 nm and / or includes a nanocrystalline structure having dimensions of less than 100 nm.
14. The method according to any one of claims 1 to 13, further comprising the step of preparing the first surface of the sample before the milling step in order to produce a deformation-free surface.
15. The method according to claim 2 or 3, further comprising the step of milling the first surface of the sample to produce a polished surface on the region of interest of the surface before milling the inclined trench on the second surface of the sample.
16. The method according to any one of claims 1 to 15, further comprising the step of analyzing the region of surface interest on the first surface of the sample using a surface property evaluation technique.
17. An apparatus for preparing a sample for analysis, wherein the sample includes a region of surface interest on a first surface of the sample and a second surface, the second surface is oriented at a certain angle to the first surface around a common edge between the first surface and the second surface, and the second surface extends between the common edge and the second edge of the sample on the opposite side of the second surface. The aforementioned device A milling beam system is configured to mill a second surface of a sample to create a trench on the surface of the second surface, wherein the trench extends from a first position on the second surface between the common edge and the second edge to a second position adjacent to the common edge, and the trench is configured to provide an electron-transmissive sample layer including a surface region of interest. An electron beam system arranged to direct a focused electron beam through the aforementioned electron-transparent sample layer, A detector arranged to collect the signal generated by the interaction between the focused electron beam and the electron-transparent sample layer, A device equipped with the following features.
18. The apparatus according to claim 17, further comprising an electron detector configured to collect scattered electrons from the electron-transparent sample layer and provide an electron intensity image.
19. The focused electron beam is directed towards the milled surface of the electron-transparent sample layer. The electron detector is directed toward the first surface of the sample opposite to the milled surface. The apparatus according to claim 18, configured to enable the orientation of the sample.
20. The apparatus according to claim 18 or 19, further comprising a processing unit configured to analyze the electron intensity image and identify one or more Kikuchi bands in the signal.
21. The aforementioned processing unit is The distance to the pattern center of the identified Kikuchi band is measured. Based on the measured distance, a correction is applied to the determined width of the Kikuchi band to mitigate the effects of gnomonic distortion. The apparatus according to claim 20, configured as follows.
22. The aforementioned processing unit is Based on the grid spacing of one or more candidate phases, calculate the minimum and / or maximum Kikuchi bandwidth. If the determined width is less than the calculated minimum Kikuchi band width or greater than the maximum Kikuchi band width, the identified Kikuchi band is ignored. The apparatus according to claim 20 or 21, configured as follows.
23. The apparatus according to any one of claims 17 to 22, comprising an X-ray detector configured to collect X-rays emitted from the electron-transparent sample layer and provide an X-ray energy spectrum.
24. The focused electron beam is directed towards the region of interest on the first surface of the sample. The X-ray detector is directed towards the region of interest on the first surface of the sample. The apparatus according to claim 23, configured to enable the orientation of the sample as described above.
25. An electron detector configured to collect scattered electrons from the electron-transparent sample layer and provide an electron intensity image, An X-ray detector configured to collect X-rays emitted from the electron-transmissive sample layer and provide an X-ray energy spectrum, A sample holder configured to move the sample relative to the electron beam system, electron imaging detector, and X-ray detector between a first orientation in which the focused electron beam is directed towards the milled surface of the electron-transparent sample layer and the electron detector is directed towards the first surface of the sample opposite to the milled surface, and a second orientation in which the focused electron beam and the X-ray detector are directed towards the region of interest on the first surface of the sample, The apparatus according to any one of claims 17 to 24, comprising the above.
26. The apparatus according to any one of claims 17 to 25, further comprising a memory for holding computer-readable instructions for causing the apparatus to perform the steps of the method according to any one of claims 1 to 16 during execution.