Adaptive Shape Dimensions for Optimal Focused Ion Beam Etching

Abstract

A method for evaluating a region of a sample that includes alternating layers of different materials, the method including using a focused ion beam to mill a portion of the sample that includes alternating layers of different materials, reducing the milled area, and repeating the milling and reducing steps multiple times during the de-layering process until the process is complete.

Description

Technical Field

[0001] Cross - reference to Related Applications This application claims the benefit of priority of U.S. Patent Application No. 16 / 859,974, filed on April 27, 2020. The disclosure of this U.S. Patent Application is hereby incorporated by reference in its entirety for all purposes.

Background Art

[0002] In the research of electronic materials and the process of manufacturing electronic structures from electronic materials, specimens of electronic structures can be used for microscopic tests for failure analysis and device verification. For example, a specimen such as a silicon wafer on which one or more electronic structures are formed can be milled and analyzed using a focused ion beam (FIB) to study specific characteristic features of the structures formed on the wafer.

[0003] Many of the latest electronic structures include many different material layers that alternately overlap in one or more parts of the structure. To analyze or study the characteristics of structures on samples formed to have multiple layers, removing one or more selected layers, known as delayering, can be performed using an FIB tool. When the delayering process is executed, secondary electrons are generated from the milled material. These secondary electrons can be detected to analyze the characteristics of the milled layer and the structure.

[0004] FIB tools are used in the delayering of various different structures, but improvements in delayering techniques are desirable.

Summary of the Invention

[0005] Embodiments of the present disclosure provide an improved method and system for delayerring a structure. Standard FIB techniques are used to delayer a specimen that includes a number of multilayers of different materials, such as an electronic structure formed on a semiconductor wafer. In order to control the exact milling depth, it is currently important to accurately know which layer of the structure is exposed and being milled. One clue that can be used to control the depth of milling is to measure the secondary electrons (SE) emitted during milling. As the process mills deeper into the specimen through a number of sets of multilayers, it becomes difficult, and sometimes even impossible, to accurately identify when the layer of the first material ends and the layer of the second material beneath it begins.

[0006] Embodiments of the present disclosure improve such standard techniques and enable the delayerring process to accurately identify the multilayers even when a deep recess is milled into the sample. The embodiments can be beneficially used whenever the multilayers produce a yield of secondary electrons that appear alternately during focused ion beam milling of the specimen.

[0007] In some embodiments, an improved signal-to-noise ratio of the secondary ion signal is obtained by gradually and repeatedly reducing the area to be milled according to a predefined milling recipe. In other embodiments, this improved signal-to-noise ratio is adaptively obtained by feedback from the measured secondary ion signal and the ratio of the noise of that signal. In other embodiments, this improved signal-to-noise ratio is obtained by counting the secondary electron signal when a generally flat area of the specimen is being milled and blocking the secondary electron signal when milling is being performed over a small non-flat area outside the generally flat region. In other embodiments, this improved signal-to-noise ratio is obtained by milling the specimen with an ion beam whose beam current profile is pre-distorted to be inversely proportional to the expected milling rate of a non-distorted uniform beam.

[0008] Some embodiments of the present disclosure are particularly effective in retarding electronic structures formed on a semiconductor wafer that includes a number of interlayers of different materials, such as those found in modern 3D NAND type flash memory devices, as well as other semiconductor logic, memory and circuitry, and MEMS systems and other structures. However, the embodiments are not limited to such, and can be useful in delaminating any sample having a number of sets of interlayers of different materials.

[0009] Some embodiments relate to a method of evaluating a region of a sample that includes interlayers of different materials. The method can include milling a portion of a sample that includes interlayers of different materials using a focused ion beam, reducing a milling area, and repeating the milling and reducing steps a number of times until the process is complete during a delamination process.

[0010] Some embodiments relate to a system for evaluating a region of a sample that includes interlayers of different materials. The system can include a vacuum chamber, a sample support configured to hold a sample within the vacuum chamber during a sample evaluation process, a focused ion beam (FIB) column configured to direct a charged particle beam into the vacuum chamber, and a processor and a memory coupled to the processor. The memory can include a plurality of computer-readable instructions that, when executed by the processor, cause the system to mill a portion of a sample that includes interlayers of different materials using a focused ion beam, reduce a milling area, and repeat the milling and reducing steps a number of times until the process is complete during a delamination process.

[0011] Some embodiments relate to a non-transitory computer-readable memory storing instructions for evaluating regions of a sample that include alternating layers of different materials, the evaluation being performed using a focused ion beam to mill a portion of the sample that includes alternating layers of different materials, reducing the milling area, and during the delayering process, repeating the milling and reducing steps a number of times until the process is complete.

[0012] Various implementations of the embodiments described herein can include one or more of the following features. During the milling process, secondary electrons from the alternating layers of different materials can be collected and used to determine the end point of the milling operation. Reducing the milling area can be performed according to a milling recipe defined prior to the milling operation. Reducing the milling area can include reducing the milling area with each iteration of the milling process operation. Reducing the milling area can include reducing the milling area every X iterations of the milling process, where X is between 2 and 1000. The alternating layers of different materials can include first and second layers that generate different numbers of secondary electrons when milled. The first layer can include a dielectric material and the second layer can include a metal. Reducing the milling area can be performed adaptively based on feedback from the signal-to-noise ratio of a signal representing secondary electrons generated during the milling process. The milling area can be reduced by reducing the scanning pattern of the ion beam in both the X and Y directions. The sample can be a semiconductor wafer. The sample can include at least 10 sets of alternating layers, and the milling process can be an iterative process of repeatedly scanning a portion of the sample with a focused ion beam to mill recesses that penetrate each of the 10 sets of alternating layers.

[0013] Some embodiments relate to a method for evaluating a region of a sample that includes alternating layers of different materials, the method comprising milling a portion of the sample that includes alternating layers of different materials using a focused ion beam, the milling being an iterative process of repeatedly scanning the portion of the sample with the focused ion beam to mill a recess to a depth of the sample, such that a small sub-region of the portion of the sample remains generally flat as the milling progresses, milling, and during the milling process, measuring and integrating data generated by a secondary ion detector when scanning a generally flat small region of the portion of the sample with the focused ion beam, while ignoring data collected by the secondary ion detector when scanning some regions outside the small region of the portion of the sample. According to some embodiments, the shape dimensions of the small region can be determined prior to milling. Also, in some embodiments, the size of the small region can be gradually reduced according to a predetermined formula over a number of iterations of the milling, or the size of the small region can be gradually reduced in response to data generated by the secondary ion detector over a number of iterations of the milling to maintain the signal-to-noise ratio of the data within a predetermined range.

[0014] Other embodiments relate to a method for evaluating a region of a sample that includes alternating layers of different materials, the method comprising milling a portion of the sample that includes alternating layers of different materials using a focused ion beam, the milling being an iterative process of repeatedly scanning the portion of the sample with the focused ion beam to mill a recess to a depth of the sample, such that a curved edge forms at the bottom of the recess as the milling process progresses, milling, and during the milling process, measuring and integrating data generated by a secondary ion detector when scanning a generally flat small region with the focused ion beam, while ignoring data collected by the secondary ion detector when scanning a curved region.

[0015] Another embodiment relates to a method of evaluating a region of a sample that includes alternating layers of different materials, the method comprising milling a portion of the sample that includes alternating layers of different materials using a focused ion beam, the milling being an iterative process of repeatedly scanning the portion of the sample with the focused ion beam to mill a recess to a depth of the sample, and increasing the beam current of the milling to a level inversely proportional to a predicted decrease in the etching rate at the periphery of the portion being milled of the sample relative to a predicted etching rate in a central region of the portion being milled of the sample in a region surrounding the portion being milled of the sample during the milling process.

[0016] To more fully understand the nature and advantages of the present disclosure, reference should be made to the following description and the accompanying drawings. However, it should be understood that each of the drawings is provided for illustrative purposes only and is not intended to define the boundaries of the scope of the present disclosure. Further, generally, unless it is apparent to the contrary in this description, where elements in different figures are identified by the same reference numerals, those elements generally have the same function or purpose, or at least similar functions or purposes.

Brief Description of the Drawings

[0017]

Figure 1

Figure 2A

Figures 2B - 2D

Figure 3

Figure 4

Figure 5

Figures 6A - 6C

Figures 7A - 7B

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

[0018] Embodiments of the present disclosure provide an improved method and system for delayered structuring. Standard FIB techniques used to delayer a specimen, such as an electronic structure formed on a semiconductor wafer, can be inadequate when the specimen includes a number of interlayers of different materials. As the standard FIB process mills the specimen, secondary electrons are generated from the milled material. These secondary electrons can be detected to analyze the characteristics of the milled layers and structures. In a multilayer structure that includes a number of interlayers of different materials, it can be difficult to accurately identify when a layer of a first material ends and a layer of a second material beneath it begins. Embodiments of the present disclosure improve such standard techniques and enable the delayered structuring process to accurately identify the interlayers even when a deep recess is milled into the sample through a number of sets of interlayers.

[0019] To more fully understand and appreciate the present disclosure, first refer to FIG. 1. FIG. 1 is a simplified schematic diagram of a focused ion beam (FIB) evaluation system 100 according to some embodiments of the present disclosure. As shown in FIG. 1, the system 100 can include, among other things, a focused ion beam (FIB) column 110, a sample support element 140, and a secondary electron detector 150. The FIB column 110 is operable to generate a parallel charged particle beam 120 and direct the particle beam toward a sample 130 (sometimes also referred to herein as an “object” or “specimen”) to mill the sample or otherwise process the sample. This sample, such as a semiconductor wafer, can be supported on a support element 140 within a vacuum chamber 105.

[0020] The FIB column 110 can mill the sample 130 (e.g., form a recess in the sample 130) by irradiating the sample with the charged particle beam 120 to form a cross-section, and can smooth the cross-section if desired. The FIB milling process typically functions by placing the specimen in a vacuum environment and emitting a focused beam of ions towards the specimen to remove the material on the specimen by etching or milling. In some examples, this vacuum environment can be purged with a controlled concentration of background gas that serves to assist in controlling the etching rate and quality or to assist in controlling the material deposition. The ions to be accelerated can be generated from xenon, gallium, or other suitable elements, and typically, the ions are accelerated towards the specimen by a voltage in the range of 500 volts to 100,000 volts, more typically in the range of 5,000 volts to 50,000 volts. The beam current is typically in the range of a few picoamperes to a few microamperes depending on the FIB instrument configuration and application, and the pressure is typically controlled between 10 -10 ~10 -5 mbar.

[0021] The delayer process can be carried out by (i) locating the position of interest to be milled to remove a certain thickness of material from the sample, (ii) moving the sample (e.g., by a mechanical support element) so that the sample is positioned under the field of view of the FIB unit, and (iii) milling the sample to remove the desired amount of material at the position of interest. This delayer process can include forming a recess (usually a recess with lateral and longitudinal dimensions in the range of several μm to several tens of μm) in the sample.

[0022] This milling process typically involves reciprocating a charged particle beam over a specific area of the sample (e.g., in a raster scan pattern) at a constant speed during imaging or milling. As is known to those skilled in the art, this scan pattern can be implemented by one or more lenses (not shown) coupled to the charged particle column. The area to be scanned is typically a very small portion of the overall area of the sample. For example, the sample may be a semiconductor wafer with a diameter of 200 or 300 mm, while each area scanned on the wafer may be a rectangular area having a width and / or length of several μm or several tens of μm.

[0023] During the milling operation, the charged particle beam 120 generated by the FIB column 110 propagates through the evacuated environment formed within the vacuum chamber 105 and then impinges on the sample 130. When ions collide with the sample, secondary ions 125 are generated and detected by the secondary ion detector 150. The characteristics of the milled layers and structures can be analyzed using the detected secondary electrons.

[0024] Although not shown in FIG. 1, the FIB evaluation system 100 can include several additional components, including but not limited to one or more gas nozzles for supplying process gas to the chamber 105, vacuum valves and other valves for controlling the pressure within the chamber 105, and one or more lenses for guiding the charged particle beam. As would be known to those skilled in the art, the system 100 can further include one or more controllers, processors, or other hardware units that control the operation of the system 100 by executing computer instructions stored in one or more computer-readable memories. By way of example, this computer-readable memory can include solid-state memory (e.g., random access memory (RAM) and / or read-only memory (ROM), which can be programmable, flash-updatable, and / or of other similar types), disk drives, optical storage devices, or other similar non-transitory computer-readable storage media.

[0025] Next, refer to FIGS. 2A - 2D. FIG. 2A is a simplified cross-sectional view of a sample semiconductor wafer 200, and FIGS. 2B - 2D are simplified cross-sectional views of the semiconductor wafer 200 at different stages of the delayer process. The sample 200 includes a number of sets of interlayers 220, 230. Specifically, FIGS. 2A and 2B show 5 sets, and FIGS. 2C and 2D show 9 sets of interlayers, but it is understood that the sample 200 can have a reasonable number of sets of interlayers and, in many instances, can have 20 or more layers 220, 230.

[0026] Layer 220 and layer 230 have different chemical compositions, and layer 220 and layer 230 each have different yields of secondary electrons when milled using a focused ion beam such as charged particle beam 120. A non-limiting list of examples according to various embodiments of the present disclosure can include an interlayer of silicon oxide and silicon nitride or polycrystalline silicon and silicon oxide, or a suitable structure formed from two or more other materials that alternately overlap and have different yields of secondary electrons. These two or more other materials that alternately overlap can be polycrystalline silicon, silicon oxide, silicon nitride, aluminum oxide or other oxides or composite materials, aluminum, tungsten, copper, gold, platinum or other metals or alloys or composite materials characterized by metals, photoresist materials, or other materials currently used or to be used in the art in manufacturing or research. In some embodiments, each layer 220 can be a dielectric material and each layer 230 can be a conductive material, and in some embodiments, one or more of layers 220 and 230 can include a thin adhesive layer or a thin barrier layer.

[0027] With respect to layer 220 or 230, embodiments of the present disclosure are not limited to specific materials as long as these two layers that overlap each other include materials with different chemical compositions and generate different amounts of secondary electrons during the milling process. For example, FIG. 2A schematically shows that when sample 200 is milled using ion beam 210, the number of secondary electrons 225 generated by layer 220 of the first type of material is greater than the number of secondary electrons 235 generated by milling layer 230 of the different second type of material. Since layer 220 of the first type of material generates more secondary electrons than layer 230 of the second type of material when milled, the output signal generated by secondary detector 150 can be used to distinguish when layer 220 is being milled from when layer 230 is being milled.

[0028] Next, referring to FIG. 2B, between the initial portions of the delayer process, the delayered portion is relatively shallow and the surface of the milled recess 240 of the substrate 200 is generally flat. At this stage of the milling process, the signal from one layer overrides the signal from the other layer, and it is relatively easy to detect when the milling process transitions from milling layer 320 to milling layer 330, or vice versa when the milling process transitions from milling layer 330 to milling layer 320.

[0029] As the milling process progresses and mills the recess deeper into the sample 200, due to the etching characteristics of the milling process, the contour of the recess begins to change. Specifically, at the bottom of the milled area 240, a rounded edge 245 begins to form. As a result of the rounded edge 245, both the regions of layer 220 and 230 are etched simultaneously, and as a result, secondary electrons 225 and 235 are generated simultaneously from those different layers, and those secondary electrons interfere with each other.

[0030] Finally, when the milling process reaches the stage shown in FIG. 2D, the signals from the different layers are sufficiently mixed such that the signal generated by the secondary electron detector becomes noisy enough that it becomes difficult or even impossible to detect which layer is currently being milled at the bottom of the recess.

[0031] Figure 3 is a graph showing the signal 300 generated by the secondary electron detector during the milling process shown in FIGS. 2B - 2D. In FIG. 3, the X - axis indicates the time directly related to the milling depth, and the Y - axis indicates the signal intensity. As can be seen from FIG. 3, at the beginning of the milling process (area 310), each of the layers 220, 230 has a strong contribution to this signal that is easily distinguishable. As the milling process progresses and the depth of the milled recess changes from the recess shown in FIG. 2B to the recess shown in FIG. 2C, the signal 300 begins to degrade (area 320), but it is still easy to detect when the milling process transitions from one of the layers 220, 230 to the other. Then, as the milling process makes the recess deeper into the sample, the signal 300 degrades and finally becomes an unclear signal (area 330), from which it is not possible to easily determine whether the milling process is milling one of the layers 220 or one of the layers 230.

[0032] Due to the curved nature of the edge of the bottom of the recess 240, the recess may begin to take on the shape of a bowl when viewed from above. FIG. 4 is a simplified top view of the sample within area 410 of the sample 200 in which the recess 240 has been milled. As shown in FIG. 4, within the area of the recess where the edge of the bottom of the recess 240 is curved, the layers 220, 230 are partially visible. At the center of the milled recess, there is a relatively flat portion 420, and the relatively flat portion 420 represents the surface of the recess where only one of these two layers is exposed.

[0033] Embodiments of the present disclosure provide an improved method for delaying a sample and a system for delaying a sample. In this method and system, throughout the milling process, the secondary ion signal remains strong even when a recess is being milled deeply into the sample. Thus, embodiments of the present disclosure enable the delayer process to identify different types of materials that overlap alternately, such as the materials of layers 220 and 230, even when a recess is being milled very deeply into the sample.

[0034] In some embodiments, an adaptive geometry technique is used in which the FIB scan pattern gradually changes as the recess in the sample becomes smaller and smaller as it is etched deeper. In such embodiments, the FIB mills only a relatively flat small area (e.g., area 420) in each iteration, so that mainly one of the layers 220, 230 is exposed to the ion beam, and in each iteration, secondary ions are generated mainly from the layer exposed to the ion beam or only from the layer exposed to the ion beam.

[0035] FIG. 5 is a flowchart showing the steps of an adaptive geometry milling method 500 according to some embodiments of the present disclosure. As shown in FIG. 5, the method 500 includes milling a desired area (block 510) and repeatedly reducing the milling area until the milling process is complete (block 520). During this milling process, secondary electrons can be continuously collected and used to analyze the milled portion of the sample, both to determine the end point of the milling process.

[0036] Each iteration of the milling process can include reciprocating the milling area with the ion beam and scanning it at a constant speed in a raster scan pattern. Thus, the scan pattern, which can be stored in the memory coupled to the FIB system 100, continuously becomes smaller as the recess in the sample is milled deeper and deeper. One iteration removes a very thin layer of the sample material. In each iteration, a single atomic layer or a thinner layer may be removed. In a typical electronic structure, the thickness of the interlayers 220, 230 of the structure can vary from a single atomic layer to several micrometers. Thus, it may take from one iteration to thousands of iterations of scanning to delayer each of the individual layers 220, 230.

[0037] Figures 6A - 6C are schematic diagrams showing a shrinking scanning pattern according to some embodiments. As shown in Figure 6A, at the initial stage of the milling process for milling a recess in the sample area 610, the scanning pattern 620 can substantially cover the entire surface area of the area 610. In this process, later (for example, after thousands of iterations), the size of the scanning pattern can be gradually reduced to the pattern 630 shown in Figure 6B. Then, near the end of this process (for example, after thousands of additional iterations), the final scanning pattern can be made even smaller to the pattern 640.

[0038] In some embodiments, in method 500, the milling area can be reduced with each iteration, resulting in a generally smooth contour along the sidewalls of the milled recess. Figure 7A is a schematic cross - sectional view of a sample 700 having a recess 710 milled according to such a technique. As shown in Figure 7A, the recess 700 is milled through a number of interlayers 720, 730 of different materials, forming a smooth sidewall 740 in the recess 710 over a particular area of the sample during milling.

[0039] In other embodiments, in method 500, the milling area can be reduced after a predetermined number of iterations, such as every 10 iterations or every 50 iterations. In some embodiments, this predetermined number of iterations can be between 2 and 1000. In other embodiments, the milling area can be reduced after milling through each layer. FIG. 7B is a simplified cross-sectional view of a sample 750 having a recess 760 milled according to such a technique. Similar to sample 700, sample 750 includes a number of sets of interlayers 720, 730 made of different materials. When milling the first layer of material 720, the milling area per iteration is the same until the milling process mills through the bottom of the first layer of material 720 and reaches the first layer of material 730. The milling area is then reduced, and the first layer of material 730 is milled according to the scanning pattern of this reduced milling area until the milling process mills through the bottom of the first layer of material 730 and reaches the second layer of material 720. This process can be continued in this way, reducing the milling area each time the next successive layer is reached until the milled recess 760 is completed. The recess 760 formed in this way can exhibit sidewalls 770 showing a stepped contour as shown in FIG. 7B.

[0040] FIG. 8 is a graph showing a signal 800 generated by a secondary electron detector during a milling process based on process 500 shown in FIG. 5, according to some embodiments. In FIG. 8, the X-axis represents time that is directly related to the milling depth, and the Y-axis represents signal intensity. As can be seen from FIG. 8, which can be contrasted with FIG. 3, the intensity of signal 800 decreases over time, but different layers of the sample (e.g., layers 720, 730) each show a strong contribution to this signal that is readily distinguishable throughout the milling process.

[0041] FIG. 9 is a flow diagram showing the steps of an adaptive shape dimension milling method 900 according to some additional embodiments of the present disclosure. Method 900 is similar to method 500 in that the area to be milled can be reduced over time. Instead of reducing the area each iteration, or every X iterations, or every time individual layers are delayered as performed in method 500, method 900 continuously monitors the quality of the secondary ion signals (e.g., signal 300 and signal 800) and based on this signal determines when to reduce the milling area. For example, in some embodiments, method 900 reduces the milling area (block 930) only after method 900 detects that the signal quality has deteriorated to a degree that it is difficult to detect which layer is being milled (block 925). In other embodiments, method 900 can reduce the milling area only after method 900 detects that the signal quality has deteriorated to a previously determined noise level (block 925). In any of the embodiments, after the signal degradation exceeds a predetermined threshold level, the area to be milled can be reduced and milling can continue until milling is complete or the signal degrades again (block 910), and if the signal degrades again, the milling area can be reduced again.

[0042] As shown in FIG. 8, the secondary ion signal can be an alternating signal characterized by its peak and valley values or alternating extreme values. The ratio between these values defines a basic signal-to-noise level assuming a steady level of additive noise. Embodiments of the present disclosure can use an algorithm that samples the signal and compares adjacent maximum and minimum levels of the signal, where the adjacent maximum and minimum levels correspond to the passage of two adjacent layers. For an initial clear signal, there will be a certain ratio between the adjacent maximum and minimum signals. Embodiments of the present disclosure can set a threshold for this ratio in a configuration file of an algorithm stored in a computer-readable memory. Then, the embodiment can reduce the milling area by a certain magnification when this ratio becomes smaller than the threshold, thereby increasing this ratio. Then, this newly adjusted area (i.e., the reduced milling area) can be maintained until this ratio becomes smaller than the threshold again.

[0043] In additional embodiments, the scanning pattern can be kept invariant throughout the milling process. Instead, the signal generated by the secondary ion detector can be measured and integrated only when scanning a relatively flat small area within the milling process with an ion beam, and this relatively flat small area represents the bottom of a depression during milling within a larger area being milled. This way, the signal (e.g., signal 300) used to analyze the properties of the material within the milled depression represents only the portion where milling is being performed over an exposed portion (e.g., area 420) of one of the layers 220, 230 of the milling process and not when milling is being performed over the curved portion of the depression, and this signal does not degrade over time.

[0044] For further explanation, reference is made to FIGS. 10 and 11. FIG. 10 is a simplified flowchart showing the steps associated with a method 1000 for delaying a sample according to some embodiments, and FIG. 11 is a simplified top view of a portion of an electronic structure 1100 formed on a wafer after a recess has been partially milled into the semiconductor wafer, similar to the figure shown in FIG. 4. As shown in FIG. 10, method 1000 is an iterative process similar to methods 500 and 900. However, method 1000 monitors (block 1020) whether, for each iteration of the milling process, an ion beam is directed over a generally flat small area of the sample towards the sample.

[0045] For example, referring to FIG. 11, an area 1110 that has already been milled through several different alternatingly overlapping layers of materials 220, 230 is shown by a dashed line. At the bottom of the milled recess there is a small area 1120 (also shown by a dashed line) that has a top surface where a layer made of material 220 is exposed. FIG. 11 further shows a scanning pattern where, in this case, the FIB tool is operated in the forward mode (shown by line 1130) but not in the reverse mode (shown by the dotted line 1140). That is, the sample is bombarded with an ion beam when moving the beam from left to right within area 1110, and the bombardment pauses when returning the field of view of the ion beam in the incoming direction to start the next scanning operation.

[0046] During each iteration of the scanning pattern (blocks 1010, 1020, 1030), the embodiment according to method 1000 ignores (blocks 1024) the signal generated by the secondary electron detector when the ion beam is outside the small region 1120 (cuts it off), and measures and integrates the signal only when the ion beam is within the small region 1120 (block 1022). Then, when the iteration is complete, the next iteration begins (blocks 1040, milling complete = no), and this cycle is repeated until the entire recess is milled (blocks 1040, complete = yes). Thus, method 1000 effectively ignores the portion of the signal that would have been generated when the ion beam is on the edge of the curved bottom of a recess, such as the curved region 245 shown in FIG. 2C, if other methods were used.

[0047] In another embodiment, method 1200 according to the present disclosure measures and integrates the secondary ion signal when scanning the entire region 1110 with the ion beam, but the processing circuit of the FIB evaluation tool adds a pre-distortion value to the signal that is inversely proportional to the expected milling profile. Thus, the resulting effect can be to increase the beam current in regions near the periphery where the etching rate is normally low, inversely proportional to the expected decrease in the etching rate, so that the resulting etching rate is uniform across the entire recessed area. In some examples, method 1200 may be similar to method 1000, except that in method 1200, instead of collecting or ignoring the secondary ion signal in blocks 1022 and 1024, a pre-distortion value is added (block 1224) or no pre-distortion value is added (block 1222).

[0048] In the above description, for purposes of explanation, specific terms were used to provide a thorough understanding of the described embodiments. However, it will be apparent to those skilled in the art that those specific details are not necessary to practice the described embodiments. For example, some of the specific embodiments of the present disclosure described above used semiconductor wafers as samples, but the present disclosure is not limited to samples that are semiconductor wafers, and the present disclosure can be used to mill other types of samples on which or over which a number of multilayers of different materials are formed. Further, the exemplary FIB system shown in FIG. 1 includes a single FIB column, but in other embodiments, the evaluation system can include an SEM column and / or an optical microscope within system 100 in addition to FIB column 110. Further, the various examples of the present disclosure discussed above include multiple sets of multilayers in which the layers alternate between a first layer and a second layer, but embodiments of the present disclosure are not limited to just two layers that alternate, for example, in some embodiments, a set of multilayers can include a first, a second, and a third layer that each generate different numbers of secondary electrons when milled, and other embodiments can include an appropriate number of different layers that alternate.

[0049] Accordingly, the foregoing description of the specific embodiments described herein is presented for purposes of illustration and description. It is not intended to be exhaustive or to limit the embodiments to the exact forms disclosed. Further, while different embodiments of the present disclosure have been disclosed above, the specific details of the specific embodiments can be appropriately combined without departing from the spirit and scope of the embodiments of the present disclosure. Further, it will be apparent to those skilled in the art that many modifications and variations are possible in light of the above teachings.

[0050] To the extent that it is possible to implement the illustrated embodiments of the present disclosure using electronic components and circuits known to those of ordinary skill in the art without obscuring or departing from the teachings of the present disclosure, the details of those electronic components and circuits are not described more than is considered necessary as illustrated above in order to understand and recognize the underlying ideas of the present disclosure.

Claims

1. A method for evaluating a region of a sample containing interlayers of different materials, the method comprising: milling, using a focused ion beam, an area of the sample containing the interlayers of different materials; reducing the milling area; and during the evaluation process, repeating the milling step and the reducing step a number of times until the milling is complete; including the reducing of the milling area being performed adaptively by feedback from a signal-to-noise ratio of a signal representing secondary electrons generated during the milling; method.

2. The method for evaluating a region of a sample according to claim 1, wherein during the milling, secondary electrons from the interlayers of different materials are collected and used to determine an end point of the milling.

3. The method for evaluating a region of a sample according to claim 1, wherein reducing the milling area is performed by a predefined milling recipe defined prior to the milling.

4. The method for evaluating a region of a sample according to claim 3, wherein reducing the milling area includes reducing the milling area for each iteration of the milling.

5. The method for evaluating a region of a sample according to claim 3, wherein the milling recipe includes a plurality of iterations, and the milling recipe reduces the milling area in some of the plurality of iterations and does not reduce the milling area in other of the plurality of iterations.

6. The method for evaluating a region of a sample according to claim 1, wherein the interlayers of different materials include first and second layers that generate different numbers of secondary electrons when milled.

7. The method for evaluating a region of a sample according to claim 6, wherein the interlayers of different materials include at least first, second, and third layers that each generate different numbers of secondary electrons when milled.

8. The method for evaluating a region of a sample according to claim 1, wherein the sample includes at least 10 sets of interlayers, and the milling is an iterative process of repeatedly scanning a portion of the sample with the focused ion beam to mill depressions that penetrate each of the 10 sets of interlayers.

9. The method for evaluating a region of a sample according to any one of claims 1 to 8, wherein the sample is a semiconductor wafer.

10. A method for evaluating a region of a sample according to claim 9, wherein the interlayer is a part of a 3D-NAND flash memory structure.

11. A method for evaluating a region of a sample according to any one of claims 1 to 8, wherein the contour of the side wall of the recess formed by the milling is angled inward, and the contour is substantially smooth from the top of the recess to the bottom of the recess.

12. A method for evaluating a region of a sample according to any one of claims 1 to 8, wherein the contour of the side wall of the recess formed by the milling includes a plurality of steps from the top of the recess to the bottom of the recess that is smaller than the top.

13. A system for evaluating a region of a sample containing an interlayer of different materials, the system comprising: a vacuum chamber, a sample support configured to hold a sample within the vacuum chamber during a sample evaluation process, a focused ion beam (FIB) column configured to direct a charged particle beam into the vacuum chamber, a processor and a memory coupled to the processor wherein the memory includes a plurality of computer-readable instructions that, when executed by the processor, mill an area of the sample containing the interlayer of different materials using a focused ion beam, reduce the milling area, and during the sample evaluation process, repeat the milling step and the reducing step a number of times until the milling is complete including the reducing of the milling area is performed adaptively by feedback from a signal-to-noise ratio of a signal representing secondary electrons generated during the milling, system.

14. A method for evaluating a region of a sample containing an interlayer of different materials, the method comprising: milling a portion of the sample containing the interlayer of different materials using a focused ion beam, wherein the milling is an iterative process of repeatedly scanning the portion of the sample with the focused ion beam to mill a recess to a depth of the sample, even as the milling progresses, a small area of the portion of the sample remains substantially flat, the milling, and during the milling, When scanning the generally flat small region of the portion of the sample with the focused ion beam, measure and integrate the data generated by the secondary ion detector. On the other hand, when scanning some regions outside the small region of the portion of the sample with the focused ion beam, ignore the data collected by the secondary ion detector. Repeat this many times until the milling is completed. including Over many repetitions of the milling, gradually reduce the size of the small region in response to the data generated by the secondary ion detector to maintain the signal-to-noise ratio of the data within a predetermined range. Method.

15. The method for evaluating a region of a sample according to claim 14, wherein the shape dimensions of the small region are determined prior to the milling.

16. The method for evaluating a region of a sample according to claim 14, wherein the size of the small region is gradually reduced according to a predetermined formula over many repetitions of the milling.

17. During the milling, collect and use secondary electrons from the multilayers of different materials to determine the end point of the milling. The method for evaluating a region of a sample according to claim 14.

18. The sample is a semiconductor wafer, The multilayers of different materials include at least three sets of multilayers of a first layer and a second layer that generate different numbers of secondary electrons when milled. The milling is an iterative process of repeatedly scanning the portion of the sample with the focused ion beam, thereby milling a recess that penetrates each of the three sets of multilayers. The method for evaluating a region of a sample according to any one of claims 14 to 17.

Images