Method and system for generating a diffraction image

Abstract

Methods and systems for generating a diffraction image include acquiring a plurality of frames from a direct detection detector in response to irradiating a sample with an electron beam. A plurality of diffraction peaks in the plurality of frames are identified. A first dose rate of at least one of the identified diffraction peaks is estimated in a counting mode. If the first dose rate does not exceed a threshold dose rate, a diffraction image containing the diffraction peak is generated by counting electron detection events. Values of pixels belonging to the diffraction peak are determined with a first set of count parameter values corresponding to a first coincidence region. Values of pixels not belonging to any of the plurality of diffraction peaks are determined using a second set of count parameter values corresponding to a second, different coincidence region.

Description

Technical Field

[0001] This specification generally relates to methods and systems for generating diffraction images, and more specifically to generating electron diffraction images using a direct detection camera. Background Technology

[0002] For example, the structure of macromolecules such as proteins can be determined based on the electron diffraction patterns of their crystals. Accurate 3D structure resolution based on crystallographic data requires accurate capture of diffraction peak intensities. Detection of electron diffraction patterns generally requires cameras or detectors with high dynamic range. Traditional scintillator-based cameras convert received electrons into photons before they are detected by the imaging sensor. These cameras struggle to detect weak diffraction peaks. Direct detection cameras directly detect imaging electrons without scintillators and can capture weak diffraction peaks. However, the applicant recognizes that the dynamic range of electron counting using some direct detection cameras can be limited. For example, for relatively strong low-resolution diffraction peaks, under typical experimental conditions, the dose rate at the peak location may exceed the permissible dose rate, and the detection quantum efficiency may be reduced due to overlap losses at high dose rates. Summary of the Invention

[0003] In one embodiment, a method includes irradiating a sample with an electron beam; acquiring multiple frames from a detector; identifying multiple diffraction peaks in the multiple frames; estimating a first dose rate for at least one of the identified diffraction peaks in a counting mode; and generating a diffraction image containing the diffraction peaks by counting electron detection events in response to the first dose rate not being greater than a threshold dose rate, wherein a first set of counting parameter values ​​corresponding to a first overlapping region is used to determine the values ​​of pixels belonging to the diffraction peaks, and a second set of counting parameter values ​​corresponding to a second distinct overlapping region is used to determine the values ​​of pixels not belonging to any of the multiple diffraction peaks. This expands the dynamic range of a direct detection camera operating in electron counting mode and yields high-quality electron diffraction images.

[0004] It should be understood that the above overview is provided to present in a simplified form some concepts further described in the detailed description. It is not intended to identify key or essential features of the claimed subject matter, the scope of which is uniquely defined by the claims following the detailed description. Furthermore, the claimed subject matter is not limited to embodiments that address any shortcomings mentioned above or in any part of this disclosure. Attached Figure Description

[0005] Figure 1 Example of transmission electron microscopy.

[0006] Figure 2 Display the integrated image and the detector pixels corresponding to the diffraction peaks in the integrated image.

[0007] Figure 3 This demonstrates various acquisition mechanisms used to read data from a camera.

[0008] Figure 4 It is an instance method used to generate diffraction images.

[0009] Figure 5 It is a flowchart used to determine the weighting factor of the diffraction peak and the optimal estimated dose rate.

[0010] Figure 6 This paper presents examples of the relationship between detection quantum efficiency and dose rate in different signal processing modes.

[0011] Figure 7 This is another instance method for generating diffraction images.

[0012] Throughout the various views of the diagram, the same reference numerals refer to the corresponding parts. Detailed Implementation

[0013] The following description relates to systems and methods for generating diffraction images (e.g., electron diffraction images) using a direct detection (DD) camera in a charged particle microscopy system. Examples of this microscopy system are described in... Figure 1 The image is shown in the diagram. A DD camera captures scattered electrons from a sample in response to illumination from an electron beam. The DD camera comprises a pixelated image sensor with multiple detector pixels. Pixel voltages are generated at the corresponding detector pixels in response to electron impacts on the image sensor. Pixel voltages of frames (which can contain all detector pixels in the image sensor) can be read out at high frame rates. Data from each frame is stored and analyzed to generate a diffraction image.

[0014] One method for generating diffraction images is based on particle counting. In particle counting, each electron detection event (i.e., the detection of an electron arriving at the detector) is assigned to or attributed to a detector pixel. Particle counting can produce high detection quantum efficiency (DQE) due to its lack of detector dark noise. However, it is impossible to determine whether one event or multiple events produced the measured signal. DQE may be attributed to the decrease in coincidence loss with increasing dose rate, where multiple electrons arrive between two consecutive readouts that cannot be spatially separated. Particle counting is suitable for capturing weak diffraction peaks in beam-sensitive samples, such as protein crystals and sensitive organic crystals, where the sample is irradiated with a low electron dose. Some low-resolution diffraction peaks may have high dose rates, thus leading to coincidence loss and reduced DQE.

[0015] To address the aforementioned issues, different sets of counting parameter values ​​are used for high-dose-rate and low-dose-rate regions in a diffraction image generated by counting electron detection events. In other words, in the diffraction image, the values ​​of pixels belonging to diffraction peaks are determined using a first set of counting parameter values, and the values ​​of pixels not belonging to diffraction peaks are determined using a second, different set of counting parameter values. Diffraction peaks can be identified from multiple acquired frames. For example, pixels belonging to diffraction peaks are identified from an integrated image obtained by adding multiple frames. The counting parameters may include one or more filter coefficients and camera settings. Each set of counting parameter values ​​can be mapped to, and therefore corresponds to, a coincident region. The coincident region is the smallest detector region that spatially resolves two electron detection events. The coincident region is a statistical property of the detection process. The first set of counting parameters corresponds to a first coincident region, and the second set of counting parameters corresponds to a second coincident region. By processing camera readouts (i.e., raw measurement data) using different sets of counting parameter values ​​corresponding to different coincident regions, different regions of the detector may have different sensitivities to spatial coincidence. The first and second sets of counting parameter values ​​can be determined experimentally or through simulation. In one example, a first set of count parameter values ​​corresponds to a first smaller overlap region, and a second set of count parameter values ​​corresponds to a second larger overlap region. Thus, the smaller overlap region is used for diffraction peaks, and the larger overlap region is used for background. In this way, DQE at high dose rates is increased for particle counting, and overlap loss is reduced.

[0016] DQE can be further or alternatively increased in the high dose range by improving the dose rate estimation of diffraction peaks. In one example, a first dose rate of at least one diffraction peak identified in multiple acquired frames is estimated in counting mode. In counting mode, the dose rate of the diffraction peak (el / peak / s) is estimated based on statistics of individual electronic events in each camera frame during a series of frames. For example, the dose rate is calculated based on the event rate of electronic detection events at one or more detector pixels. The first dose rate is compared to a threshold dose rate. If the first dose rate does not exceed the threshold dose rate, the first dose rate is the optimal estimated dose rate of the diffraction peak. If the first dose rate is higher than the first threshold dose rate, a second dose rate of the diffraction peak is estimated in integrated mode. The optimal estimated dose rate of the diffraction peak is generated based on both the first and second dose rates. The first and second dose rates are the average number of electrons per diffraction peak per second, in el / peak / s. The optimal estimated dose rate corresponds to the optimal signal-to-noise ratio. At the optimal estimated dose rate, the DQE is not less than the first DQE at the first dose rate in counting mode or the second DQE at the second dose rate in integrated mode. The intensity of diffraction peaks in a diffraction pattern can be determined based on the optimal estimated dose rate.

[0017] In another example, after identifying diffraction peaks from multiple acquired frames, a second dose rate for at least one diffraction peak is estimated in an integrated mode. This second dose rate is compared to a second threshold dose rate. If the second dose rate is not less than the second threshold dose rate, then the second dose rate is the optimal estimated dose rate. If the second dose rate is less than the second threshold dose rate, a first dose rate for the diffraction peak is estimated in a counting mode. The optimal estimated dose rate is calculated based on both the first and second dose rates. In the counting mode, the dose rate is estimated based on the event rate of electronic detection events corresponding to the diffraction peak. In the integrated mode, the dose rate is estimated based on the cumulative pixel voltage corresponding to the diffraction peak and the detector conversion efficiency. Figure 6 As shown, the DQE in the counting mode decreases with increasing dose rate, while the DQE in the integrated mode remains constant with increasing dose rate. By determining the optimal estimated dose rate in the combined mode (or both the counting and integrated modes), the DQE corresponding to the optimal estimated dose rate is no less than the first DQE at the first dose rate in the counting mode, or the second DQE at the second dose rate in the integrated mode. Therefore, the intensity of diffraction peaks with a large dynamic range can be accurately captured. Compared to conventional particle counting, the dynamic range of the DD camera can be extended to higher dose rates without adjusting the data readout mechanism.

[0018] In some instances, instead of dose rate, first and second average doses (e1 / peak / frame) of diffraction peaks can be estimated in counting and / or integration modes. The optimal estimated average dose can be determined based on the first and second average doses.

[0019] In one instance, the dose rate of the diffraction peak is estimated in both counting and integration modes to generate a first dose rate and a second dose rate, respectively. The optimal estimated dose rate of the diffraction peak is calculated by weighting the first and second dose rates using a weighting factor. The weighting factor can be determined based on the first and second dose rates. In some instances, the initial dose rate is determined based on the first and second dose rates. The weighting factor is calculated based on the first DQE at the initial dose rate in counting mode and the second DQE at the initial dose rate in integration mode. The first and second DQEs can be determined based on known relationships between DQE and dose rate, such as... Figure 6 As shown. In some instances, the optimal estimated dose rate can be determined iteratively through an optimization process. For each iteration, the weighting factor is updated based on the optimal estimated dose rate calculated from the previous iteration, and the optimal estimated dose rate is updated based on the updated weighting factor.

[0020] In one instance, after acquiring multiple frames of scattered electrons from a sample, an integrated image is generated by summing the frames. One or more diffraction peaks are identified in the integrated image. In other words, the detector pixel in the imaging sensor corresponding to each diffraction peak is located. Processing the signal of a particular diffraction peak involves processing the pixel voltage acquired from the detector pixel corresponding to the diffraction peak. In some instances, the dose rate of the diffraction peaks is estimated in both counting and integrated modes to determine the optimal estimated dose rate. In some instances, multiple diffraction peaks are grouped into one or more groups, and a weighting factor determined from one diffraction peak in a group can be applied to the other diffraction peaks in that group. Diffraction peaks can be grouped by the estimated dose rate in both counting and integrated modes. Alternatively, diffraction peaks can be grouped based on their position, such as their distance from the optical axis of the microscope or the center of the diffraction pattern in the integrated image.

[0021] In one example, the count image of the diffraction pattern is formed based on the counting of electron detection events, and the integrated image of the diffraction pattern is formed by integrating (or summing) the detected signals. The count image can be generated based on the counting parameters corresponding to the same or different overlapping regions of the diffraction peaks and background. The diffraction image is generated based on the count image, the integrated image, and the weighting factors of the diffraction peaks and the optimal estimated dose rate. The diffraction image can be generated by scaling the weighted sum of the count image and the integrated image with the optimal estimated dose rate.

[0022] In one instance, for each diffraction peak, a first dose rate is estimated in counting mode and a second dose rate is estimated in integrated mode. In response to a first dose rate below a first threshold dose rate, an optimal estimated dose rate is determined in counting mode. The diffraction image is a counting image scaled to the optimal estimated dose rate. In response to a second dose rate above a second threshold dose rate, an optimal estimated dose rate is determined in integrated mode. The diffraction image is an integrated image scaled to the optimal estimated dose rate. In some instances, the first threshold dose rate and the second threshold dose rate are the same. In other instances, the second threshold dose rate is higher than the first threshold dose rate. In response to a first dose rate greater than the first threshold dose rate and / or a second dose rate lower than the second threshold dose rate, an optimal estimated dose rate is determined in combined mode based on a weighted sum of the first and second dose rates. The weighting factor used to generate the optimal estimated dose rate can also be applied to the counting image and the integrated image to generate the diffraction image. The diffraction peaks in the diffraction image are scaled based on the optimal estimated dose. Therefore, in a single diffraction image containing multiple diffraction peaks, the intensities of at least two diffraction peaks can be estimated in different modes. Therefore, noise in different regions of a diffraction image can have different statistical distributions. For example, noise in the first region, which contains signals processed in counting mode, can have a Poisson distribution, while noise in the second region, which contains signals processed in integration mode, can have a Landau distribution.

[0023] In one instance, the image sensor can be reset after each readout. By resetting, the pixel voltage of each detector pixel can be reset to a reset value (e.g., zero). The raw data read from each frame can be directly used to analyze the electrons received in each frame. In another instance, the camera can be read out based on multi-frame correlated dual sampling (mfCDS), as disclosed in U.S. Application 13 / 645,725, filed October 5, 2012, entitled "Method for acquiring data with an image sensor," which is incorporated herein by reference in its entirety for all purposes. In mfCDS, multiple frames of raw data are read out sequentially from the image sensor before resetting it. The difference in the sequentially acquired pixel voltages is used to analyze the electrons received in each frame. Therefore, the data collected via the mfCDS mechanism is preprocessed to obtain differential frames, and then processed to estimate the optimal dose rate.

[0024] Figure 1 A transmission electron microscopy (TEM) system 100 operating in a selected region (SA) diffraction mode is shown. The TEM system 100 includes an electron source 10 that emits an electron beam 11 along an optical axis 110 toward a condenser system 12. The electron source 10 can generate high-energy electrons, i.e., electrons with typical energies between about 10 keV and 1,000 keV. In some embodiments, the condenser system 12 may include one or more condenser lenses and one or more apertures. A deflector 19 positioned downstream of the condenser system 12 deflects and / or tilts the electron beam relative to the optical axis 110. A sample front objective 16 positioned downstream of the deflector 19 collimates the electron beam and guides it onto a sample 14. The sample 14 may be held in a sample plane 111 by a sample holder 13. In some instances, the sample is positioned on a TEM grid attached to the sample holder. The sample holder 13 can adjust the sample position by tilting the sample relative to the optical axis and / or translating the sample within the sample plane. Scattered electrons originating from sample 14 sequentially pass through sample back objective 123 and projector system 21, and are collected by detector 25 positioned on the opposite side of sample 14 relative to electron source 10. Detector 25 can detect the received electrons and send the signal to image processor 24 to form an image. Detector 25 may include an amplifier for amplifying the signal before sending it to image processor 24. In one example, detector 25 is a direct detection camera. In some embodiments, other detectors may be used for diffraction pattern acquisition and sample image acquisition.

[0025] Dashed line 41 illustrates the beam path of unscattered electrons from sample 14 to detector 25. Projector system 21 images the back focal plane 43 of sample rear objective 123 onto detector 25. Beam stopper 17 is inserted into optical axis 110 to block the unscattered beam. SA aperture may be positioned in SA plane 44. Alternatively, condenser aperture in condenser system 12 may act as beam-limiting aperture. In another example, image deflector may be positioned between sample and detector to move and tilt electrons transmitted through sample back to optical axis, such that the ED pattern remains centered on detector during beam tilting and the image remains centered on detector during beam shifting. In some embodiments, the TEM system does not include a beam stopper, and the detector receives the unscattered beam.

[0026] The controller 30 can control the operation of the TEM system 100 manually or automatically in response to operator instructions or according to computer-readable instructions stored in non-transitory memory (or computer-readable medium) 32. The controller 30 may include a processor and is configured to execute computer-readable instructions and control various components of the TEM system 100 to implement any of the methods described herein. The controller 30 can adjust the beam position and / or beam incident angle on the sample by adjusting the deflector 19. The controller 30 can adjust the electron dose of each ED pattern by adjusting one or more settings of the illumination optics, the exposure time of each frame acquired by the detector, and the angular velocity of beam tilt / sample rotation. The controller 30 can adjust the magnification by adjusting the projector system 21. The controller 30 can also be coupled to a display 31 to display notifications and / or signals detected by the detector 25. The controller 30 can receive user input from a user input device 33. The user input device 33 may include a keyboard, mouse, or touchscreen.

[0027] Although the TEM system has been described by way of example, it should be understood that sample images and diffraction patterns can be acquired using other charged particle microscopy systems. As another example, a charged particle microscopy system is a scanning transmission electron microscope (STEM) system. This discussion of the TEM system is provided only as an example of a suitable form of imaging.

[0028] Figure 2An example integrated image 201 is shown, illustrating a detector pixel corresponding to a diffraction peak. The shadow 210 at the center of the integrated image is caused by a beam stopper used to block high-intensity unscattered beams. Integrated image 201 contains multiple local diffraction peaks with a wide dynamic dose rate range. The dose rate of each peak of the low-resolution diffraction peaks located near the center of the integrated image is typically higher than that of the high-resolution diffraction peaks farther from the center. An integrated image can be obtained from a pixelated camera containing multiple detector pixels. Each of the diffraction peaks corresponds to multiple detector pixels. For example, curve 204 shows a portion of the pixelated camera. A signal from diffraction peak 203 is received from the shadow detector pixel 205. A pixel voltage is generated in response to an electron impact on the detector pixel. In some instances, multiple detector pixels may generate non-zero pixel voltages in response to an electron impact on the camera.

[0029] Figure 3 Example camera readout sequences 301-303 are shown, illustrating diffraction peaks at various dose rates. The x-axis indicates time. Solid arrow 311 indicates frame readout, and dashed arrow 312 indicates camera reset. Each star indicates that the electron arriving at the detector pixel corresponds to a diffraction peak. The camera is reset immediately after each frame readout. In sequence 301, the dose rate is low, such that no more than one electron arrives between two consecutive readouts. Sequence 302 corresponds to mfCDS, where the camera is reset after multiple frame readouts. Similar to 301, the dose rate in 302 is low, and no more than one electron arrives between two consecutive readouts. In sequence 303, similar to 301, the camera operates at the same frame rate and is reset after each frame readout. However, unlike 301, the dose rate is higher, and multiple electrons can arrive between consecutive readouts. If the signal received in sequence 303 is processed in counting mode, the DQE will be lower due to overlap loss.

[0030] Figure 4 A method 400 for generating a diffraction image based on a signal acquired using a DD camera is demonstrated. Diffraction peaks are identified. The intensity of each identified diffraction peak in the diffraction image is determined based on an optimally estimated dose rate, which is based on dose rate estimates in both counting and integration modes.

[0031] At position 402, set the system parameters. These parameters may include the electron dose rate at the sample location, the camera readout sequence, and the readout frame rate. The electron dose rate can be determined based on the sample properties. For example, the electron dose rate will be lower for beam-sensitive samples. The camera readout sequence can be adjusted based on the expected dose rate at the camera location. Figure 3As shown, in different readout sequences, the camera can be reset after each readout or after multiple readouts. Furthermore, at position 402, the region of interest (ROI) of a sample can be identified and moved into the field of view. For example, the ROI can be determined based on a low-resolution sample image.

[0032] At 404, the electron beam is directed to the ROI. In one instance, the electron beam is a parallel or quasi-parallel beam. In another instance, the electron beam is a nanoprobe beam. At 406, the camera acquires multiple frames at the frame rate determined at 402. Each frame contains the pixel voltages of multiple detector pixels of the camera. Each frame is stored separately.

[0033] At position 408, an integrated image is generated, and diffraction peaks in the acquired frames are identified and located within the integrated image. In one example, the integrated image is generated by summing all frames acquired from the ROI at position 406. In the integrated image, a single electron may contribute to the pixel value of more than one pixel. Diffraction peaks correspond to high-intensity regions in the integrated image. Diffraction peaks can be identified by their local peak intensities. The detector pixel corresponding to each diffraction peak is identified. For example, such as... Figure 2 As shown, detector pixel 205 corresponds to a pixel in the integrated image 201 that has a pixel value greater than the threshold intensity.

[0034] At 410, one of the identified diffraction peaks is selected. In other words, the detector pixel corresponding to the selected diffraction peak is selected, and the signal received from these selected detector pixels is used to generate a weighting factor and the optimal estimated dose rate for the selected diffraction peak at 412.

[0035] At 412, a weighting factor and optimal estimated dose rate for a selected diffraction peak are determined based on a first dose rate estimated in count mode and / or a second dose rate estimated in integrated mode. The optimal estimated dose rate can be a weighted sum of the first and second dose rates estimated in both count and integrated modes. The weighting factor ranges from zero to one. The sum of the weighting factors for both count and integrated modes is one. In one example, the optimal estimated dose rate... Calculated separately by weighting factors and (1- Weighted first and second dose rates: Equation 1.

[0036] exist Figure 5 The document describes a detailed procedure for determining the weighting factor and the optimal estimated dose rate.

[0037] At 414, method 400 checks whether the signal from any other diffraction peak needs to be processed. If the answer is "yes," another diffraction peak is selected at 410. Otherwise, method 400 proceeds to 416.

[0038] At 415, a count image is generated from multiple frames. The count image is generated by counting electron detection events, where each event in which an electron reaches the camera (i.e., the selected detection event) is assigned to one pixel. In one instance, it is useful to distinguish between diffraction peaks and a continuous background within the diffraction pattern for an optimal count image. Spatial overlap (two events in a single frame being so close that they cannot be separated) will occur more frequently in the diffraction peak region than in the background region. To reduce overlap loss, different sets of count parameter values ​​are used to process the signals in different regions of the count image. The count parameter values ​​are camera / detector specific. The count parameters may include filter coefficients used to perform filtering operations on the camera readout. The count parameters may also include one or more camera settings, including gain, readout order, signal quantization, and timing. Each set of count parameter values ​​corresponds to a region of overlap. The region of overlap is related to the spatial resolution of two electron detection events. The count parameter values ​​are chosen such that the region of overlap in the diffraction peak region is small, while the region of overlap in the background region is large. For example, the overlapping region of the diffraction peaks is 4×4 detector pixels, and the overlapping region of the background is 16×16 detector pixels. The smaller overlapping region makes it easier to spatially distinguish between the two events, which is useful in the peak region. The disadvantage of a small overlapping region is increased spurious detection. The trade-off between these two effects in the background and peak regions will vary. By using a smaller overlapping region for the diffraction peak region, DQE increases at higher dose rates. In another example, a set of counting parameters is used to generate the entire counting image. In yet another example, the counting image is generated using the counting parameters corresponding to the same overlapping region of the entire image.

[0039] At 416, a diffraction image containing multiple diffraction peaks can be generated based on the integrated image from step 408, the count image from step 415, the weighting factor, and the optimal estimated dose for each diffraction peak. In one example, a weighted image can be generated using a weighted sum of the integrated image and the count image, using the weighting factor. The diffraction peaks are then scaled using the optimal estimated dose to generate the diffraction image. In another example, for each pixel belonging to a specific diffraction peak, the pixel value in the weighted image is calculated as a weighted sum of the pixel values ​​in the count image and the pixel values ​​in the integrated image. For example, the pixel value in the weighted image is the pixel value in the count image multiplied by the weighting factor from 412. And integrate the pixel values ​​in the image by multiplying by (1- The sum of all pixel values ​​in the weighted image is used to calculate the diffraction peaks. For pixels that do not belong to any diffraction peak (i.e., background pixels), the pixel values ​​in the weighted image are the same as the corresponding pixel values ​​in the counted image. The diffraction peaks in the weighted image are then scaled based on the best estimated dose rate to generate a diffraction image. Through scaling, the sum of all pixel values ​​for each diffraction peak in the diffraction image is proportional to its best estimated dose rate.

[0040] In some embodiments, at 406, a motion estimation (or generation) of the best-estimated dose rate and / or diffraction image is performed on the acquired frames. For example, at one time point, a subset of the acquired frames acquired at 406 (e.g., frames 1 to N) is processed to generate a diffraction image. For a later time point, later-acquired frames (e.g., frames n to n+N (n≥1)) are processed to generate another best-estimated dose rate and diffraction image. In some instances, changes in the dose rate can be monitored based on changes in a second dose rate estimated in the integrated mode, and this change can be added to Poisson statistics to better estimate the dose rate in the counting mode.

[0041] In this way, the dose rate of diffraction peaks can be estimated using counting mode, integration mode, or combined mode. In combined mode, the dose rates estimated from counting mode and integration mode are combined to generate the optimal estimated dose rate. The estimation described in method 600 extends the dynamic range of conventional counting mode.

[0042] The diffraction image generated by method 400 may contain diffraction peaks generated based on dose rate estimates in any of the following modes: counting mode (where the best estimated dose rate equals a first dose rate), integration mode (where the best estimated dose rate equals a second dose rate), and combination mode (where the best estimated dose rate includes a fraction of the first dose rate and a fraction of the second dose rate). In one example, the dose rate of the low-resolution diffraction peaks is estimated in the integration mode, and the dose rate of the high-resolution diffraction peaks is estimated in the counting mode. In another example, the dose rate of the medium-resolution diffraction peaks is estimated in the combination mode. Therefore, different regions of the generated diffraction image may have different noise distributions. For example, the counting mode produces regions with a Poisson noise distribution, and the integration mode produces regions with a Landau noise distribution.

[0043] Figure 5 A method 500 is presented for determining the weighting factor and the optimal estimated dose rate of a selected diffraction peak based on the first dose rate and / or the second dose rate estimated in counting mode and integration mode, respectively.

[0044] At position 501, the first dose rate of the selected diffraction peak is estimated in counting mode. In one instance, the first dose rate is calculated based on the event rate of electron detection events. The electron arrival rate is described by a Poisson distribution. The probability that k electrons from a single diffraction peak hit the camera in a single frame is... Equation 2.

[0045] Where D is the average dose (el / peak / frame). Dose rate The average dose (el / peak / s) can be obtained by multiplying the average dose D by the frame rate F (frames / s). As an example, for a dose rate of 10 el / peak / s and a frame rate of 250 frames / s, the average dose is 10 / 250 = 0.04 el / peak / frame. At this low dose rate, an event consists of one electron from a diffraction peak hitting the camera. At higher dose rates, some events will involve multiple electrons, such as... Figure 3 As shown in 303.

[0046] The probability that no event (no electrons reaching the camera) occurs between consecutive frame readouts is: Equation 3 And the probability of an event (one or more electrons from a diffraction peak reaching the camera) between consecutive frame readouts is: Equation 4.

[0047] Therefore, the first dose rate It can be calculated from the probability of no event or the probability of the event: Equation 5.

[0048] At point 502, the first dose rate is compared to a first threshold dose rate. If the first dose rate does not exceed the first threshold dose rate, then the first dose rate is set as the optimal estimated dose rate. A weighting factor can be set accordingly. For example, the weighting factor in Equation 1... It is set to 1. The first threshold dose rate can be determined based on one or more of the electronic dose rate at the sample, the sample type, and the frame rate. If the first dose rate is greater than the first threshold dose rate, then method 500 proceeds to 503.

[0049] At position 503, the second dose rate is estimated in the integrated mode. The second dose rate can be calculated based on the pixel voltage of the detector pixel corresponding to the selected diffraction peak and the conversion efficiency of the camera. For example, the second dose rate is the product of the second average dose and the frame rate, where the second average dose is calculated by dividing the sum of the pixel values ​​corresponding to the selected diffraction peak in the integrated image by the conversion efficiency.

[0050] At point 504, the second dose rate is compared to the second threshold dose rate. If the second dose rate is not less than the second threshold dose rate, then the second dose rate is set as the optimal estimated dose rate. Weighting factors can be set accordingly. For example, the weighting factors in Equation 1... It is set to 0. The second threshold dose rate can be equal to or greater than the first threshold dose rate. The second threshold dose rate can be determined based on one or more of the electronic dose rate at the sample, the sample type, and the frame rate. If the second dose rate is less than the second threshold dose rate, the initial dose rate is determined at 506.

[0051] At position 506, an initial dose rate is determined. The optimal estimated dose rate is initialized using this initial dose rate. In one instance, the initial dose rate is calculated based on a first dose rate and / or a second dose rate. The initial dose rate can be equal to either the first or the second dose rate. The initial dose rate can be the average of the first and second dose rates. In one instance, the initial dose rate can be a predetermined dose rate, for example, the dose rate corresponding to the average dose rate of 1 el / peak / frame.

[0052] At point 508, the DQE is determined based on the predetermined DQE-dose rate relationship. cnt and DQE int These are the DQE at the current best estimated dose rate in counting mode and integrated mode, respectively. The DQE-dose rate relationship varies for different cameras and different camera operating conditions. The DQE-dose rate relationship can be determined through mathematical calculation, simulation, or experimentation. DQE at a specific dose rate cnt and DQE int Depend on Figure 6 The curves 602 and 603 shown in the figure or the lookup table derived from the curves are used to determine this.

[0053] Figure 6 Example curves of DQE versus dose rate are shown in count mode, integrated mode, and combined mode. The camera operates at a frame rate of 250 frames / s. In this example, the DQE 602 in count mode is close to 1 at very low dose rates and begins to decrease rapidly when the dose rate is greater than 25 el / peak / s. The DQE 603 ​​in integrated mode is relatively constant at 0.5. At a dose rate of approximately 300 el / peak / s, the DQE 602 in count mode is lower than the DQE 603 ​​in integrated mode. At any dose rate, the DQE 601 in combined mode is not lower than the DQE 602 in count mode or the DQE 603 ​​in integrated mode. For example, at the current best estimated dose rate... Below, DQE cnt It is 604 and DQE. int The value is 605.

[0054] At point 516, based on the DQE from step 508. cnt and DQE int Calculate the weighting factors. For example, the variance of the output signal in integrated mode and counting mode is calculated from the corresponding DQE: Equation 7 Equation 8 in and These are the variances of the output signals in integration mode and counting mode, respectively. } represents the variance of the input signal. The weighting factor can be obtained. for: Equation 9 in and .

[0055] At point 518, the optimal estimated dose rate is updated based on the weighting factor, the first dose rate, and the second dose rate. In one instance, as shown in Equation 1, the weighting factor... It can be applied to the first and second dose rates of a specific diffraction peak to obtain the optimal estimated dose rate.

[0056] At 520, method 500 checks whether the weighting factor needs to be updated. In one instance, the weighting factor is calculated only once, and at 524, the weighting factor and optimal estimated dose rate generated at 518 are output. In some instances, the weighting factor and optimal estimated dose rate are determined iteratively via an optimization process, where the updated optimal estimated dose rate is set to the current optimal estimated dose rate and used to calculate the variance of the input signal at 514. The weighting factor update can be terminated after a predetermined number of iterations. Alternatively, the weighting factor update can be terminated when the difference between the change in the optimal estimated dose rate and / or the weighting factor is less than a threshold. If method 500 determines that the weighting factor does not need to be updated, the current weighting factor and optimal estimated dose rate are output to method 400. Otherwise, if the weighting factor and optimal estimated dose rate need to be updated, method 500 proceeds to 508, where the DQE is updated based on the current optimal estimated dose rate. cnt and DQE int .

[0057] By processing the readout frame in combined mode, the DQE at the optimal estimated dose rate is no less than the DQE at the first dose rate in counting mode or the DQE at the second dose rate in integrated mode. When the dose rate at the camera is low (e.g., below 25 el / peak / s), the optimal estimated dose rate is close to the first dose rate. When the dose rate is high (e.g., above 750 el / peak / s), the optimal estimated dose rate is close to the second dose rate. In some instances, the first and second dose rates for all diffraction peaks can be determined before selecting the diffraction peak at 410.

[0058] Figure 7 Another example method 700 is shown for generating diffraction images based on signals acquired using a DD detector. Unlike... Figure 4In method 400, and in method 700, diffraction peaks are grouped such that the signals of diffraction peaks in the same group are processed in the same mode (counting, integration, or combination). In some instances, the same weighting factor is applied to diffraction peaks in the same group.

[0059] At positions 702-708, similar to Figure 4 Steps 402-408 involve acquiring multiple frames in response to irradiating the sample with an electron beam after setting the system parameters. The detector pixel corresponding to each diffraction peak in the multiple frames is then identified.

[0060] At 710, similar to Figure 4 415, generating count images from multiple frames.

[0061] At position 712, the first and second dose rates of all diffraction peaks are optionally calculated in both counting and integration modes. The first and second dose rates of each diffraction peak can be calculated as follows: Figure 5 Calculate as described in steps 501 and 503.

[0062] At 714, the diffraction peaks are grouped into one or more groups. In one example, the diffraction peaks are grouped based on their position relative to the optimal estimation axis, resolution, or distance from the center of the diffraction pattern. Diffraction peaks within a certain distance range from the optimal estimation axis can be grouped together. In another example, the diffraction peaks are grouped based on their first dose rate and second dose rate. Diffraction peaks with a first dose rate and / or a second dose rate within a predetermined dose rate range can be grouped together.

[0063] At position 716, select a group determined at position 714. At position 718, calculate the weighting factor for the group. The weighting factor can be calculated based on the first and second dose rates of the diffraction peaks in the group, such as... Figure 5 As shown.

[0064] At 720, method 700 checks if there are any remaining groups for calculating the weighting factor. If the answer is "yes", method 700 proceeds to 716 to select another group. If the weighting factors for all groups have been determined, method 700 proceeds to 722.

[0065] At position 722, the optimal estimated dose rate for all diffraction peaks is calculated based on the corresponding first and second dose rates of all diffraction peaks and the weighting factors of the corresponding groups. The optimal estimated dose rate can be calculated according to Equation 1. Furthermore, a diffraction pattern containing multiple diffraction peaks can be generated, where the intensity of each diffraction peak is proportional to its corresponding optimal estimated dose rate.

[0066] By processing diffraction peaks in groups, the time from the overall signal to the image can be reduced without sacrificing image quality.

[0067] The technique of generating diffraction images using counting parameters corresponding to different overlap regions reduces overlap loss at diffraction peaks. Comparing the estimated dose rate in counting mode with a threshold dose rate improves DQE at high dose rates by estimating the optimal estimated dose rate in combined mode. The technique of estimating the dose rate of diffraction peaks in both counting and combined modes increases DQE over a wider dose rate range. In combined mode, DQE at higher dose rates is less affected by overlap loss compared to estimating the dose rate only in counting mode. The technique of generating a weighting factor combines the dose rates calculated in counting and combined modes to generate the optimal estimated dose rate. The intensity of diffraction peaks in the diffraction image is then determined based on the optimal estimated dose rate. The technique of iteratively searching for the optimal estimated dose rate finds the dose rate with the highest DQE. The technique of identifying diffraction peaks in the combined image accurately identifies detector pixels corresponding to specific diffraction peaks due to the high contrast of the diffraction peaks in the combined image.

[0068] In another presentation, a method includes: irradiating a sample with an electron beam; acquiring multiple frames from a detector; identifying multiple diffraction peaks in the multiple frames; estimating a first dose rate for at least one of the identified diffraction peaks in a counting mode, wherein the first dose rate does not exceed a threshold dose rate; and generating a diffraction image containing the diffraction peaks by counting electron detection events, wherein the values ​​of pixels belonging to the diffraction peaks are determined using a first set of counting parameter values ​​corresponding to a first overlapping region, and the values ​​of pixels not belonging to any of the multiple diffraction peaks are determined using a second set of counting parameter values ​​corresponding to a second distinct overlapping region.

[0069] In another presentation, a method includes: irradiating a sample with an electron beam; acquiring multiple frames from a detector; identifying multiple diffraction peaks in the multiple frames; estimating a first dose rate of at least one of the identified diffraction peaks in a counting mode, wherein the first dose rate is higher than a threshold dose rate; a second dose rate of the diffraction peak in an integrated mode; determining an optimal estimated dose rate of the diffraction peak based on the first dose rate and the second dose rate; and generating a diffraction image based on the optimal estimated dose rate.

Claims

1. A method for generating a diffraction image, comprising: Irradiate the sample with an electron beam; Multiple frames are acquired from the detector; Identify multiple diffraction peaks in the multiple frames; Estimate the first dose rate of at least one of the identified diffraction peaks in counting mode; as well as In response to the first dose rate not being greater than a threshold dose rate, a diffraction image containing the diffraction peaks is generated by counting electron detection events, wherein a first set of count parameter values ​​corresponding to a first overlapping region is used to determine the values ​​of pixels belonging to the diffraction peaks, and a second set of count parameter values ​​corresponding to a second overlapping region is used to determine the values ​​of pixels not belonging to the plurality of diffraction peaks, the second overlapping region being different from the first overlapping region.

2. The method according to claim 1, wherein the first overlapping region is smaller than the second overlapping region.

3. The method of claim 1, wherein the first set of count parameter values ​​and the second set of count parameter values ​​are detector-specific.

4. The method of claim 1, wherein estimating the first dose rate of the diffraction peak in the counting mode comprises estimating the first dose rate based on the event rate of the electron detection events corresponding to the diffraction peak.

5. The method according to any one of claims 1-4, further comprising: In response to the first dose rate being greater than the threshold dose rate, a second dose rate of the diffraction peak is estimated in the integrated mode; The optimal estimated dose rate for the diffraction peak is determined based on the first dose rate and the second dose rate; as well as A diffraction image containing the diffraction peaks is generated, wherein the intensity of the diffraction peaks is determined based on the optimal estimated dose rate.

6. The method of claim 5, wherein estimating the second dose rate of the diffraction peak in the integrated mode comprises estimating the second dose rate based on the accumulated signal corresponding to the diffraction peak from the plurality of frames and the detector conversion efficiency.

7. The method of claim 5, wherein estimating the optimal estimated dose rate of the diffraction peak based on the first dose rate and the second dose rate comprises calculating the optimal estimated dose rate as a weighted sum of the first dose rate and the second dose rate.

8. The method of claim 5, wherein the first dose rate and the second dose rate are weighted by a weighting factor, and the method further comprises: Determine the first detection quantum efficiency in the counting mode; The second detection quantum efficiency is determined in the integrated mode; as well as The weighting factor is calculated based on the first detection quantum efficiency and the second detection quantum efficiency.

9. The method of claim 8, wherein the first detection quantum efficiency and the second detection quantum efficiency are determined based on the known relationship between the detection quantum efficiency and the dose rate in the counting mode and the integration mode, respectively.

10. The method of claim 5, wherein the optimal estimated dose rate is determined iteratively by updating the first detection quantum efficiency in the counting mode and the second detection quantum efficiency in the integrated mode based on a previously determined optimal estimated dose rate.

11. The method of claim 1, wherein identifying the plurality of diffraction peaks in the plurality of frames comprises: generating an integrated image by summing the plurality of frames; and identifying one or more diffraction peaks in the integrated image.

12. A method for generating a diffraction image, comprising: Irradiate the sample with an electron beam; Multiple frames are acquired from the detector; Identify multiple diffraction peaks in the multiple frames; as well as A diffraction image containing the plurality of diffraction peaks is generated from the plurality of frames by counting electron detection events, wherein the values ​​of pixels belonging to at least a first diffraction peak among the plurality of diffraction peaks are determined using a first set of counting parameter values ​​corresponding to a first overlapping region, and the values ​​of pixels not belonging to the first diffraction peak are determined using a second set of counting parameter values ​​corresponding to a second overlapping region, which is different from the first overlapping region, and the intensity of the first diffraction peak is determined based on a first dose rate of the first diffraction peak estimated in the counting mode.

13. The method of claim 12, wherein estimating the first dose rate of the first diffraction peak in the counting mode comprises estimating the first dose rate based on the event rate of electron detection events corresponding to the first diffraction peak.

14. The method of claim 12 or 13, wherein the intensity of at least the second diffraction peak is determined based on a second dose rate of the second diffraction peak among the plurality of diffraction peaks estimated in the integrated mode.

15. The method of claim 14, wherein estimating the second dose rate of the second diffraction peak in the integrated mode comprises estimating the second dose rate based on the accumulated signal corresponding to the second diffraction peak from the plurality of frames and the detector conversion efficiency.

16. The method of claim 15, wherein the intensity of the second diffraction peak is further determined based on a third dose rate of the second diffraction peak estimated in the counting mode.

17. The method of claim 14, wherein the different regions of the diffraction image contain noise with different statistical distributions.

18. A charged particle microscope, comprising: An electron source, used to generate an electron beam; Sample holder, used to position the sample; A pixelated detector for receiving electrons from the sample; as well as A controller, comprising non-transitory memory for storing computer-readable instructions, is configured to, by executing the instructions with a processor,: The sample is irradiated with the electron beam; Multiple frames are acquired using the detector; In counting mode, estimate the first dose rate of at least one diffraction peak in the acquired frame; In response to the first dose rate not being greater than a threshold dose rate, a diffraction image containing the diffraction peaks is generated by counting electron detection events, wherein a first set of counting parameter values ​​corresponding to a first overlapping region is used to determine the values ​​of pixels belonging to the diffraction peaks, and a second set of counting parameter values ​​corresponding to a second overlapping region is used to determine the values ​​of pixels not belonging to any of the plurality of diffraction peaks, the second overlapping region being different from the first overlapping region; In response to the first dose rate being higher than a threshold dose rate, a second dose rate of the diffraction peak is estimated in the integrated mode; The optimal estimated dose rate for the diffraction peak is determined based on the first dose rate and the second dose rate; as well as The diffraction image is generated, wherein the intensity of the diffraction peak is determined based on the optimal estimated dose rate.

19. The charged particle microscope of claim 18, wherein acquiring a plurality of frames with the detector comprises continuously reading from a plurality of detector pixels multiple times before resetting the detector; and obtaining the plurality of frames by subtracting the continuous readings.

20. The charged particle microscope of claim 18 or 19, wherein determining the optimal estimated dose rate of the diffraction peak based on the first dose rate and the second dose rate comprises: determining an initial dose rate; determining a first detection quantum efficiency corresponding to the initial dose rate in the counting mode; determining a second detection quantum efficiency corresponding to the initial dose rate in the integration mode; calculating a weighting factor based on the first detection quantum efficiency and the second detection quantum efficiency; and determining the optimal estimated dose rate by calculating a weighted sum of the first dose rate and the second dose rate using the weighting factor.

Images