Methods and systems for determining molecular structure

By acquiring multiple diffraction patterns and electron microscope images within the inclination angle range, combined with Fourier transform technology, the problem of phase information loss in the diffraction patterns is solved, and the precise analysis of the molecular structure is achieved.

CN112129795BActive Publication Date: 2024-11-12FEI CO
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202010578893.3
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Priority Date
2019-06-24
Filing Date
2020-06-23
Publication Date
2024-11-12
Estimated Expiration
2040-06-23

AI Technical Summary

Technical Problem

The prior art is difficult to capture the phase information of the molecular structure from the diffraction pattern, resulting in defects in molecular structure analysis.

Method used

By acquiring multiple diffraction patterns and electron microscopy images within the tilt angle range, combined with Fourier transform technology, the amplitude and phase of the structural factor are determined, thereby analyzing the molecular structure.

Benefits of technology

The amplitude and phase information of structural factors are effectively captured and parsed, improving the accuracy and integrity of molecular structure determination.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN112129795B_ABST
    Figure CN112129795B_ABST
Patent Text Reader

Abstract

Methods and systems for determining molecular structure. The molecular structure can be determined based on structure factors solved from diffraction patterns and electron microscope images of a sample. Specifically, the amplitude of the structure factor can be determined based on the intensity of each diffraction peak in a plurality of diffraction patterns. The phase of the structure factor can be determined based on the electron microscope image and the intensity of each diffraction peak.
Need to check novelty before this filing date? Find Prior Art

Description

[0001] The present specification relates generally to methods and systems for determining molecular structure, and more particularly to determining molecular structure based on diffraction patterns and electron microscope images. Background Art

[0002] The molecular structure of an organic molecule can be determined by its corresponding structure factor. The structure factor describes how the lattice of a crystal scatters incident radiation. The 3D molecular structure of the molecules that form a crystal can be resolved by determining the structure factor of the crystal. The scattered waves from the crystal interfere with each other and a two-dimensional diffraction pattern can be formed. When the scattering angle meets the Bragg condition, high intensities in the diffraction pattern are constructively formed. The structure factor contains amplitude and phase. Although the amplitude of the structure factor can be obtained from the diffraction pattern, the phase information is lost. In order to resolve the molecular structure, the phase information can be extracted from the electron microscope image of the crystal. Specifically, the phase information can be read out from the Fourier transform of the electron microscope image. Summary of the invention

[0003] In one embodiment, a method for determining a molecular structure comprises: acquiring a plurality of diffraction patterns of the sample while tilting the sample within a tilt angle range, determining an amplitude of a structure factor based on the intensity of diffraction peaks in the plurality of diffraction patterns; acquiring an electron microscope (EM) image of the sample; determining a phase of the structure factor based on the EM image and the intensity of the diffraction peaks in the plurality of diffraction patterns; and determining the molecular structure of the sample based on the amplitude of the structure factor and the phase of the structure factor. In this way, the molecular structure can be solved from the diffraction pattern and the EM image of the sample. The sample can be a thin slice prepared by milling a macro-crystal.

[0004] It should be understood that the above summary is provided to introduce in simplified form a series of concepts that are further described in the detailed description. It is not meant to identify key or essential features of the claimed subject matter, the scope of which is uniquely defined by the claims that follow the detailed description. Furthermore, the claimed subject matter is not limited to implementations that address any disadvantages noted above or in any part of this disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0005] Figure 1 A system for determining a molecular structure according to an exemplary embodiment is presented.

[0006] Figure 2A Example positions showing reflections of a 3D crystal and an Ewald sphere in reciprocal space.

[0007] Figure 2B Shown in Figure 2A Diffraction patterns obtained under .

[0008] Figure 3 is a high-level flow chart illustrating an example method for determining molecular structure based on diffraction patterns and electron microscope images.

[0009] Figure 4 Example positions of the lamellae used to obtain diffraction patterns and electron microscopy images are shown.

[0010] Figure 5 Thin sections at different tilt angles are shown.

[0011] Figure 6 is an example subroutine for acquiring a diffraction tilt series.

[0012] Fig. 7A Diffraction peaks sampled in multiple diffraction patterns are shown.

[0013] Figure 7B Shown for Fig. 7A The rocking curve generated by the diffraction peak.

[0014] Figure 8 An example subroutine for acquiring an electron microscope tilt series is shown.

[0015] Fig. 9 An example subroutine for determining the phase of a structure factor is shown.

[0016] Fig. 10A Example diffraction patterns are shown.

[0017] Fig. 10B Shown with Fig. 10A Fourier transform of low-dose electron microscopy images acquired at the same tilt angle.

[0018] Like reference numerals refer to corresponding parts throughout the several views of the drawings. DETAILED DESCRIPTION

[0019] The following description relates to systems and methods for determining molecular structure based on diffraction patterns and electron microscope (EM) images of a sample. In one example, the diffraction patterns and EM images can be obtained from Figure 1 The image is acquired using an electron microscope imaging system such as an electron microscope system.

[0020] The structure of a molecule can be solved based on its structure factor in reciprocal space, which is the Fourier transform of real space. The structure factor contains amplitude and phase. Both the amplitude of the structure factor and the phase of the structure factor are three-dimensional (3D) data sets defined in reciprocal space. Scattered waves from the planes of the lattice in real space form reflections in reciprocal space. Each reflection has a Friedelmate positioned symmetrically relative to the origin center of the reciprocal space. A reflection and its Friedelmate form a Friedel pair. Reflections can be captured or sampled by a 2D diffraction pattern. As Figure 2A As shown in , a diffraction peak in a diffraction pattern can appear only when its corresponding reflection overlaps with (or is within a threshold distance from) a curved Ewald sphere in reciprocal space. For weak phase objects, the amplitudes of the reflected Friedel pairs in a diffraction pattern can have the same intensity. However, due to the curvature of the Ewald sphere, a single diffraction pattern may not capture the amplitudes of the reflected Friedel pairs. In other words, Figure 2B As shown in the diffraction pattern of , the diffraction peaks in the diffraction pattern may not be centrosymmetric. The Friedel pairs of the diffraction peaks in the diffraction pattern may have different intensities. Therefore, the intensities of the Friedel pairs of the structure factor may be different. Further, when determining the phase of the structure factor based on the EM image, it is necessary to consider the additional phase shift caused by the unequal intensities of the Friedel pairs of the diffraction peaks.

[0021] Figure 3 An example method for determining the molecular structure of a 3D crystal based on a diffraction tilt series and an EM tilt series of a sample is shown. The diffraction tilt series includes diffraction patterns acquired at multiple tilt angles, and the EM tilt series includes EM images acquired at multiple tilt angles. The sample can be in the form of a thin slice. The amplitude of the structure factor can be determined based on the intensity of the diffraction peak in the diffraction tilt series. The phase of the structure factor can be solved based on the EM image of the sample and the intensity of the diffraction peak from the diffraction tilt series. Figure 4 Example positions on the lamella used to acquire diffraction patterns and EM images are shown. Figure 5 Thin sections at different tilt angles are shown. Figure 6 A subroutine for determining the amplitude of the structure factor is shown. Fig. 7A and 7B The procedure used to generate rocking curves of diffraction peaks is shown. Figure 8 A subroutine for acquiring an EM tilt series is shown. Fig. 9 A subroutine for determining the phase of the structure factor is shown. Fig. 10A and 10B Diffraction patterns and Fourier transforms of low-dose EM images acquired at the same tilt angle are shown, respectively.

[0022] Go to Figure 1, a transmission electron microscope (TEM) system 100 according to an embodiment of the present disclosure is shown. The TEM system 100 includes an electron source 10 that emits charged particles along an emission axis 110 (such as an electron beam 11) toward a column 12. The electron source 10 can generate high-energy electrons, that is, electrons with typical energies between about 10 keV and 1,000 keV. In some embodiments, the column 12 can include one or more condenser lenses such as condenser lenses 120 and 121, and one or more apertures such as aperture 122. The column 12 collimates the electrons generated from the electron source 10 and directs the electron beam onto a sample 14. The sample 14 can be held by a sample holder 13. The sample holder 13 can adjust the sample position by tilting and / or translating the sample. A condenser aperture 15 can be optionally inserted into the beam path to allow the electron beam to irradiate only selected areas of the sample.

[0023] TEM system 100 can be used to obtain both diffraction patterns and EM images of a sample. In diffraction imaging mode, scattered electrons can be transmitted through sample 14 and collected by detector 25 after sequentially traveling through objective lens 123, intermediate lens 126 and projector lens 127, as shown by solid lines. The unscattered beam may be blocked by beam stopper 17. The collected scattered electrons are used to form an electron diffraction pattern, such as a selected area electron diffraction (SAED) pattern. In EM imaging mode, beam stopper 17 is removed from the beam path. The electrons transmitted from sample 14 sequentially travel through objective lens 123, intermediate lens 126 and projector lens 127, and form a TEM image of the sample on detector 25, as shown by dotted lines.

[0024] The detector 25 can detect the received electrons and send a signal to the image processor 24 to form an image. The detector 25 can include an amplifier for amplifying the signal before sending the signal to the image processor 24. In one example, the detector 25 can be a CCD camera or a CMOS camera. In some embodiments, different detectors can be used for diffraction pattern acquisition and EM image acquisition.

[0025] The controller 30 can control the operation of the TEM system 100 manually in response to operator instructions or automatically according to computer readable instructions stored in a 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 so as to implement any method described herein. For example, the controller can adjust the imaging mode by adjusting the position of the selected area aperture 15 and the objective lens aperture 124 to obtain a SAED map or an EM image. The controller 30 can adjust the dose of the electron beam irradiated toward the sample by adjusting the current density. For example, the current density can be adjusted by adjusting the condenser optical device (such as condenser lenses 120 and 121). The controller 30 can adjust the profile of the incident beam by adjusting one or more apertures and / or lenses in the column 12. The controller 30 can adjust the position and / or orientation of the sample relative to the incident beam by adjusting the sample holder 13. The controller 30 can be further coupled to a display 31 to display a notification and / or image of the sample. The controller 30 can receive user input from a user input device 33. The user input device 33 may include a keyboard, a mouse or a touch screen.

[0026] Although a TEM system is described by way of example, it should be understood that the diffraction pattern and the EM image may be acquired by other microscope systems. In one embodiment, the diffraction pattern and the EM image may be acquired by different microscope systems. As an example, the diffraction pattern may be an X-ray diffraction pattern acquired from an X-ray diffraction system. As another example, the EM image may be acquired from a scanning transmission electron microscope (STEM) system. This discussion of the TEM system is provided only as an example of a suitable imaging modality.

[0027] Figure 2A The positions of example reflections of a 3D crystal and an Ewald sphere in a reciprocal lattice are shown. Figure 2B Shown in Figure 2A The corresponding diffraction pattern obtained under the conditions of Figure 2A In the reciprocal space, k x -k z The plane shows the position of the reflection. The reciprocal space is composed of the axis k x , k y and k z Defined. In direction 212, the 3D crystal is illuminated with a charged particle beam. Reflections (201, 202, 203, 204, and 205) represent a subset of reflections of the 3D crystal in reciprocal space. In this paper, reflections (201-205) are along k x -k zThe dotted line 210 in the plane is aligned. The arcuate shape 211 is part of the Ewald sphere. Although the reflected Friedel pairs may have the same amplitude, the Friedel pairs of the diffraction peaks may have different intensities due to the curvature of the Ewald sphere. Reflections close to (i.e., within a small distance of) the Ewald sphere appear as high-intensity diffraction peaks in the diffraction pattern. In this example, the reflections are aligned with the axis k. z The reflection on the left side (i.e., k x < 0) compared to axis k z The reflection on the right side (i.e., k x > 0) is closer to the Ewald sphere. In the corresponding diffraction pattern Figure 2B In the diffraction peaks (223, 224, 225 and 226) along the k x -k y The lines 222 in the plane are aligned. Most of the diffraction peaks appear on the right side of the image, to the right of the beam stop 221 located at the origin of the diffraction pattern (i.e., the origin of the reciprocal space). Diffraction peaks 223 and 224 are Friedel pairs. The Friedel pairing of diffraction peaks 225 and 226 is not captured in the diffraction pattern. In other words, the intensity of the Friedel pairing of diffraction peaks 225 and 226 is lower than the intensity of diffraction peaks 225 and 226, respectively.

[0028] Figure 3 An example method 300 for determining the molecular structure of a molecule is shown. The molecule can be in the form of a 2D crystal or a 3D crystal. The molecular structure can be determined by solving for the amplitude and phase of its structure factor. The amplitude of the structure factor can be determined based on a diffraction tilt series. The phase of the structure factor can be determined based on a diffraction tilt series and an EM tilt series.

[0029] At 302, a crystal sample is prepared. The sample may be in the form of a thin slice. The thin slice may be milled from a macroscopic crystal. For example, a thin slice may be formed from a macroscopic crystal by milling the macroscopic crystal using a charged particle beam, such as a focused ion beam. Macroscopic crystals may be prepared by standard protein crystallization methods such as sitting drop or hanging drop vapor diffusion or batch crystallization. The thickness of the thin slice may be determined based on the type of molecule being studied. For example, the thin slice thickness may be greater than the size of the unit cell of the crystal to ensure that the molecular structure is intact. The thin slice thickness may be below 300 nm. Thinner thin slices may reduce the effect of dynamic scattering on the amplitude of determining the structure factor. In some embodiments, the sample may include multiple thin slices.

[0030] Figure 4 An example slice 401 prepared by milling a macrocrystal with a focused ion beam is shown. For each slice, multiple locations can be imaged to obtain diffraction patterns and EM images. Figure 4In the figure, the solid squares on the sheet indicate the locations for EM image acquisition, and the dashed squares on the sheet indicate the locations for diffraction pattern acquisition. The imaging area for the diffraction pattern and the EM image can be the same. The pixel resolution of the image can be determined based on the desired resolution of the molecular structure. For example, a desired resolution of 5 angstroms in the molecular structure may require a pixel resolution of 2.5 angstroms in the EM image to meet the Nyquist condition. The imaging area can be determined by the radiation hardness of the crystal. In one example, 0.01-1 um can be in the imaging area. 2 .

[0031] At 303, set Figure 1 Setting up the imaging system may include loading the lamella into the system. Setting up the imaging system may include performing a coma-free alignment prior to imaging the sample.

[0032] At 304, a diffraction tilt series of the sample is acquired. A diffraction tilt series is a 3D data set containing multiple 2D diffraction patterns. These diffraction patterns can be obtained with a resolution of less than 0.1 el / Å. 2 For example, the current density is 0.01 el / Å. 2 .

[0033] In one embodiment, the thin slice can be continuously tilted at a tilt speed around the tilt axis. The tilt axis can be orthogonal to the incident beam. The tilt speed can be determined based on the desired sampling resolution and acquisition speed of the diffraction pattern. When each tilt angle difference Δθ tilts the thin slice, a diffraction pattern can be formed. In other words, each diffraction pattern spans the tilt angle difference Δθ. The tilt angle difference Δθ between the continuously acquired diffraction patterns can be determined based on the desired sampling resolution. For example, the tilt angle difference Δθ can decrease as the desired sampling resolution increases. Further, the tilt angle difference Δθ can be adjusted based on the oversampling rate to obtain a finer slice. The tilt angle difference Δθ can be inversely proportional to the oversampling rate. In one example, the tilt angle difference Δθ can be 3 degrees. That is, a diffraction pattern is acquired during every 3 degrees of tilt. In another embodiment, the thin slice can be tilted discretely. For example, the thin slice can be first tilted by the tilt angle difference Δθ. Then, when the sample is stationary relative to the beam, the charged particle beam is irradiated toward the sample to obtain a diffraction pattern.

[0034] Figure 5The positions of the thin slices in the real space defined by the axes x, y and z are shown at different tilt angles. The thin slices can be irradiated along the z-axis with a charged particle beam indicated by arrow 501. The thin slices can be tilted around an axis orthogonal to the incident beam. For example, the thin slices can be tilted or rotated around the y-axis, as indicated by arrow 510. The y-axis can correspond to the rotation axis of the sample holder. The tilt angle can be defined by the angle between the thin slice and the x-axis. For example, at position 502, the tilt angle of the thin slice is zero. At position 504, the positive tilt angle of the thin slice is 505. At position 503, the negative tilt angle of the thin slice is 506.

[0035] In some embodiments, the range of the tilt angle can be limited by the maximum rotation range of the sample holder. The tilt angle range can be the same as the maximum rotation range of the sample holder. For example, a diffraction tilt series can be acquired while tilting the thin slice from -70 degrees to +70 degrees. In one embodiment, the entire diffraction tilt series can be acquired at the same position of the sample. In another embodiment, the diffraction tilt series can be acquired from multiple positions of the sample. For example, a diffraction pattern with a tilt angle ranging from -70 degrees to 0 degrees can be acquired at a first position of the sample, and a diffraction pattern with a tilt angle ranging from 0 degrees to 70 degrees can be acquired at a second position of the sample. In order to avoid destroying the molecular structure, the number of diffraction patterns acquired at each position may be limited by the electron beam dose. In some embodiments, the diffraction tilt series can be acquired from multiple thin slices prepared at 302.

[0036] In some embodiments, the diffraction tilt series may include diffraction patterns acquired by tilting along two orthogonal axes. Figure 5 As shown in , a first set of diffraction patterns can be obtained by tilting the sample along the y-axis, and a second set of diffraction patterns can be obtained by tilting the sample along the x-axis. By including diffraction patterns acquired while tilting along two orthogonal axes, a more accurate profile of the diffraction peaks in 3D reciprocal space can be obtained. Further, the effect of dynamic scattering on the amplitude of the determined structure factor can be reduced.

[0037] At 306, the amplitude of the structure factor can be calculated based on the diffraction tilt series. The amplitude of the structure factor can be determined based on the intensity of the diffraction peaks in the diffraction tilt series. Each diffraction peak corresponds to a reflection in reciprocal space. Due to the finite size and / or beam divergence of the crystal, the diffraction peaks may not be delta functions, but occupy a volume in reciprocal space. In other words, the intensity of the diffraction peaks extends within the volume in reciprocal space (i.e., has non-zero intensity). Therefore, a single diffraction pattern cannot capture the full-angle spot profile of the diffraction peak, and the intensity of the diffraction peaks in multiple tilt angles is used to solve the structure factor amplitude.

[0038] Figure 6and Figure 7A-7B The structure factor amplitude calculation is shown in detail. The amplitude of the structure factor can be determined based on the angular spot profile (or rocking curve) of each diffraction peak over a subset of tilt angles at which the intensity of the diffraction peak is non-zero. Each rocking curve is generated by fitting the intensity of the diffraction peak over multiple tilt angles.

[0039] At 601, the intensity of each diffraction peak in the diffraction tilt series is determined. Determining the intensity of the diffraction peak may include identifying each diffraction peak by grouping high intensity spots belonging to a specific diffraction peak, and calculating the intensity of the diffraction peak in the diffraction pattern (or at the tilt angle) based on the total intensity of the high intensity spots of the diffraction peak in the diffraction pattern.

[0040] At 602, the diffraction tilt series can be thresholded. For example, after removing the background, low-intensity data points (or pixels) in the diffraction tilt series are filtered out by thresholding. The background may be caused by inelastic scattered electrons, electrons scattered from the sample support film, or amorphous parts of the sample. In one example, the background can be removed by low-pass filtering. After removing the low-intensity data points, only data points with intensities above a threshold level are retained. In one embodiment, the threshold level can be predetermined. In another embodiment, the threshold level can be determined based on the maximum intensity in the diffraction tilt series. For example, the threshold level can be 10% of the maximum intensity. In yet another embodiment, the threshold level can be determined by the sensitivity of the imaging system. For example, the threshold level can be the noise level of the diffraction tilt series.

[0041] At 604, diffraction peaks in the diffraction tilt series are identified. For example, identifying the diffraction peaks may include identifying high intensity spots in each diffraction pattern, and identifying the high intensity spots belonging to the diffraction peaks. A high intensity spot is an area in the diffraction pattern that contains a set of high intensity data points (or pixels). Identifying the high intensity spots may include determining the location and total intensity of the high intensity spots. The high intensity spots may not overlap with each other. In one embodiment, identifying the high intensity spots may include identifying the boundaries of each high intensity spot. The total intensity of each high intensity spot may be calculated by averaging or summing the intensities of all data points (or pixels) within the boundaries of the high intensity spot. The location of the high intensity spot may be the geometric center of the high intensity spot boundary in reciprocal space.

[0042] Identifying the high intensity spots belonging to each diffraction peak may include grouping the high intensity spots based on their intensity and position. Diffraction peaks may be captured in multiple diffraction patterns. In one embodiment, the high intensity spots belonging to a diffraction peak may be identified based on the distance of the high intensity spots in reciprocal space. For example, high intensity spots within a threshold distance from each other in reciprocal space are determined to belong to a diffraction peak. The distance between the high intensity spots may be determined based on the position of the high intensity spots in the diffraction pattern and the tilt angle of the diffraction pattern. The high intensity spots of each diffraction peak may be further identified based on the total intensity of the high intensity spots. For example, the total intensity increases first and then decreases as the tilt angle increases.

[0043] At 606, the intensity of the diffraction peak in each diffraction pattern is calculated. For example, the intensity of a diffraction peak in a diffraction pattern (or at an oblique angle) is the total intensity of the high intensity spots of the diffraction peak within the diffraction pattern.

[0044] Fig. 7A The intensities of example diffraction peaks in a diffraction tilt series are shown. Fig. 7A , the diffraction tilt series 701 includes multiple diffraction patterns acquired at multiple tilt angles θ. An example high intensity spot is shown on each diffraction pattern 711, 712, 713 and 714. The high intensity spots in the diffraction patterns 711-714 are determined to belong to the same diffraction peak because they are located close to each other in reciprocal space. The gray level of the high intensity spot indicates the total intensity of the high intensity spot, or the intensity of the diffraction peak in the diffraction pattern. The darker the high intensity spot, the higher the total intensity. The total intensity increases first and then decreases as the tilt angle increases.

[0045] At 608, a rocking curve (or angular spot profile) for each diffraction peak may be generated based on the intensity of the diffraction peak in the plurality of diffraction patterns. For example, the rocking curve may be generated by fitting the intensity of the diffraction peak to a curve such as a Gaussian curve.

[0046] Figure 7B Shown based on Fig. 7A Example rocking curve 725 generated by the intensity of the diffraction peak with a tilt angle θ1 to θ2. Figure 7BIn , the x-axis is the tilt angle. The y-axis is the intensity of the diffraction peak, and the intensity increases as shown by the arrow. In this example, because the diffraction pattern is acquired while the sheet is continuously rotated, the total intensity of each high-intensity spot is in the form of a bar with a width of the tilt angle difference Δθ. The heights of bars 721, 722, 723, and 724 correspond to the total intensities of the high-intensity spots in diffraction patterns 711, 712, 713, and 714, respectively. The intensity of the diffraction peak is above zero from tilt angles θ1 to θ2, and is zero elsewhere. A rocking curve can be generated by fitting the intensity of the diffraction peak to a curve. For example, the total intensity of the high-intensity spot can be fitted to a Gaussian curve ,in is the maximum intensity of the diffraction peak, Is has The tilt angle of maximum intensity, and is the standard deviation of the Gaussian curve. Standard deviation The position of the diffraction peak in the reciprocal space can be determined based on the rocking curve. For example, the position of the diffraction peak in the reciprocal space can be determined based on the tilt angle (e.g. Figure 7B In ) and the location of the high intensity spot with maximum intensity (e.g. Fig. 7A The position of the diffraction peak in the diffraction pattern 713 is determined.

[0047] At 610, the amplitude of the structure factor can be determined based on the rocking curve. The amplitude of the structure factor at a spatial frequency k can be calculated by taking the square root of the rocking curve at the corresponding tilt angle θ'. That is, , where F(k) is the amplitude of the structure factor at spatial frequency k (i.e., the position of the diffraction peak) in reciprocal space, and I(θ') is the intensity of the rocking curve at tilt angle θ'. The tilt angle θ' can be calculated based on the spatial frequency k. Note that for absolute structure factor determination, appropriate integration and scaling of the rocking curve is required. For the present invention, only the relative intensity of the structure factor corresponding to Friedel pairing is of interest, so integration and scaling are not required.

[0048] In this way, the amplitude of the structure factor can be determined based on the rocking curve of the diffraction peak. The rocking curve is generated by fitting the intensity of the diffraction peak in the diffraction tilt series. The position of the diffraction peak in reciprocal space can be determined based on the position of its corresponding high intensity spot in the diffraction pattern. The intensity of the diffraction peak at a specific tilt angle can be the total intensity of the high intensity spot at the tilt angle.

[0049] Return to Figure 3At 308, an EM tilt series is acquired. The EM tilt series may contain both low-dose and high-dose EM images of the sample at multiple tilt angles. For example, at each tilt angle, a low-dose and a high-dose EM image are acquired at the same position of the sample. Figure 8 An example subroutine 800 for acquiring multiple EM images is shown.

[0050] At 802, the tilt angles of a plurality of EM images are determined. The range of tilt angles of the EM tilt series may be Figure 3 The range of tilt angles of the diffraction tilt series at 304 is the same. The spacing between EM tilt series, i.e. the tilt angle difference between consecutive EM images, can be different from the diffraction tilt series. In one embodiment, the spacing of the tilt angles used for EM imaging can be determined based on the tilt angle difference Δθ of the diffraction tilt series. The spacing of the tilt angles can further depend on the available imaging area.

[0051] At 804, the lamella sample may be shifted to a new position and tilted to one of the tilt angles determined at 802. The lamella may be shifted and tilted by operating the sample holder. In some embodiments, the new position may be a position on a different lamella. In one embodiment, the lamella may be tilted about the same tilt axis as the diffraction tilt series. In another embodiment, the lamella may be tilted about a different tilt axis than the diffraction tilt series. The tilt axis may be orthogonal to the incident beam axis.

[0052] At 806, a low dose rate EM image of the sample may be acquired. The low dose EM image may be acquired using a reduced current density. For example, the low dose EM image may be acquired using a current density less than 0.1 el / angstrom. 2 Low-dose EM images can be acquired with a current density of 0.01 el / angstrom. 2 Low-dose EM images contain Bragg reflections from the crystal lattice.

[0053] At 808, immediately after acquiring the low-dose EM image, a high-dose EM image of the sample may be acquired without adjusting the position of the sample. The high-dose EM image may be acquired using the current density acquired with the low-dose EM image. For example, the current density used for high-dose EM imaging may be 10-100 el / angstrom. 2 By irradiating a region of a thin-section sample with high-energy electrons, the molecular structure within the imaged region may be damaged. As such, high-dose EM images do not contain Bragg reflections from the lattice. From high-dose EM images, a contrast transfer function (CTF) can be derived that reflects the aberrations of the imaging system. Fig. 10B The Fourier transform of the low-dose EM image is shown. Fig. 10BThe Thon rings in the center of the image may be caused by aberrations such as defocus, spherical aberration and astigmatism of the imaging system. Fig. 10A Thorne rings are only poorly visible in EM images compared to the diffraction pattern of the low-dose EM image, which is acquired at the same tilt angle as the low-dose EM image. Therefore, high-dose EM images are required to calculate the CTF. Fig. 10A The diffraction peak in region 1001 is Fig. 10B corresponds to the high intensity signal in region 1002 .

[0054] At 810, the subroutine 800 determines whether EM images have been acquired at all tilt angles determined at 802. If all tilt angles have been imaged, the subroutine 800 ends. Otherwise, an EM image is acquired at another position and tilt angle.

[0055] In one embodiment, a diffraction tilt series is acquired while the sample is continuously tilted relative to the incident beam, and each of the plurality of EM images is acquired while the sample is stationary relative to the incident beam. Thus, each diffraction pattern corresponds to a range of tilt angles, and each EM image corresponds to a discrete tilt angle. The sample is held stationary during EM acquisition, as tilting the sample during EM acquisition may smear high-resolution information in these EM images and reduce resolution in the Fourier transformed EM images.

[0056] Return to Figure 3 At 310 , the phase of the structure factor may be determined based on the EM tilt series acquired at 308 and the intensity of the diffraction peaks in the diffraction pattern. Fig. 9 An example subroutine 900 for determining a structure factor phase is shown. Specifically, a 2D structure factor phase at a tilt angle can be solved based on a Fourier transform of a low-dose EM image, a CTF derived from a high-dose EM image released at a tilt angle, and the intensity of the Friedel pairs of the diffraction peaks at the tilt angle. The 2D structure factor phase at multiple tilt angles can then be reconstructed into a 3D data set of structure factor phases.

[0057] At 902, the high-dose and low-dose EM images acquired at the selected tilt angle from step 308 may be loaded into memory. The selected angle may be Figure 8 One of the tilt angles determined at 802.

[0058] At 904, a CTF may be derived from the high dose EM image. In one embodiment, the CTF may be derived based on a Fourier transform of the high dose EM image. The CTF may be further adjusted based on the tilt angle to account for non-uniform focus across the imaging region in a tilted sample such as a thin section.

[0059] At 906 , a structure factor phase (ie, 2D structural phase) at a selected tilt angle may be calculated based on the Fourier transform of the low-dose EM image, the CTF derived at 904 , and the intensity of the Friedel pairs of the diffraction peaks at the selected tilt angle.

[0060] For example, a single Fourier component of a low-dose EM image can be expressed as:

[0061] , Equation 1

[0062] in is obtained with a tilt angle θ, x is the position in real space, k is the spatial frequency in reciprocal space, and is the amplitude of the Friedel pair of the structure factor, is the CTF derived at 904, and is the phase of the structure factor at spatial frequency k. and Proportional to the square root of the intensity of the Friedel pair of diffraction peaks in the diffraction pattern. Figure 2B As shown in , the intensities of the Friedel pairs of diffraction peaks in the diffraction pattern may be different because the distances between the reflections and the Ewald sphere are not centrosymmetric with respect to the origin. Therefore, the structure factor, and The amplitudes of the Friedel pairs may be different. In one embodiment, the amplitude of the structure factor at any given spatial frequency can be obtained based on the rocking curve of the diffraction peak. For example, Figure 7B As shown in , the amplitude of the structure factor F(k') can be the square root of the rocking curve I(θ'). In another embodiment, without generating a rocking curve, It can be directly obtained from the intensity of the diffraction peak in the diffraction pattern. Figure 7B In the is set to the square root of the intensity 731. In another example, if the diffraction pattern and the EM image are taken at the same divergence angle θ', the amplitude of the structure factor can be set to the square root of the intensity of the diffraction peak in the diffraction pattern.

[0063] Equation 1 can be further simplified:

[0064] ,

[0065] Equation 2

[0066] in , ,

[0067] ,and . Thus, the structure factor is The phase at θ can be based on low-dose EM images The Fourier transform of the structure factor Friedel pair intensity and CTF are used to solve. The intensity difference of the structure factor Friedel pair (or the Friedel pair of the diffraction peak) causes Additional phase shift. When the strengths of the Friedel pairs are equal, that is, , , there is no additional phase shift.

[0068] At 908, a phase origin is determined in the 2D structure factor phase. Specifically, the 2D structure factor phase is adjusted to a common origin of the coordinate system in real space so that the 2D structure factor phases with different tilt angles can be combined to obtain a 3D phase of the structure factor. In one embodiment, the phase origin can be determined by requiring the 2D structure factor phases on the tilt axis to have the same phase. For example, if the EM image is acquired by tilting the sample about the y-axis in real space, the 2D structure factor phase can be adjusted so that the k-axis along the reciprocal space is the same as the 2D structure factor phase. y The phases of the reflections from the axis are the same.

[0069] At 910, the subroutine 900 checks whether all EM images have been processed. In other words, the subroutine 900 checks whether all tilt angles used for EM imaging have been processed. If the answer is no, another tilt angle is selected at 912, and the phase of the structure factor at the selected tilt angle is calculated. Otherwise, if all EM images have been processed, a 3D phase of the structure factor can be constructed based on the 2D structure factors at different angles at 914. Constructing the structure factor phase can include merging the 2D structure factor phase based on the tilt angle relative to the crystal axis.

[0070] Return to Figure 3 , at 312, the molecular structure can be reconstructed based on the amplitude of the structure factor determined at 306 and the phase of the structure factor determined at 310. The molecular structure can include atomic coordinates of the crystal in real space. In one example, the molecular structure can be determined based on the inverse Fourier transform of the 3D structure factor amplitude and the 3D structure factor phase. In some embodiments, the molecular structure can be further refined based on the acquired diffraction tilt series. For example, the atomic positions in the molecular structure can be fine-tuned to minimize the difference between the calculated diffraction pattern and the observed diffraction pattern. The calculated diffraction can be determined by the molecular structure solved by Fourier transform and the observed diffraction pattern from the diffraction tilt series.

[0071] In some embodiments, the EM tilt series can be acquired before acquiring the diffraction tilt series. In some embodiments, the tilt angle range in the diffraction tilt series or the EM tilt series can be reduced based on the known symmetry of the crystal. For example, a first diffraction tilt series covering a first tilt angle range can be acquired. A second diffraction tilt series covering a second tilt angle range greater than the first tilt angle range can be constructed based on the first diffraction tilt series and the known symmetry of the crystal.

[0072] In this way, the molecular structure of a 3D crystal can be solved based on a diffraction tilt series and an EM tilt series. The diffraction tilt series can include diffraction patterns acquired at a first tilt angle range, and the EM tilt series can include low-dose and high-dose EM images acquired at a second tilt angle range. The 3D structure factor amplitude can be determined based on the integral of the rocking curve over the tilt angle. The 3D structure factor phase can be solved based on the EM tilt series and the 3D structure factor amplitude. The molecular structure can be obtained based on the amplitude and phase of the structure factor.

[0073] The technical effect of acquiring a diffraction pattern is to obtain the amplitude of the structure factor. The technical effect of determining the amplitude of the structure factor based on the intensity of the diffraction peaks in multiple diffraction patterns is to capture the total intensity of the diffraction peaks. The technical effect of determining the phase of the 2D structure factor based on the intensity of the Friedel pairs of the diffraction peaks in the diffraction pattern is that the additional phase shift caused by the unequal intensity of the Friedel pairs can be considered. The technical effect of acquiring a high-dose EM image and a low-dose EM image at each tilt angle is that the CTF corresponding to the low-dose EM image can be determined based on the high-dose EM image.

[0074] In one embodiment, a method for determining a molecular structure of a sample comprises: acquiring a plurality of diffraction patterns of the sample while tilting the sample within a tilt angle range; determining an amplitude of a structure factor based on the intensity of diffraction peaks in the plurality of diffraction patterns; acquiring an electron microscope (EM) image of the sample; determining a phase of the structure factor based on the EM image and the intensity of the diffraction peaks in the plurality of diffraction patterns; and determining the molecular structure of the sample based on the amplitude of the structure factor and the phase of the structure factor. In a first instance of the method, after acquiring the EM image, a second EM image of the sample is acquired at a current density higher than the current density used to acquire the EM image, and the phase of the structure factor is further determined based on the second EM image. The second instance of the method optionally comprises the first instance, and further comprises determining a contrast transfer function (CTF) based on the second EM image, and determining the phase of the structure factor based on the second EM image comprises determining the phase of the structure factor based on the CTF. The third example of the method optionally includes one or more of the first example and the second example, and further includes: wherein acquiring the EM image includes acquiring the EM image at a certain tilt angle within a tilt angle range, and determining the phase of the structure factor based on the EM image and the intensity of the diffraction peak in the plurality of diffraction patterns includes determining the phase of the structure factor at the tilt angle based on the EM image and the intensity of the diffraction peak at the tilt angle. The fourth example of the method optionally includes one or more of the first to third examples, and further includes wherein the phase of the structure factor at the tilt angle is further determined based on the intensity of the Friedel pairing of the diffraction peak. The fifth example of the method optionally includes one or more of the first to fourth examples, and further includes wherein the intensity of the diffraction peak and the intensity of the Friedel pairing of the diffraction peak are different. The sixth example of the method optionally includes one or more of the first to fifth examples, and further includes: wherein determining the amplitude of the structure factor based on the intensity of the diffraction peak includes calculating the amplitude of the structure factor at a specific tilt angle by taking the square root of the intensity of the diffraction peak at the tilt angle. A seventh example of the method optionally includes one or more of the first to sixth examples, and further includes: wherein determining the amplitude of the structure factor based on the intensity of the diffraction peak includes fitting the intensity of the diffraction peak to a curve, and calculating the amplitude of the structure factor at a specific tilt angle by taking the square root of the intensity of the curve at the tilt angle.The eighth example of the method optionally includes one or more of the first to seventh examples, and further includes identifying the diffraction peak by thresholding the plurality of diffraction patterns. The ninth example of the method optionally includes one or more of the first to eighth examples, and further includes wherein the intensity of the diffraction peak in the plurality of diffraction patterns is non-zero. The tenth example of the method optionally includes one or more of the first to ninth examples, and further includes wherein the sample is a thin sheet.

[0075] In another embodiment, a non-transitory computer readable medium includes instructions that, when executed, cause a processor to perform the following: acquiring a plurality of diffraction patterns of the sample while tilting the sample within a range of tilt angles; determining an amplitude of a structure factor based on the intensity of each diffraction peak in the plurality of diffraction patterns; acquiring a plurality of electron microscope (EM) images of the sample at a plurality of tilt angles; determining a phase of the structure factor based on the plurality of EM images and the intensity of the diffraction peaks; and determining a molecular structure of the sample based on the amplitude of the structure factor and the phase of the structure factor. In a first example, wherein determining the phase of the structure factor based on the plurality of EM images and the intensity of the diffraction peaks includes determining the phase of the structure factor based on the plurality of EM images and the intensity of Friedel pairs of the diffraction peaks. A second example of the non-transitory computer-readable medium optionally includes the first example, and further includes wherein determining the phase of the structure factor based on the multiple EM images and the intensity of the diffraction peaks includes: determining the intensity of the Friedel pairs of the structure factor based on the intensity of the diffraction peaks, and determining the phase of the structure factor based on the multiple EM images and the intensity of the Friedel pairs of the structure factor. A third example of the non-transitory computer-readable medium optionally includes one or more of the first example and the second example, and further includes: instructions that, when executed, cause the processor to acquire a plurality of diffraction patterns at a first position of the sample, and acquire a plurality of EM images at a second position of the sample.

[0076] In yet another embodiment, a system for determining a molecular structure of a sample comprises: an electron source for generating an electron beam; a sample holder for tilting the sample relative to the electron beam; a first detector for acquiring a diffraction pattern of the sample; a second detector for acquiring an electron microscope (EM) image of the sample; a controller having computer-readable instructions stored in a non-transitory medium, the controller being configured to: tilt the sample within a tilt angle range through the sample holder; direct the electron beam toward a first position of the sample when the sample is tilted, and acquire a plurality of electron diffraction patterns of the sample through the first detector; determine an amplitude of a structure factor based on the intensities of diffraction peaks in the plurality of diffraction patterns; direct the electron beam to a second position of the sample, and acquire a low-dose EM image of the sample through the second detector; determine a phase of the structure factor based on the low-dose EM image and the intensities of the diffraction peaks; and determine the molecular structure of the sample based on the amplitude of the structure factor and the phase of the structure factor. In a first example of the system, the controller is further configured to: position the sample at a first tilt angle before directing the electron beam to the second position of the sample and acquiring the low-dose EM image; acquire a high-dose EM image at the second position of the sample after acquiring the low-dose EM image; and wherein determining the phase of the structure factor based on the low-dose EM image and the intensity of the diffraction peak includes: determining the phase of the structure factor at the first tilt angle based on the low-dose EM image, the intensity of the diffraction peak, and the high-dose EM image. A second example of the system optionally includes the first example, and further includes wherein the controller is further configured to tilt the sample along two orthogonal axes, and acquiring the plurality of diffraction patterns includes: acquiring a first set of diffraction patterns while tilting the sample along one of the two orthogonal axes, and acquiring a second set of diffraction patterns while tilting the sample along the other of the two orthogonal axes.A third example of the system optionally includes one or more of the first example and the second example, and further includes wherein the controller is further configured to: after acquiring the high-dose EM image, position the sample at a second tilt angle; direct the electron beam to a third position of the sample positioned at the second tilt angle, and sequentially acquire a second low-dose EM image and a second high-dose EM image of the sample; determine the phase of the structure factor at the second tilt angle based on the second low-dose EM image, the second high-dose EM image, and the intensity of the diffraction peak; and wherein determining the molecular structure based on the amplitude of the structure factor and the phase of the structure factor includes: determining the molecular structure based on the amplitude of the structure factor, the phase of the structure factor at the first tilt angle, and the phase of the structure factor at the second tilt angle. A fourth example of the system optionally includes one or more of the first example and the third example, and further includes wherein the controller is configured to acquire the plurality of diffraction patterns while continuously tilting the sample by the sample holder.

Claims

1. A method for determining the molecular structure of a sample, comprising: acquiring a plurality of diffraction patterns of the sample while tilting the sample relative to an incident beam; determining an amplitude of a structure factor based on the intensities of diffraction peaks in the plurality of diffraction patterns; acquiring a first electron microscope (EM) image of the sample; After acquiring the first EM image, acquiring a second EM image of the sample at a current density higher than the current density used to acquire the first EM image; determining a phase of the structure factor based on the first EM image, the second EM image, and the intensities of the diffraction peaks in the plurality of diffraction patterns; as well as The molecular structure of the sample is determined based on the amplitude of the structure factor and the phase of the structure factor.

2. The method of claim 1, further comprising determining a contrast transfer function (CTF) based on the second EM image, and determining the phase of the structure factor based on the second EM image comprises determining the phase of the structure factor based on the CTF.

3. The method according to any one of claims 1 to 2, wherein acquiring the first EM image includes acquiring the first EM image at a certain tilt angle, and determining the phase of the structure factor based on the first EM image, the second EM image, and the intensities of the diffraction peaks in the multiple diffraction patterns includes determining the phase of the structure factor at the tilt angle based on the first EM image and the intensities of the diffraction peaks at the tilt angle. 4 . The method according to claim 3 , wherein the phase of the structure factor at the tilt angle is further determined based on the intensity of a Friedel mate of the diffraction peak. The method of claim 4 , wherein the intensity of the diffraction peak is different from the intensity of the Friedel pairing of the diffraction peak.

6. The method of any one of claims 1 to 2, wherein determining the amplitude of the structure factor based on the intensity of the diffraction peak comprises calculating the amplitude of the structure factor at a specific tilt angle by taking the square root of the intensity of the diffraction peak at the tilt angle.

7. The method according to any one of claims 1 to 2, wherein determining the amplitude of the structure factor based on the intensity of the diffraction peak comprises: fitting the intensity of the diffraction peak to a curve, and calculating the amplitude of the structure factor at a specific tilt angle by taking the square root of the intensity of the curve at the tilt angle.

8. The method of any one of claims 1-2, further comprising identifying the diffraction peaks by thresholding the plurality of diffraction patterns.

9. The method according to any one of claims 1 to 2, wherein the intensity of the diffraction peaks in the plurality of diffraction patterns is non-zero.

10. A system for determining the molecular structure of a sample, comprising: An electron source, the electron source being used to generate an electron beam; a sample holder for tilting the sample relative to the electron beam; a first detector, the first detector being used to obtain a diffraction pattern of the sample; a second detector, the second detector being used to obtain an electron microscope (EM) image of the sample; a controller having computer-readable instructions stored in a non-transitory medium, the controller configured to: direct the electron beam toward a first position of the sample while the sample is tilted relative to the electron beam by the sample holder, and acquire a plurality of electron diffraction patterns of the sample by the first detector; determining an amplitude of a structure factor based on the intensities of diffraction peaks in the plurality of diffraction patterns; directing the electron beam to a second location of the sample and acquiring a low-dose EM image of the sample via the second detector; determining a phase of the structure factor based on the low-dose EM image and the intensity of the diffraction peak; and The molecular structure of the sample is determined based on the amplitude of the structure factor and the phase of the structure factor.

11. The system of claim 10, wherein the controller is further configured to: before directing the electron beam to the second position of the sample and acquiring the low-dose EM image, positioning the sample at a first tilt angle; After acquiring the low-dose EM image, acquiring a high-dose EM image at the second position of the sample; and Wherein determining the phase of the structure factor based on the low-dose EM image and the intensity of the diffraction peak includes determining the phase of the structure factor at the first tilt angle based on the low-dose EM image, the intensity of the diffraction peak and the high-dose EM image.

12. A system according to any one of claims 10 to 11, wherein the controller is further configured to tilt the sample along two orthogonal axes, and acquiring the multiple diffraction patterns comprises: acquiring a first set of diffraction patterns while tilting the sample along one of the two orthogonal axes, and acquiring a second set of diffraction patterns while tilting the sample along the other of the two orthogonal axes.

13. The system of claim 11, wherein the controller is further configured to: After acquiring the high-dose EM image, positioning the sample at a second tilt angle; directing the electron beam to a third position of the sample positioned at the second tilt angle, and sequentially acquiring a second low-dose EM image and a second high-dose EM image of the sample; determining the phase of the structure factor at the second tilt angle based on the second low-dose EM image, the second high-dose EM image, and the intensity of the diffraction peak; and Wherein determining the molecular structure based on the amplitude of the structure factor and the phase of the structure factor comprises determining the molecular structure based on the amplitude of the structure factor, the phase of the structure factor at the first tilt angle, and the phase of the structure factor at the second tilt angle.

14. The system of any one of claims 10 to 11 and 13, wherein the controller is configured to acquire the plurality of diffraction patterns while continuously tilting the sample by the sample holder.