Search

CN-121978061-A - Reflection type coherent diffraction imaging method and system based on inclined phase correction

CN121978061ACN 121978061 ACN121978061 ACN 121978061ACN-121978061-A

Abstract

The application discloses a reflection type coherent diffraction imaging method and system based on inclined phase correction, and relates to the technical field of extreme ultraviolet optical imaging. The method comprises the steps of acquiring an extreme ultraviolet diffraction signal to obtain an original data matrix, acquiring an initial inclination angle value of a sample, executing a phase recovery iteration process, namely substituting a current inclination angle parameter into a coordinate mapping relation to calculate corrected frequency domain coordinates, determining a reflectivity compensation coefficient, carrying out coordinate resampling and amplitude correction on the original data to generate corrected data, calculating a reconstruction error and dynamically updating the inclination angle parameter, outputting a reconstructed image if the convergence condition is judged to be met, and otherwise, returning to execute again. According to the application, the inclination angle is used as a variable optimization variable to be embedded into an iterative closed loop, and the problems of geometric distortion and intensity artifact caused by sample inclination are effectively solved by combining extreme ultraviolet multilayer film reflectivity compensation, so that sub-pixel level angle correction precision and high-fidelity image reconstruction are realized, and imaging efficiency and stability are remarkably improved.

Inventors

  • HAO JINLIANG
  • WANG MIN

Assignees

  • 北京量紫瞬测科技有限公司

Dates

Publication Date
20260505
Application Date
20260320

Claims (10)

  1. 1. A reflective coherent diffraction imaging method based on tilt phase correction, comprising the steps of: Irradiating a sample to be detected by using an extreme ultraviolet light source with the wavelength of 20nm to 50nm, and collecting diffraction pattern signals of the sample to be detected through a reflection type optical path to obtain an original diffraction data matrix, wherein each pixel point in the original diffraction data matrix is provided with a position coordinate under a detector plane coordinate system; obtaining an initial inclination angle value of the sample to be tested ; Executing a phase recovery iteration flow until convergence conditions are met, and outputting a reconstructed complex amplitude image; Wherein the phase recovery iterative process comprises the following sub-steps: (a) Parameter of inclination angle Substituting the position coordinates into a preset coordinate mapping relation, calculating to obtain corrected frequency domain coordinates, and determining reflectivity compensation coefficients corresponding to the frequency domain coordinates; according to the corrected frequency domain coordinates and the reflectivity compensation coefficient, carrying out coordinate resampling and amplitude correction on the original diffraction data matrix to obtain corrected diffraction data; wherein the tilt angle parameter is the first time step (a) is performed Is the initial inclination angle value ; (B) Inputting the corrected diffraction data into a phase recovery algorithm unit, reconstructing a complex amplitude image of a sample, and calculating a reconstruction error of a current iteration step, wherein the reconstruction error represents the consistency degree between the frequency domain amplitude distribution of the reconstructed image and the corrected diffraction data, and dynamically updating the inclination angle parameter based on the reconstruction error ; (C) Judging the reconstruction error of the current iteration step or the inclination angle parameter Whether the convergence condition is satisfied; if the convergence condition is met, terminating iteration and outputting a reconstructed complex amplitude image; if the convergence condition is not satisfied, the updated inclination angle parameter is updated As a new tilt angle parameter, and returns to step (a).
  2. 2. The method of claim 1, wherein the coordinate mapping relationship is constructed as: In the middle of (a) , ) The position coordinates of the detector under the plane coordinate system are [ (] ) For the corresponding corrected frequency domain coordinates, For the wavelength of the incident light, And the distance from the sample to be measured to the detector is set.
  3. 3. The method according to claim 1, wherein the initial tilt angle value of the sample to be measured is obtained Comprising the following steps: Measuring an initial inclination angle value of the sample to be measured through a gesture detection module; or determining the initial inclination angle value of the sample to be detected according to the central symmetry deviation of the original diffraction data matrix.
  4. 4. The method of claim 1, wherein determining the reflectivity compensation coefficient corresponding to the frequency domain coordinate in the step (a) comprises the steps of: Establishing reflectivity data of an extreme ultraviolet multilayer film reflector under 20nm to 50nm wave bands The database stores reflectivity values corresponding to different inclination angles and incident light wavelengths; inquiring the reflectivity database according to the inclination angle parameter and the incident light wavelength to obtain a reflectivity value ; The reflectivity compensation coefficient is calculated as 。
  5. 5. The method of claim 4, wherein said performing coordinate resampling and amplitude correction of said original diffraction data matrix based on said corrected frequency domain coordinates (u, v) and said reflectivity compensation coefficients in step (a) comprises: Mapping the original amplitude distribution at the position coordinates (x, y) in the original diffraction data matrix to an integer grid determined by the corrected frequency domain coordinates (u, v) by adopting a bilinear interpolation or bicubic interpolation algorithm to obtain resampled amplitude values; Multiplying the resampled amplitude value by the reflectivity compensation coefficient to obtain a corrected amplitude value, and taking the corrected amplitude value as the corrected diffraction data.
  6. 6. The method of claim 1, wherein the dynamically updating the tilt angle parameter based on the reconstruction error in step (b) Comprising: In the phase recovery iterative process, monitoring the change trend of the reconstruction error, determining the adjustment direction of the inclination angle parameter for reducing the reconstruction error, and updating the inclination angle parameter along the adjustment direction.
  7. 7. The method of claim 6, wherein said updating said tilt angle parameter in said adjustment direction in step (b) A multi-scale step strategy is adopted, and the multi-scale step strategy comprises the following steps: At the initial stage of iteration, updating the inclination angle parameter along the adjustment direction with a first update step ; At the later stage of iteration, when the change rate of the reconstruction error is lower than a preset rate threshold value, switching to a second updating step length to the inclination angle parameter Fine tuning is carried out; wherein the value of the first update step is greater than the value of the second update step.
  8. 8. The method of claim 1, wherein the convergence condition in step (c) comprises any of the following: The reconstruction error of the current iteration step is smaller than a preset error threshold; The inclination angle parameter in a plurality of continuous iterations The variation of (2) is smaller than a preset angle threshold; The iteration number reaches a preset maximum iteration number.
  9. 9. A reflective coherent diffraction imaging system based on tilt phase correction, comprising: an extreme ultraviolet light source module configured to generate a coherent light beam having a wavelength of 20nm to 50 nm; The sample bearing module comprises a nano displacement table and a gesture detection module, and the gesture detection module is configured to monitor the gesture of a sample in real time; the reflection type optical module comprises an extreme ultraviolet multilayer film focusing lens group and is arranged between the extreme ultraviolet light source module and the sample bearing module; The detection module is configured to receive the reflected diffraction light and convert the reflected diffraction light into a digital signal; a process control module electrically connected to the detection module and the gesture detection module, respectively, configured to perform the method of any one of claims 1 to 8.
  10. 10. A computer readable storage medium, on which a computer program is stored, characterized in that the program, when being executed by a processor, implements the steps of the method according to any one of claims 1 to 8.

Description

Reflection type coherent diffraction imaging method and system based on inclined phase correction Technical Field The invention relates to the technical field of Extreme Ultraviolet (EUV) optical imaging and computational imaging, in particular to a reflection type coherent diffraction imaging method and system based on inclined phase correction. The present invention also relates to a computer-readable storage medium storing a program for implementing the above method. Background Coherent diffraction imaging (CDI, coherent Diffraction Imaging) is a lensless computational imaging technique that breaks through the resolution limitations of conventional lenses. The basic principle is that the complex amplitude image of the sample is reconstructed by recording the far field diffraction pattern intensity of the scattering or reflection of the sample and utilizing a phase recovery algorithm (such as a hybrid input-output algorithm, a stacked scanning imaging algorithm and the like). Because extreme ultraviolet light (Extreme Ultraviolet, EUV, wavelength is usually between 10nm and 50 nm) has the characteristics of short wavelength, high photon energy, strong interaction with substances and the like, the reflective CDI technology based on EUV has great application potential in the fields of nano-scale resolution detection, biological macromolecular structure analysis, extreme ultraviolet lithography mask defect detection and the like. In existing reflective CDI imaging systems, in order to obtain high quality diffraction signals, it is often required that the sample surface be strictly perpendicular to the optical axis or at a specific bragg reflection angle. However, during practical experiments, the sample tends to have a slight inclination angle with respect to the ideal position, limited by the mechanical accuracy of the sample stage, the mounting error of the sample, and the micro-vibration during scanning. This tilting can cause two serious physical problems: First, geometric distortion problems. Tilting of the sample can result in non-linear geometric deformation of the diffraction pattern acquired at the detection plane. If the deformed diffraction data are directly input into a traditional phase recovery algorithm, stretching, blurring and even artifact can occur to a reconstructed image, and the spatial resolution and positioning accuracy of the image are seriously reduced. Second, the problem of uneven reflectivity. In the EUV band of 20nm to 50nm, imaging systems typically employ multilayer film mirrors to direct and focus light. The reflectivity of the multilayer film mirror is extremely sensitive to the angle of incidence, and when the sample is tilted, the equivalent angle of incidence at different locations changes, resulting in a non-uniform distribution of reflectivity across the field of view. Such tilt-induced intensity modulation can be mistaken by the algorithm for an absorption or phase change of the sample itself, thereby introducing serious intensity artifacts in the reconstructed image, interfering with the determination of the true structure of the sample. Aiming at the problems, the prior art mainly adopts the following treatment modes, but has obvious limitations: Some existing schemes measure the inclination angle of a sample through an external sensor (such as a laser interferometer and a capacitance inclinometer), and then perform one-time coordinate transformation correction on the diffraction pattern by using a geometric optical formula. However, this method has two drawbacks, firstly, that the measurement accuracy of the external sensor is limited (usually on the order of 0.01 °), and that the micro-vibration error of the sample at the moment of exposure cannot be eliminated, resulting in incomplete correction, and secondly, that the method corrects only the geometrical distortion of coordinates, completely ignores the specific angle-dependent effect of reflectivity in the EUV band, and cannot eliminate intensity artifacts. Another part of the schemes try to optimize the tilt angle as an unknown parameter by means of a "grid search" during the phase recovery. A series of angle values are preset, reconstruction is carried out respectively, and the angle with the smallest error is selected as the optimal solution. Although the method can simultaneously optimize angles theoretically, the calculation complexity is extremely high, the calculation amount increases exponentially along with the improvement of the search precision, and the requirement of actual imaging on efficiency is difficult to meet. In addition, grid search is prone to be trapped in a locally optimal solution, and convergence stability is poor. The conventional algorithms such as the extended stack scan imaging engine (extended Ptychographical ITERATIVE ENGINE, EPIE) have certain fault tolerance, but do not embed the tilt angle as an explicit optimization variable into the iterative loop. In the face of large initia