Sample Alignment via Diffraction Pattern Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for aligning crystalline samples in charged particle microscopes are inefficient, particularly for thin samples, as they rely on heuristic circle-fitting and dynamic diffraction models, which can fail due to dynamic diffraction effects and are not accurate enough for precise metrology, especially when sample misalignment is less than 1 degree or in thin samples like those used in the electronics industry.
Innovation Solution
A method that compares the sample diffraction pattern with reference patterns having known misalignments to determine the closest match, using a simplified kinematical model to generate reference patterns, which reduces computational burden and eliminates the need for extensive libraries, allowing for precise alignment of the sample zone axis to within 0.5 degrees without requiring multiple diffraction patterns or indexing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If circle-fitting method is used to align sample, then alignment can be achieved for some patterns, but the method fails for patterns with dynamic diffraction effects or when misalignment is less than 1 degree
Solution Approach 1:
The invention changes the approach from geometric circle-fitting to intensity-based pattern matching by correlating diffraction spot intensities with simulated patterns. This parameter change enables reliable alignment for thin samples with dynamic diffraction effects and small misalignments (<1 degree) that were previously unsuitable for circle-fitting methods.
Solution Approach 2:
The invention replaces the mechanical/geometric circle-fitting approach with a computational intensity correlation method. By substituting geometric shape analysis with intensity-based pattern recognition, the system achieves universal applicability across different diffraction patterns including those affected by dynamic diffraction.
2Measurement precision
If dynamic diffraction model is used, then theoretical accuracy can be achieved, but computational burden increases and extensive libraries are required
Solution Approach 1:
The invention uses a simplified kinematical diffraction model instead of complex dynamic diffraction simulations. This 'cheaper' computational approach generates sufficient reference patterns without requiring extensive pre-computed libraries, reducing computational burden while maintaining adequate alignment precision for thin samples.
Solution Approach 2:
The invention applies partial action by using only the necessary level of diffraction model complexity (kinematical rather than dynamic) sufficient for thin sample alignment. This avoids the excessive computational resources required by full dynamic diffraction models while achieving the needed measurement precision.
3Ease of operation
If circle-fitting routine is used, then alignment can be performed, but it is difficult to determine accuracy of tilt measurement
Solution Approach 1:
The invention introduces feedback through iterative intensity correlation, where the measured diffraction pattern is compared with simulated patterns at different tilt angles. This feedback mechanism enables determination of both the optimal alignment and the accuracy of tilt measurement, overcoming the limitation of circle-fitting routines.
4Measurement precision
If multiple diffraction patterns are collected for indexing, then alignment accuracy can be improved, but processing time increases
Solution Approach 1:
The invention segments the alignment problem into intensity correlation of individual diffraction spots rather than requiring complete indexing of multiple patterns. This segmentation approach achieves accurate alignment using a single diffraction pattern, eliminating the time loss associated with collecting and processing multiple patterns for indexing.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach enables accurate and efficient alignment of crystalline samples by correlating lists of diffraction spots rather than images, reducing computational complexity and achieving precise alignment without the need for extensive libraries or multiple diffraction patterns, making it suitable for thin samples and improving metrological accuracy.
Implementation Method 1
Because of the wave nature of electrons, electrons that pass through a crystalline sample interfere with each other, reinforcing the electron beam in some regions and cancelling the beam in other regions. The interference forms a diffraction pattern on the back focal plane of an objective lens positioned below the sample.
Data Source
AI summary
A method and apparatus are provided for aligning a sample in a charged particle beam system. The charged particle beam is directed toward the sample to obtain a sample diffraction pattern. The sample diffraction pattern is compared with reference diffraction patterns having known misalignments to determine which reference pattern most closely matches the sample pattern. The known alignment of the best-matching reference diffraction pattern is used to correct the tilt of the sample. The “patterns” compared can be lists of bright spots with corresponding intensities rather than images.