RHEED Analysis
How does AtomCloud help me get more from my reflection high-energy electron diffraction (RHEED) data?
Last updated
How does AtomCloud help me get more from my reflection high-energy electron diffraction (RHEED) data?
Last updated
After receiving RHEED images or videos, AtomCloud immediately starts extracting and standardizing the physical diffraction features using our curated pipelines (see consolidated example literature). You can see the status of your files on the Data Management page, in the Data section. For RHEED data, the analysis pipeline is broken down into a few steps outlined below:
For data visualization see: Pattern Analysis and Timeseries and Transitions
If RHEED is captured while the stage is rotating, AtomCloud will automatically extract the rotational frequency up to the symmetrically equivalent patterns in the video.
Standardizes RHEED data across different detector configurations.
Data can be exported at this stage for independent analysis. Use the 'Export Processed Selected' button on the Data Management page to export the transformed files for the selected data.
From each individual RHEED frame, if diffraction features are present, a Fingerprint pattern is extracted and quantified.
Feature regions are segmented from the background using a series of models:
UNet for RHEED, adapted from Ref. 1 and fine-tuned on additional RHEED patterns. This model produces separate segmentations of RHEED streaks and RHEED spots.
Medsam 2, taken from Ref. 2.
Patterns from images and videos are treated the same.
There is no differentiation at this time between surface effects (streaks and spots) and bulk effects (Kikuchi lines). If Kikuchi effects are strongly observed, they may be included and merged with other diffraction features.
For each feature extracted, quantified metrics are measured, including:
Horizontal and vertical FWHM.
Relative position of the intensity maxima relative to the geometric center of the feature.
Eccentricity, total area, streak-classified area, spot-classified area.
Specular spot labeled using intensity and relative position.
Position of each feature relative to the specular spot.
Timeseries can be generated from video frames or a related sequence of images allowing further automated results that depend on the relationship and relative changes between two or more patterns, such as:
Lattice spacing evolution -- strain, surface reconstruction
Intensity oscillation period -- growth rate
Streak to spot Ratio -- growth mode
Transitions -- manifold changes detected in diffraction pattern
Fingerprint pattern metrics can be used to track the evolution of lattice spacing allowing for efficient monitoring of strain throughout processing.
Pattern spacing can also be used to identify changes in the surface reconstruction.
Timeseries with sufficient framerate provide a profile of growth rate from the intensity oscillation period of the pattern.
This value quantifies whether you are closer to an island-like growth mode (high streak/spot ratio) or a layer by layer growth mode (low streak/spot ratio)
Principal component analysis (PCA) for dimensionality reduction, and Clustering algorithms are applied to identify and group statistically similar frames within the RHEED timeseries. This clustering is done without knowledge of the frame sequence (time order) providing robust statistical grouping.
Clusters are used to identify changes in the growth phase as they signal significant changes in the diffraction pattern's evolution.
