0000000001117876
AUTHOR
Sebastian Schlitt
Application of clustering techniques to electron-diffraction data: determination of unit-cell parameters.
A new approach to determining the unit-cell vectors from single-crystal diffraction data based on clustering analysis is proposed. The method uses the density-based clustering algorithm DBSCAN. Unit-cell determination through the clustering procedure is particularly useful for limited tilt sequences and noisy data, and therefore is optimal for single-crystal electron-diffraction automated diffraction tomography (ADT) data. The unit-cell determination of various materials from ADT data as well as single-crystal X-ray data is demonstrated.
Introduction to ADT/ADT3D
ADT approach was successfully used within the last years to solve crystal structure of nanocrystalline material ab-inito using programs initially written for X-ray diffraction data. The methodology of ADT data analysis including data extraction in a format suitable for established structure solution programs was developed. A collection of processing routines was implemented into dedicated software ADT3D. Here the main steps of ADT data processing are explained, and the complete sequence of steps from importing raw data to saving the hklI file is demonstrated.