Electron crystallography has been proven to be effective for structure determination of nano- and micron-sized crystals. In the past few years, 3D electron diffraction (3DED) techniques were used for the structure solution of various types of complex structures such as zeolites, metal-organic frameworks (MOF) and pharmaceutical compounds. However, unlike X-ray crystallography, electron diffraction has not yet become an independent technique for a complete structure determination due to relatively poorer diffraction intensities and often powder X-ray diffraction data are used for structure validation and refinement.

Electron beam damage to the structures that are sensitive to high energy electrons and dynamical scattering are important factors to lead to the deviation of electron diffraction intensities from the squared amplitudes of the structure factors. In this thesis, we investigate various aspects around the 3D electron diffraction data quality and strategies for obtaining better data and structure models. We combined 3D electron diffraction methods and powder X-ray diffraction to determine the structure of an open-framework material and discussed the difficulties and limitations of electron diffraction for beam sensitive materials. Next, we illustrated the structure determination of a pharmaceutical compound, bismuth subgallate, using 3D electron diffraction. While severe beam damage and diffuse scattering were observed in the dataset collected with the conventional rotation electron diffraction (RED) method, the continuous rotation electron diffraction (cRED) method coupled with sample cooling significantly improved the data quality and made the structure solution possible. In order to better understand the potentials and limitations of the continuous rotation method, we collected multiple datasets from different crystals of a known structure and studied the data quality by evaluating the accuracy of the refined structure models. To tackle dynamical scattering in electron diffraction data, we explored a routine for structure refinement with dynamical intensity calculation using RED data from a known structure and discussed its potentials and limitations.