Quantitative Electron Diffraction In Two Dimensional Materials And Automatic Data Processing In Electron Microscope

Nanoscience has been very successful since its emergence in the last century,which is indispensable from rapid developments of electron microscope.Here we,based on quantitative electron microscopy,completed two research works with two-dimensional(2D)materials,which has drawn a lot of interest in the research community due to their novel physical properties.In the first part:having considered the important effects of layer number on 2D materials,we proposed a method,based on quantitative electron diffraction,to determine monolayer rhenium diselenide(ReSe2)and its vertical orientation,which plays an important role in determining the properties of anisotropic devices;In the second part:atomically resolved electron microscope images provided microstructure information at the atomic scale while lacking statistical information at a larger scale,to address the problem of processing data from two different scales with efficiency and accuracy,we developed an image processing algorithm,based on techniques in data science,to automatically acquire the information of atom distributions and local crystallography in atomically resolved images.In the current extensive studies of 2D materials,compared to the hexagonal ones,like graphene,hBN and MoS2,low symmetry 2D materials have shown great potential for applications in anisotropic devices.ReSe2 has the bulk space group P1 and belongs to triclinic crystal system with a deformed cadmium iodide type structure.Here we propose a quantitative electron diffraction based method to distinguish monolayer ReSe2 membrane from multilayer ReSe2,and its two different vertical orientations,our method could also be applicable to other low symmetry crystal systems,including both triclinic and monoclinic lattices,as long as their third unit-cell basis vectors are not perpendicular to their basal planes.Our experimental results confirmed the inferences from kinematical electron diffraction theory and multislice simulations.Generalization of our method to other 2D materials,like graphene,is also discussed.Developments of electron microscopy has pushed the limits of resolution to sub-atomic level,and atomically resolved structural and functional imaging are becoming more and more popular among research communities.With the increasing volumn of information of multi-dimensional spectral data,as well as the structural data of atom positions,types,and arrangements,data science is playing a more and more important role in post data processing.Atomically resolved scanning transmission electron microscopy(STEM)imaging directly visualizes atomic structures in 2D materials,particularly suitable for investigations into doping problems in transition metal dichalcogenides(TMDCs).To efficiently discover knowledge from these images,we have developed image processing routines to identify atom positions and classify their types in 2D alloys,based on which we further analyzed the local crystallography,and calculated alloying degree.