Abstract
Electron microscopy is undergoing a transition; from the model of producing only a few micrographs, through the current state where many images and spectra can be digitally recorded, to a new mode where very large volumes of data (movies, ptychographic and multi-dimensional series) can be rapidly obtained. Here, we discuss the application of so-called “big-data” methods to high dimensional microscopy data, using unsupervised multivariate statistical techniques, in order to explore salient image features in a specific example of BiFeO3 domains. Remarkably, k-means clustering reveals domain differentiation despite the fact that the algorithm is purely statistical in nature and does not require any prior information regarding the material, any coexisting phases, or any differentiating structures. While this is a somewhat trivial case, this example signifies the extraction of useful physical and structural information without any prior bias regarding the sample or the instrumental modality. Further interpretation of these types of results may still require human intervention. However, the open nature of this algorithm and its wide availability, enable broad collaborations and exploratory work necessary to enable efficient data analysis in electron microscopy.
Highlights
Scanning transmission electron microscopy (STEM) and associated spectroscopies have emerged as powerful tools for the visualization of structure and functionality of materials at atomic resolution[1,2]
The camera and the microscope were integrated through a custom FPGA control system to synchronize frame capture and beam positioning to acquire 4D scanning-scattering data sets
We have utilized a capture rate of ~300 frames per second to give a total acquisition time approximately 1 minute for the entire 4D dataset. We note that this acquisition speed is comparable to STEM spectrum imaging and allows a transition to “full data” acquisition imaging in STEM using existing instrumental infrastructure once the associated data streaming pipelines and data analytic tools are established, similar to the approach recently demonstrated for scanning probe microscopy[28,29,30]
Summary
Ptychography received: 19 February 2016 accepted: 26 April 2016 Published: 23 May 2016. K-means clustering reveals domain differentiation despite the fact that the algorithm is purely statistical in nature and does not require any prior information regarding the material, any coexisting phases, or any differentiating structures While this is a somewhat trivial case, this example signifies the extraction of useful physical and structural information without any prior bias regarding the sample or the instrumental modality. The widespread implementation of aberration correction, and the associated increase in spatial resolution has allowed sub-50 pm resolution and, determination of atomic positions with sub-10 pm precision[3,4,5,6] These capabilities enable direct visualization of chemical and mechanical strains[7], order parameter fields including ferroelectric polarization[8,9,10,11], and octahedral tilts[12,13,14,15,16]. We deliberate on a roadmap for data streaming and storage for ptychographic imaging of complex materials
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
