Colin Ophus1
Lawrence Berkeley National Lab1
Colin Ophus1
Lawrence Berkeley National Lab1
The past decade of development for scanning transmission electron microscopy (STEM) have been enormously successful, driven primarily by three technological innovations: hardware aberration correction, direct electron detectors, and the transformation of S/TEM into a digital science with the rise of computational imaging. In this talk, I will show how these three developments have enabled four dimensional (4D)-STEM experiments where we record full 2D images of the diffracted STEM probe over a 2D grid of probe positions. I will demonstrate large field-of-view mapping of structure, orientation and strain for materials ranging from structural oxides nanowires to 2D materials. I will also show how these methods can also be combined with complementary techniques such as STEM-electron energy loss spectroscopy or x-ray spectro-ptychography, for multimodal characterization of energy materials. I will briefly describe how modern machine learning methods can be used to extend the range of usable sample thicknesses by untangling the complex nonlinear contrast generated from multiple scattering of the electron beam. Finally, I will show results from ptychographic atomic-resolution tomography, which we use to solve the structure of complex encapsulated nanotube-nanowire materials. All of our methods, algorithms, codes and datasets are freely available to the community in order to promote widespread development of these experimental procedures.