Search

The thousand-fold increase in data-collection speed enabled by aberration-corrected optics allows us to overcome an electron microscopy paradox: how to obtain atomic-resolution chemical structure in individual nanoparticles yet record a statistically significant sample from an inhomogeneous population. This allowed us to map hundreds of Pt-Co nanoparticles to show atomic-scale elemental distributions across different stages of the catalyst aging in a proton-exchange-membrane fuel cell, and relate Pt-shell thickness to treatment, particle size, surface orientation, and ordering., (© 2011 American Chemical Society)

The short depth of focus of aberration-corrected scanning transmission electron microscopes (STEMs) could potentially enable 3D reconstruction of nanomaterials through acquisition of a through-focal series. However, the contrast transfer function of annular dark-field (ADF)-STEM depth sectioning has a missing-cone problem similar to that of tilt-series tomography. The elongation as a function of the probe-forming angle is found to be (square root of 3/2) x 1/alphamax. For existing aberration-corrected STEMs operated at optimal imaging conditions, the elongation factor for depth sectioning is larger than 30. This large elongation factor results in highly distorted shapes of 3D objects and unexpected artifacts due to the loss of information. Depth-sectioning experiments using a 33-mrad 100 keV C(5)-corrected aberration-corrected STEM demonstrate the elongation effect and the missing-cone problem in real and reciprocal space. The performance limits of different S/TEM-based imaging modes are compared. There is a missing cone of information for bright-field S/TEM, ADF-STEM, hollow-cone ADF-STEM and coherent scanning confocal electron microscopy (SCEM). Only incoherent SCEM fills the missing cone.