Automated Electron Nanocrystallography
- PDF / 5,389,695 Bytes
- 10 Pages / 612 x 792 pts (letter) Page_size
- 86 Downloads / 235 Views
1026-C19-01
Automated Electron Nanocrystallography John Spence1, Joseph McKeown1, Haifeng He2, and Jinsong Wu3 1 Physics, Arizona State University, Tempe, AZ, 85287 2 NCEM, Lawrence Berkeley Laboratory, Cyclotron Rd, Berkeley, CA, 94720 3 Materials Science, Northwestern University, Evanston, IL, 60208 ABSTRACT Methods for solving crystalline nanostructures in real time at the electron microscope are reviewed, based on automated collection of microdiffraction patterns in three dimensions. We compare Koehler mode “SAD” patterns, a new kinematic CBED mode, and our new precession electron diffraction system. We discuss the optimum data-collection strategy and the eucentrictilt problem. We advocate use of the new “charge-flipping” algorithm for solving the phase problem when dealing with relatively poor-quality electron diffraction data. We show an experimental demonstration of the effect of precession on the quality of diffraction data and of the use of the flipping algorithm to solve a nanocrystal. INTRODUCTION The many new crystalline nanostructures being synthesized has generated an urgent demand for a rapid method of solving the crystal structure of a single, isolated nanocrystal. It would be desirable if this could be completed in real time at the electron microscope, based on automated collection of three-dimensional electron microdiffraction data followed by the phasing of the diffraction data and presentation of a charge-density or potential map showing atom positions and perhaps types. As one example of a nanostructural synthesis method that requires a rapid technique to solve extremely small, individual crystallites is a new combinatorial method for evaluating all binary and ternary nanocrystals that may be candidates for hydrolysis of water under sunlight [1]. In the future, using a “nanolab in a TEM,” it may be possible to perform both the synthesis and characterization within the same in-situ TEM environment, following the same methods on which the discovery of the carbon nanotube was based. The thermodynamic conditions needed to produce interesting new structures observed at atomic resolution in the TEM could then be recorded at the same time the structures are examined. The characterization would include X-ray and energy-loss data, which we assume are available at the electron microscope, to give approximate estimates of stoichiometry. The method would allow a much more detailed study of the variability of particles and the extent to which they are monodispersed.
Although multiple-scattering artifacts are expected to be weak for light-element nanoparticles less than about 20 nm in thickness and nanometer-diameter electron probes are readily obtained, a number of severe challenges exist for this project. Perhaps the most important of these is the lateral displacement that occurs during tilting of the sample, which we address here by using a hybrid STEM mode for data collection and suggest the use of a three-axis holder, which reduces the crystallographic search to one dimension. We use the new flipping
Data Loading...
