Archiviertes Physik-Kolloquium:

03. Nov. 2011, 16:15 Uhr, Gebäude NW1, Raum H3

Atomic imaging in 2D and in 3D

Prof. G. van Tendeloo, Uni Antwerpen

Ernst Ruska built his first electron microscope in 1931 and the first atomic resolution images date from the beginning of the seventies, but it is only since the introduction of lens aberration correcting systems a decennium ago, that “reliable” high resolution images have become available on a routine basis. With modern corrected microscopes a resolution in TEM or STEM below 0.1 nm has become more or less standard. However one should never forget that these images, no matter how beautiful and appealing, are only a two dimensional (2D) projection of a three dimensional (3D) reality. Moreover they are function of the transfer of the microscope and the interaction between the sample and the electron beam. Imaging nanomaterials in 3D has been a wet dream for most materials scientists and electron microscopist. However, so far, imaging in 3D is possible using “classical” electron tomography, but the resolution is hampered by several practical and fundamental obstacles. A resolution of about one nanometer is already a big success, but unfortunately this is not sufficient to obtain atomic resolution. Making use of 2D STEM imaging, where the intensity of the projected atom columns is proportional to the atomic number Z and the number of atoms in the column, we have been able to determine the 3D structure of nanoparticles at an atomic scale and to quantify the number of atoms within a nanoparticle with very high precision. In order to do so, we have used STEM information along several viewing directions. The resulting images have been treated using model based statistical parameter estimation theory and discrete tomography. Recent results show that the method also works for more challenging structures including free-standing nanoclusters. Another important aspect in materials science is the exact nature of interfaces in inorganic materials and multilayers; particularly the ability to determine atom diffusion on a local scale. It is well known that STEM images show Z-contrast allowing one to visually distinguish between chemically different atomic column types. However, if the difference in atomic number is small or if the signal-to-noise ratio is poor, direct interpretation of HAADF STEM images is inadequate. A performance measure which is sensitive to the chemical composition is the total intensity of scattered electrons. We will show that these intensities can be quantified atomic column - by - atomic column and that differences in atomic number of only 3 can be identified.