Progress and Perspectives for Atomic-Resolution Electron Microscopy

Abstract

The atomic-resolution capabilities of the electron microscope, which first became widely accessible in the 1980s, have had a major impact across many disciplines [1]. With carefully prepared samples and correct operating conditions, assisted by quantitative image recording and processing, image characteristics can be interpreted directly in terms of individual atomic columns. With improvements in hardware and special attention to microscope environment, resolving powers close to or exceeding the one Ångstrom (0.1nm) barrier have been achieved by the latest generation of high-voltage HREMs [2][3][4]. Similar performance levels have been reached by dedicated scanning and conventional medium-voltage instruments [5,6], while sub-Ångstrom electron microscopy to resolutions of close to 0.8 nm has since been achieved using exit-wave retrieval [7,8]. Meanwhile, after a relatively dormant period during much of the 1990s, the last several years have witnessed a veritable explosion of further instrumentation hardware, including the successful implementation of aberration correctors for both conventional [9] and scanning [10] TEMs, as well as the design of electron monochromators which provide reduced energy spread and hence improved source coherence [11]. These latest developments have generated great interest and enthusiasm within the microscopy community, as well as attracting much attention from the broader materials community, especially for potential applications related to the rapidly emerging fields of nanoscience and nanotechnology.All electron lenses suffer from performance-limiting aberrations that must be removed in order to attain theoretical resolution limits. Two-fold astigmatism and third-order axial coma can be corrected manually by an experienced operator but for improved reliability and greater accuracy, online computer control or 'autotuning' of the microscope based, for example, on automated diffractogram analysis is recommended [12]. With the advent of aberration correctors, digital acquisition via slowscan CCD cameras has become indispensable. Computer analysis and control of corrector power supplies is essential for determining and then correcting most aberrations up to fourth order, including three-fold objective lens astigmatism and spherical aberration [13,14]. In the case of the probe-forming lens of the STEM, correction of the aberrations is done using multiple quadrupoleoctopole elements following analysis of a Ronchigram or shadow image obtained from a thin amorphous film [10,13]. For conventional TEM imaging, aberration correction is achieved with a double hexapole system and utilizes a tilt series of diffractograms, again relying on the availability of a thin amorphous film [14]. Aberration correction can also be tackled using a posteriori methods, such as exit-wave reconstruction based on through-focal image-series [7,8]. Similar resolution enhancements have been achieved using atomic-resolution electron holography [15]. Chromatic aberration remains as a serious obstacle to achievement of d...

Investigating the optical properties of dislocations by scanning transmission electron microscopy

“…In the very same year, it was also demonstrated that two diffracted beams could be passed simultaneously through the microscope objective lens to form interference fringes, or lattice fringes, the first representative image of a crystal lattice. Over the years the resolution gradually improved and the contrast mechanisms became well understood, for reviews of these developments, see Smith (1997Smith ( , 2008. Correlating the image to atomic positions is not always simple since the relative phases of the diffracted beams depend on the thickness and microscope focus and aberrations, so the imaging of dislocation core structures relied on extensive image simulations, at least until the recent introduction of aberration correctors (Kisielowski et al 2006).…”