Investigation of three-dimensional grain-boundary structures in oxides through multiple-scattering analysis of spatially resolved electron-energy-loss spectra

1998 ◽  
Vol 58 (13) ◽  
pp. 8289-8300 ◽  
Author(s):  
N. D. Browning ◽  
H. O. Moltaji ◽  
J. P. Buban
1996 ◽  
Vol 466 ◽  
Author(s):  
H. Ichinose ◽  
Y. Zhang ◽  
Y. Ishida ◽  
K. Ito ◽  
M. Nakanose

ABSTRACTA new spatially resolved electron energy-loss spectrometry (EELS) method was introduced to obtain atomic structure information of grain boundaries in diamond thin films grown by chemical vapor deposition. The electron energy-loss spectra recorded from the grain boundary regions showed different feature near the energy loss corresponding to carbon ls-to-π* transition, as compared to the spectra recorded from neighboring crystalline regions. This difference was attributed to dangling bonds in atoms with planar three-fold coordination. A series of experiments are described in this paper that exclude any possible artifact in result interpretation.


1994 ◽  
Vol 332 ◽  
Author(s):  
Haral MÜllejans ◽  
J. Bruley ◽  
R. H. French ◽  
P. A. Morris

ABSTRACTValence electron energy-loss (EEL) spectroscopy in a dedicated scanning transmission electron microscope (STEM) has been used to study the Σ11 grain boundary in α-A12O3 in comparison with bulk α-A12O3. The interband transition strength was derived by Kramers-Kronig analysis and the electronic structure followed from quantitative critical point (CP) modelling. Thereby differences in the acquired spectra were related quantitatively to differences in the electronic structure at the grain boundary. The band gap at the boundary was slightly reduced and the ionicity increased. This work demonstrates for the first time that quantitative analysis of spatially resolved (SR) valence EEL spectra is possible. This represents a new avenue to electronic structure information from localized structures.


Author(s):  
C P Scott ◽  
A J Craven ◽  
C J Gilmore ◽  
A W Bowen

The normal method of background subtraction in quantitative EELS analysis involves fitting an expression of the form I=AE-r to an energy window preceding the edge of interest; E is energy loss, A and r are fitting parameters. The calculated fit is then extrapolated under the edge, allowing the required signal to be extracted. In the case where the characteristic energy loss is small (E < 100eV), the background does not approximate to this simple form. One cause of this is multiple scattering. Even if the effects of multiple scattering are removed by deconvolution, it is not clear that the background from the recovered single scattering distribution follows this simple form, and, in any case, deconvolution can introduce artefacts.The above difficulties are particularly severe in the case of Al-Li alloys, where the Li K edge at ~52eV overlaps the Al L2,3 edge at ~72eV, and sharp plasmon peaks occur at intervals of ~15eV in the low loss region. An alternative background fitting technique, based on the work of Zanchi et al, has been tested on spectra taken from pure Al films, with a view to extending the analysis to Al-Li alloys.


Author(s):  
Eckhard Quandt ◽  
Stephan laBarré ◽  
Andreas Hartmann ◽  
Heinz Niedrig

Due to the development of semiconductor detectors with high spatial resolution -- e.g. charge coupled devices (CCDs) or photodiode arrays (PDAs) -- the parallel detection of electron energy loss spectra (EELS) has become an important alternative to serial registration. Using parallel detection for recording of energy spectroscopic large angle convergent beam patterns (LACBPs) special selected scattering vectors and small detection apertures lead to very low intensities. Therefore the very sensitive direct irradiation of a cooled linear PDA instead of the common combination of scintillator, fibre optic, and semiconductor has been investigated. In order to obtain a sufficient energy resolution the spectra are optionally magnified by a quadrupole-lens system.The detector used is a Hamamatsu S2304-512Q linear PDA with 512 diodes and removed quartz-glas window. The sensor size is 13 μm ∗ 2.5 mm with an element spacing of 25 μm. Along with the dispersion of 3.5 μm/eV at 40 keV the maximum energy resolution is limited to about 7 eV, so that a magnification system should be attached for experiments requiring a better resolution.


Sign in / Sign up

Export Citation Format

Share Document