High-Spatial and High-Energy Resolution EELS Studies of Chemical and
Electronic Properties of Interfaces and Nanostructures
K.J. Dudeck
1
, M. Couillard
1
, N. Gauquelin
1
, L. Gunawan
1
, S. Hosseini Vajargah
1
, S. Lazar
1,2
, D.
Rossouw
1
, Y. Shao
1
, G. Zhu
1
, G.A. Botton
1
1 Department of Materials Science and Engineering and Canadian Centre for Electron Microscopy,
McMaster University, 1280 Main St. W., ON, Canada, L8S 4M1
2 FEI Electron Optics, 5600 KA Eindhoven, The Netherlands
The development of aberration-corrected transmission electron microscopes capable of high-energy
resolution Electron Energy Loss Spectroscopy (EELS) has enabled the study of the chemical
composition of materials at unprecedented spatial resolution together with the prospects of providing
new insight onto the electronic properties of such materials. Elemental maps demonstrating atomic
resolution, even at low-accelerating voltages, have been established recently [1,2]. Such experiments
open the path to the study of complex electronic structure effects occurring in materials such as
complex oxides, interfaces between films and their substrates and grain boundaries. Although there
is significant interest on such studies, there are still several challenges that need to be overcome
when detailed structural and electronic information is to be retrieved. This contribution presents
examples of applications of high-spatial and high-energy resolution EELS studies and the
quantitative analysis of the data in order to highlight information retrieved as well as the pitfalls and
the difficulties in the interpretation.
Experiments were carried out on an FEI Titan 80-300 Cubed equipped with aberration-correctors of
the probe-forming lens and imaging lens as well as a monochromator and a high-brightness source.
The system is located in an ultra-stable environment allowing the reliable quantitative measurement
of strain on scanning-transmission electron microscopy images[3]. Examples of EELS elemental
maps on grain boundaries in bicrystals of SrTiO
3
have shown the detection of depletion of Sr atoms
within one unit cell from the grain boundary (figure 1). Quantitative elemental maps have also been
processed with and without the use of multivariate statistical analysis methods in order to understand
the effect of signal processing parameters on the information content of images. Detailed analysis
using measurements of signal-to-noise ratios (SNR) have been performed and show that the
improvements in the quality of the images are related to the reduction of the variance in the SNR
values. Such measurements have been carried out with an in-house parallelized MATLAB code
capable of processing spectrum images. Automated extraction of the EELS signal based on
optimization of the SNR (such as shown figure 1d) and as a function of the number of components in
the statistical analysis has been developed [4]. The quantification of the spectrum-images and the
measurement of energies of the edges have shown that changes in the valence and core-level shift
have been detected within less than 1nm form the boundary centre consistent with the local change
in composition and local charge. In the case of interfaces in ferroelectric Bi
3.25
La
0.75
Ti
3
O
12
deposited
on SrRuO
3
/SrTiO
3
, we have shown that the apparently sharp interfaces, as assessed by high-angle
annular dark-field (HAADF) STEM, are, in fact, chemically diffuse with atomic steps on only one of
the cationic sublattices hence suggesting that the EELS mapping approach can even reveal changes
in the surface termination of substrates (figure 2) and thus provide key information on the growth
mechanisms in thin films and the origin of defects. Examples of the application of high-resolution
794
doi:10.1017/S1431927611004843
Microsc. Microanal. 17 (Suppl 2), 2011
© Microscopy Society of America 2011