A Practical Method for Removing Plural Scattering from Core Edges in EELS

1981 ◽  
Vol 39 ◽  
pp. 196-197
Author(s):  
R.D. Leapman ◽  
C.R. Swyt
Keyword(s):  
Core Loss ◽  
Practical Method ◽  
Electron Excitation ◽  
Single Scattering ◽  
Low Loss ◽  
High Loss ◽  
The Core ◽  
Electron Energy Loss ◽  
Electron Excitations ◽  
Electron Energy Loss Spectra

Core edges in electron energy loss spectra are generally complicated by thickness effects involving plural inelastic scattering. The fine structure can be modified and errors can be caused in quantitation based on measured edge intensities. Sometimes plural scattering can confuse even identification of elements. In this paper we describe a practical method for eliminating these difficulties.We derive the single scattering distribution, assuming valence electron excitation is small in the core loss region, a reasonable approximation for edges above 100 or 200 eV and for thicknesses of a few hundred Å. We can then separate the spectrum into a “high loss” region H(E) consisting of core edges (less background) and a “low loss” region L(E) containing the zero loss peak, plasmons and one-electron excitations.

Background Fitting in the Low-Loss Region of Electron Energy Loss Spectra

1990 ◽  
Vol 48 (2) ◽  
pp. 56-57
Author(s):  
C P Scott ◽  
A J Craven ◽  
C J Gilmore ◽  
A W Bowen
Keyword(s):  
Energy Loss ◽  
Multiple Scattering ◽  
Simple Form ◽  
Single Scattering ◽  
Low Loss ◽  
Characteristic Energy ◽  
Normal Method ◽  
Fitting In ◽  
Electron Energy Loss ◽  
Electron Energy Loss Spectra

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.

The Thickness Dependence of Energy Loss Spectra

1982 ◽  
Vol 40 ◽  
pp. 486-487
Author(s):  
M. Sarikaya ◽  
P. Rez
Keyword(s):  
Energy Loss ◽  
Multiple Scattering ◽  
Core Loss ◽  
Single Scattering ◽  
Low Loss ◽  
High Loss ◽  
Loss Analysis ◽  
Scattering Spectrum ◽  
Exact Procedure ◽  
Quantitative Analyses

One factor limiting energy loss analysis is the effect of multiple scattering on core loss edge shapes. Multiple scattering distorts fine structures, leads to incorrect quantitative analyses and even affects analysis of extended fine structure (EXELFS). Two procedures for extracting the single scattering spectrum from a spectrum showing the effects of multiple scattering have been proposed. Johnson and Spence derive the single scattering profile by taking the logarithm of the fourier transformed spectrum. If all scattering angles are accepted by the spectrometer this is an exact procedure. Leapman and Swyt have had some success assuming that multiple scattering imposes low loss structure on the high loss part of the spectrum. It is of interest to know how stable these procedures are with thickness and whether the logarithmic deconvolution can be used in thicker specimens than the method described by Leapman and Swyt.

Fourier deconvolution of electron energy-loss spectra

1985 ◽  
Vol 398 (1815) ◽  
pp. 395-404 ◽  
Keyword(s):  
Fine Structure ◽  
Energy Loss ◽  
Electron Energy ◽  
Core Loss ◽  
Single Scattering ◽  
Electron Interaction ◽  
Edge Structure ◽  
Fourier Deconvolution ◽  
Electron Energy Loss ◽  
Electron Energy Loss Spectra

Electron energy-loss spectroscopy (e. e. l. s.) performed with an electron microscope can be used to obtain plasmon spectra, near-edge fine structure and extended electron energy-loss fine structure (ex. e. l. f. s.), as well as to do chemical analysis on truly microscopic samples. However, the very strength of the electron–electron interaction gives rise to significant and sometimes predominant plural scattering effects. To obtain consistent and reliable estimates of the single scattering distributions these effects must be accounted for. In this paper, two Fourier-transform deconvolution methods of removing plural scattering from electron energy-loss spectra and their implementation on a microcomputer, are discussed. Their relative advantages and limitations are considered together with examples of artefacts that may arise in both plasmon and core-loss spectra. As a result of deconvolution the sensitivity and accuracy of core-elemental analysis is enhanced but, more importantly, it is possible to obtain reproducible near-edge structure and plasmon spectra from relatively thick samples that would otherwise not be suitable for investigation by means of e. e. l. s.

Empirical Identification of Uranium Oxides and Fluorides Using Electron Energy-loss Spectroscopy in the Transmission Electron Microscope

1999 ◽  
Vol 5 (6) ◽  
pp. 437-444 ◽  
Author(s):  
Stephen B. Rice ◽  
Hazel H. Bales ◽  
John R. Roth ◽  
Allen L. Whiteside
Keyword(s):  
Energy Loss ◽  
Electron Energy ◽  
Electron Energy Loss Spectroscopy ◽  
Contaminated Soils ◽  
Core Loss ◽  
Uranium Oxides ◽  
Low Loss ◽  
Electron Dose ◽  
Core Losses ◽  
Electron Energy Loss

Abstract: A set of uranium compound particles relevant to contaminated soils and other environmental concerns surrounding uranium bioavailability were studied by electron energy-loss spectroscopy (EELS). Core-loss EELS results suggest that uranium 4+ compounds have an energy loss resolvable from 6+ compounds. Shoulders on the uranium O4,5 edge further distinguish UO2 from UF4. Low-loss characteristics distinguish carbon-free uranium oxide specimens on holey substrates. In the presence of carbon, correction techniques must be applied. Uranium oxides, fluorides, and minerals show a tendency toward reduction of uranium toward 4+ under the beam. The electron dose required to achieve the transformation from 6+ to 4+ is more severe than that usually required to obtain satisfactory spectra, but the possibility for reduction should be considered. The conditions for low-loss analysis need not be as vigorous as those for core losses, and can be done without altering the valence of most oxides.

Image-Spectroscopy: New Developments and Applications

1999 ◽  
Vol 5 (S2) ◽  
pp. 618-619 ◽  
Author(s):  
P.J. Thomas ◽  
P.A. Midgley
Keyword(s):  
Spectroscopic Analysis ◽  
Core Loss ◽  
Compositional Analysis ◽  
Single Scattering ◽  
Electron Spectroscopic Imaging ◽  
Low Loss ◽  
Step Size ◽  
New Developments ◽  
Analysis Techniques ◽  
Image Spectroscopy

The ability of modern TEMs to acquire a series of energy filtered images opens up new possibilities in energy loss compositional analysis. In particular, an electron spectroscopic imaging (ESI) series may be treated as a 2-D array of spectra whose resolution is dictated by the step size of the image series, as illustrated in Fig (a). This allows standard spectroscopic analysis techniques to be used on the extracted ‘image-spectra’, such as the removal of plural scattering by deconvolution. Examples of this are given in Fig (b) and (c), which show how Fourier-log and Fourier-ratio deconvolution can be used to recover the single scattering distribution (SSD) from both the low-loss and core-loss regions from a Cr specimen. A pure elemental sample is ideal for testing the validity of such analysis techniques for quantitative compositional mapping, and more details of this method will be published elsewhere. Further, for many simple metal systems, such as steels and alloys, and for simple semiconductors it is possible to model the plasmon contribution using a simple Drude-Lorentz model.

Low-loss Electron Energy-loss Spectroscopy and Dielectric Function of Biological and Geological Polymorphs of CaCO3

1999 ◽  
Vol 5 (5) ◽  
pp. 358-364 ◽  
Author(s):  
Kalpana S. Katti ◽  
Maoxu Qian ◽  
Daniel W. Frech ◽  
Mehmet Sarikaya
Keyword(s):  
Energy Loss ◽  
Electron Energy ◽  
Dielectric Function ◽  
Electronic Structures ◽  
Microstructural Characterization ◽  
Low Loss ◽  
Composite Microstructure ◽  
Abalone Shell ◽  
Electron Energy Loss ◽  
Electron Excitations

Previous work on microstructural characterization has shown variations in terms of defects and organization of nanostructures in the two polymorphs of calcium carbonate, calcite, and aragonite in mollusc shells. Large variations in mechanical properties are observed between these sections which have been attributed to variations in composite microstructure as well as intrinsic properties of the inorganic phases. Here we present local low-loss electron energy-loss spectroscopic (EELS) study of calcitic and aragonitic regions of abalone shell that were compared to geological (single-crystal) counterpart polymorphs to reveal intrinsic differences that could be related to organismal effects in biomineralization. In both sets of samples, local dielectric function is computed using Kramer-Kronig analysis. The electronic structures of biogenic and geological calcitic materials are not significantly different. On the other hand, electronic structure of biogenic aragonite is remarkably different from that of geological aragonite. This difference is attributed to the increased contribution from single electron excitations in biogenic aragonite as compared to that of geological aragonite. Implications of these changes are discussed in the context of macromolecular involvement in the making of the microstructures and properties in biogenic phases.

Highly Automated Electron Energy-Loss Spectroscopy Elemental Quantification

2014 ◽  
Vol 20 (3) ◽  
pp. 798-806 ◽  
Author(s):  
Raman D. Narayan ◽  
J. K. Weiss ◽  
Peter Rez
Keyword(s):  
User Interface ◽  
Energy Loss ◽  
Electron Energy ◽  
Electron Energy Loss Spectroscopy ◽  
Single Scattering ◽  
Ease Of Use ◽  
Low Loss ◽  
Wide Range ◽  
Fitting Algorithm ◽  
Electron Energy Loss

AbstractA model-based fitting algorithm for electron energy-loss spectroscopy spectra is introduced, along with an intuitive user-interface. As with Verbeeck & Van Aert, the measured spectrum, rather than the single scattering distribution, is fit over a wide range. An approximation is developed that allows for accurate modeling while maintaining linearity in the parameters that represent elemental composition. Also, a method is given for generating a model for the low-loss background that incorporates plural scattering. Operation of the user-interface is described to demonstrate the ease of use that allows even nonexpert users to quickly obtain elemental analysis results.

Mapping Chemical Bonds in Semiconductor Devices by Monitoring the Shifts of EELS Edges

2017 ◽  
Vol 23 (5) ◽  
pp. 926-931 ◽  
Author(s):  
Pavel Potapov ◽  
Elena L. Svistunova ◽  
Alexander A. Gulyaev
Keyword(s):  
Scanning Transmission Electron Microscopy ◽  
Core Loss ◽  
Least Square ◽  
Reference Spectrum ◽  
Low Loss ◽  
Edge Structure ◽  
Simultaneous Acquisition ◽  
Transmission Electron ◽  
Scanning Transmission ◽  
Electron Energy Loss

AbstractScanning transmission electron microscopy (STEM) in combination with electron energy-loss spectroscopy (EELS) can deliver information about variations of bonding at the nm scale. This is typically performed by analyzing the electron-loss near edge structure (ELNES) of given EELS edges. The present paper demonstrates an alternative way of a bonding examination through monitoring the EELS onset positions. Two conditions are essential for their accurate measurement. One (hardware) is using the dual EELS instrumentation that provides near simultaneous acquisition of low-loss and core-loss spectra. Another (software) is the least-square fitting of observed spectra to a reference spectrum. The combination of these hardware and software techniques reveals the positions of EELS onsets with the precision sufficient for mapping tiny variations of bonding. The paper shows that the method is capable of helping to solve practical tasks of nanoscale engineering like the analysis of modern CMOS devices.

Sign in / Sign up

Export Citation Format

Share Document