Abstract
• Electron energy loss spectroscopy (EELS) requires background fitting and removal. • Scripts for background subtraction have been written in MATLAB. • The scripts can be applied to core, low, and ultra-low loss EELS. • Statistical information on the goodness-of-fit is given. • Several examples of background subtraction are presented in the main text. Electron energy-loss spectroscopy (EELS) is a technique that can give useful information on elemental composition and bonding environments. However in practice, the complexity of the background contributions, which can arise from multiple sources, can hamper the interpretation of the spectra. As a result, background removal is both an essential and difficult part of EELS analysis, especially during quantification of elemental composition. Typically, a power law is used to fit the background but this is often not suitable for many spectra such as in the low-loss region (< 50 eV) and when there are overlapping EELS edges. In this article, we present a series of scripts written in MATLAB v. R2019b that aims to provide statistical information on the model used to fit the background, allowing the user to determine the accuracy of background subtraction. The scripts were written for background subtraction of vibrational EELS in the ultralow-loss region near the zero-loss peak but can also be applied to other kinds of EEL spectra. The scripts can use a range of models for fitting, provided by the Curve Fitting Toolbox of MATLAB, and the user is able to precisely define the window for fitting as well as for edge integration. We demonstrate the advantages of using these scripts by comparing their background subtraction of example spectra to the most commonly used software, Gatan Microscopy Suite 3. The example spectra include those containing multiple scattering, multiple overlapping peaks, as well as vibrational EELS. Additionally, a comprehensive guide to using the scripts has been included in the Supplementary Information.
Highlights
Incident electrons passing through a thin volume of matter in an electron microscope lose energy due to a variety of interactions that are dependent on the nature of the sample
energy-loss spectroscopy (EELS) can for instance probe sample thickness using the integrated intensity of the zero-loss peak relative to the total integrated intensity of the spectrum [1], valence or conduction electron density using the plasmon peaks [2], elemental composition using coreloss edges, bonding and oxidation state using energy loss near edge structure (ELNES) [3], as well as band structure [4,5] and vibrational modes [6] using near zero-loss features
We have shown that the MATLAB scripts can be used to extract signals from a variety of spectra, ranging from core-loss to low and ultra-low loss EELS
Summary
Incident electrons passing through a thin volume of matter in an electron microscope lose energy due to a variety of interactions that are dependent on the nature of the sample. The resulting energy distribution of electrons can be measured using electron energy-loss spectroscopy (EELS), and used to study structure, chemical composition or bonding. The majority of electrons passing through a suitable sample lose little to no energy, resulting in a high intensity zero-loss peak (ZLP). This peak is often in practice asymmetric, especially in the case of a cold field emitter, and often exhibits a tail extending significantly beyond the full-width at half-maximum (FWHM) of the ZLP which must be subtracted from the low-loss region (< 50 eV) before any quantification can be made.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
