Assessing electron beam sensitivity for SrTiO$_3$ and La$_{0.7}$Sr$_{0.3}$MnO$_3$ using electron energy loss spectroscopy

TL;DR

This study evaluates electron beam damage thresholds for SrTiO$_3$ and La$_{0.7}$Sr$_{0.3}$MnO$_3$ using EELS during STEM, revealing different sensitivities at 80 and 200 kV and highlighting potential misinterpretations in damage detection methods.

Contribution

It provides a refined approach to detect beam damage via EELS and compares damage thresholds at different voltages for two materials.

Findings

SrTiO$_3$ shows no damage at 80 kV but damages at 200 kV.

La$_{0.7}$Sr$_{0.3}$MnO$_3$ damages at both voltages.

STEM image contrast changes may not reliably indicate damage.

Abstract

Thresholds for beam damage have been assessed for La$_{0.7}$Sr$_{0.3}$MnO$_3$ and SrTiO$_3$ as a function of electron probe current and exposure time at 80 and 200 kV acceleration voltage. The materials were exposed to an intense electron probe by aberration corrected scanning transmission electron microscopy (STEM) with simultaneous acquisition of electron energy loss spectroscopy (EELS) data. Electron beam damage was identified by changes of the core loss fine structure after quantification by a refined and improved model based approach. At 200 kV acceleration voltage, damage in SrTiO$_3$ was identified by changes both in the EEL fine structure and by contrast changes in the STEM images. However, the changes in the STEM image contrast as introduced by minor damage can be difficult to detect under several common experimental conditions. No damage was observed in SrTiO$_3$ at 80 kV acceleration voltage, independent of probe current and exposure time. In La$_{0.7}$Sr$_{0.3}$MnO$_3$, beam damage was observed at both 80 and 200 kV acceleration voltages. This damage was observed by large changes in the EEL fine structure, but not by any detectable changes in the STEM images. The typical method to validate if damage has been introduced during acquisitions is to compare STEM images prior to and after spectroscopy. Quantifications in this work show that this method possibly can result in misinterpretation of beam damage as changes of material properties.