Significant Oxygen Underestimation When Quantifying Barium-Doped SrTiO Layers by Atom Probe Tomography

Abstract

In this paper, the capability for quantifying the composition of Ba-doped SrTiO layers from an atom probe measurement was explored. Rutherford backscattering spectrometry and time-of-flight/energy elastic recoil detection were used to benchmark the composition where the amount of titanium was intentionally varied between samples. The atom probe results showed a significant divergence from the benchmarked composition. The cause was shown to be a significant oxygen underestimation (≳14 at%). The ratio between oxygen and titanium for the samples varied between 2.6 and 12.7, while those measured by atom probe tomography were lower and covered a narrower range between 1.4 and 1.7. This difference was found to be associated with the oxygen and titanium predominantly field evaporating together as a molecular ion. The evaporation fields and bonding chemistries determined showed inconsistencies for explaining the oxygen underestimation and ion species measured. The measured ion charge state was in excellent agreement with that predicted by the Kingham postionization theory. Only by considering the measured ion species, their evaporation fields, the coordination chemistry, the analysis conditions, and some recently reported density functional theory modeling for oxide field emission were we able to postulate a field emission and oxygen neutral desorption process that may explain our results.

Keywords: atom probe tomography; barium-doped strontium titanate; composition quantification; energy elastic recoil diffraction; oxygen underestimation; perovskites; rutherford backscattering spectroscopy.