Wednesday sessions highlight EELS and EDX

On Wednesday morning I browsed some of the sessions...

The session on Quantitative Analysis at the Nanoscale chaired by Masahi Watanabe and Matthew Weyland featured nine talks on the use of EDS, EELS and HAADF in aberration-corrected electron microscopes to analyze materials.

Maria Varela and colleagues at Oak Ridge National Lab described their work on the mapping of inequivalent oxygen sites in complex oxides. They use a VG501 STEM and one of only four installed Nion UltraSTEM 100 instruments.

Maria gave an overview of the use of atomic resolution spectrum imaging using ADF to identify oxygen atoms in perovskites such as La and Ca manganates. Principal component analysis and simulations of k-shell EELS images down pseudocubic 110 axis clearly reveal columns of the three elements.

But, Maria asked, can we distinguish different O sites in the oxide? Her simulations of the fine structure at the Ok edge (2ev separation) say yes. And to prove it she showed dramatic images of the Ca/La-O columns and the Mn-O columns. The take home message?: you need simulations to quantify atomic resolution spectrum images.

In the session on Problem Solving using Micronanalysis in the Real World, chaired by Paul Carpenter, Jon Hangas and Stuart McKernan, there were seven talks on applications of EDS.

Nestor Zaluzec - he with the pork pie hat - delivered a masterclass on the tricks you can use to maximize the X-ray signal in energy-dispersive X-ray microanalysis. For thin films, - and some bulk specimens - X-ray intensity is related to four factors: 1. # atoms in sample; 2. ionization processes; 3. probe current; and 4. detector properties. Since most of us cannot change the first three Nestor said that we have to optimize the last factor which comprises the QE of the SDD or SiLi detector and the collection solid angle (SA). Given that 98% of emitted X-rays are never detected the SA is critical to improve the signal. When the SA is >pi sterradians you can collect over 50%. So Nestor showed us a few tricks to calculate the real distance of the detector from the sample using the formula lnI= ln(k.S) - 2ln(R).

Two other factors that affect intensity are the nature and construction of the window and collimator and these were exposed by Nestor's critical eye. Finally he showed that some Be windows are contaminated by steel during the Be rolling technique.