Part of the Oxford Instruments Group
4th November 2020 | Author: Dr Simon Burgess
EDS technology is developing ever more rapidly, with new hardware and software promising ever more sensitive and high-speed analysis.
This leads to a common question about the best parameters to use for acquisition and what conditions should be used to ensure accurate analysis. There is no right answer to this question as it depends on many factors, however, it's really easy to find out yourself and I always recommend this. All you need is a sample with the type of materials you want to analyse. A sample of known composition is best, but even if you are not sure of the composition you can use it to check what range of conditions you can use without affecting the results.
I often use a slag sample which contains large numbers of spinels of varying composition. Spinels are great because their composition is predictable. They have the formula AB2X4, where A are divalent cations (e.g. Iron, Magnesium, Manganese and Zinc), B are trivalent cations (e.g. Chromium and Aluminium) and X is oxygen. This means that in atomic % Oxygen should be 57.1%, Mg+Fe 14.3% and Al+Cr 28.6%. With this information I can take this sample to any system and check its ability to accurately analyse this type of material, giving me high confidence for the analysis of minerals, ceramics, steel inclusions etc.
Normally I start at relatively low count rate to check the system is working well. This means long process time and 20kV accelerating voltage. With the latest large area detector this still means throughputs of over 50,000cps, at beam currents of less than 1nA, and an acquisition time of about 10 seconds. Below is a typical standardless result collected under these conditions for two of the common spinels in this sample.
| O | Mg | Fe | A2+ | Al | Cr | B3+ | Si | Ti | Total | |
| Cr Spinel | 57.28 | 11.77 | 2.04 | 13.81 | 11.77 | 16.96 | 28.73 | 0.18 | 100 | |
| Al Spinel | 57.17 | 13.99 | 0.36 | 14.35 | 22.84 | 5.45 | 28.29 | 0.10 | 0.09 | 100 |
Table showing the standardless result for two common spinels. The summed values of cations A & B are represented in bold. Numbers are shown in Atomic %.
Confident the results are good I can now think about the conditions I actually want to use. With the latest Ultim detectors and software for correcting high count rate artefacts like pulse pile-up, it is routine to go much faster and achieve accurate results at fast mapping speeds. I know this because I commonly check it. So, let's increase the throughput 7 times to 320,000cps, using Process Time 2 – a set-up I may be using for AZtecLive. With the Ultim Max 100 detector, the beam current required should still be well below 10nA. Collecting a spectrum for quant analysis now takes less than 2 seconds and the results are shown below. The analysis totals have increased slightly but the contents of the A, B and X sites have hardly changed. Is this good enough? Well, that is a question for you to answer depending on your characterisation requirements.
| O | Mg | Fe | A2+ | Al | Cr | B3+ | Mn | Ti | Total | |
| Cr Spinel | 57.25 | 11.79 | 1.99 | 13.78 | 12.11 | 16.52 | 28.63 | 0.16 | 0.18 | 100 |
| Al Spinel | 57.14 | 14.02 | 0.28 | 14.30 | 23.09 | 5.47 | 28.56 | 100 |
Table showing the standardless result for two common spinels acquired under standard acquisition settings. The summed values of cations A & B are represented in bold. Numbers are shown in Atomic %.
Knowing the AutoID and Quantitative analysis is accurate gives me great confidence in my results. If the quant is good the system is giving stable results and the spectrum processing is working well for all types of analysis, including peak deconvolution mapping, phase mapping etc.
Using this simple method, you can test any conditions. For example, how are the results when you reconstruct a spectrum from a map or change the accelerating voltage to increase spatial resolution? Try it and take control of your analysis. If you have any questions regarding this method or AZtecLive, then please post them in the comment section below where I will get back to you.
We send out monthly newsletters keeping you up to date with our latest developments such as webinars, new application notes and product updates.
© Oxford Instruments 2024
