EDS Analysis in APEX EDS Software for Light and Heavy Elements

Gatan recently reported on improvements in quantitative analysis achieved with energy dispersive X-ray spectroscopy (EDS) measurements completed in the EDAX APEX^™ 3.0 EDS Standard and Advanced software suites with the use of the normalized standardless eZAF correction.¹

Using the analysis of over 60 certified standards as a baseline, the report shows a 3x improvement in standardless quantitative evaluation accuracy using APEX 3.0. That being said, relying on one figure of merit to comprehend the analytical algorithm performance capabilities used when quantitatively evaluating EDS spectra is challenging and nearly impossible. This article explores the analytical accuracy of APEX EDS Advanced software for samples with a mean atomic number range between 6 and 70.

For an extended period, microanalysts have accepted the wisdom that standardless evaluation offers acceptable results in numerous applications where the elements identified within the sample are of atomic number 12 and greater.

In these samples, the self-absorption of X-rays generated in the sample, the effects of changes in excitation efficiency, and secondary X-ray fluorescence for each element present (typically referred to as ZAF corrections) are understood with a reasonable amount of certainty. However, where the materials consist of or contain low-Z elements (Z <12), standardless analytical methods are used with great caution because of the potential for significant errors in analytical measurements due to uncertainty in the applied correction values.

Results and Discussion

A quantitative analysis of 34 alloy and compound certified standards received from MAC Micro-Analysis Consultants Ltd was performed. These materials were comprised of 50% that contained elements with atomic number <12, including boron nitride, silicon carbide, lanthanum hexaboride, and 12 oxides, such as quartz, albite, almandine garnet, magnesium oxide, and yttrium aluminum garnet.

The EDS spectra were captured at a 10,000 cps count rate and a live time of 40 s. The mean atomic number was calculated utilizing the modified electron fraction model.² The absolute error per element was determined for each analysis through summation of the absolute difference between the measured and known composition for each element, i, identified in the sample (Equation 1).

Equation 1

Figure 1 summarizes the variety seen in sum absolute error for each element as a function of mean atomic number. According to commonly understood wisdom, samples without light elements (blue circles), had their elemental composition determined with great accuracy—the mean error per element was only 1.0 at. % and 13 of the 17 samples analyzed showed a negligible error (<1 at. %). However, exceptional results were also arrived at for the samples containing one (or more) light elements (orange triangles); the mean error was slightly greater at 1.4 at. % and 11 of the 17 samples analyzed showed a negligible error. Furthermore, the 12 oxide samples showed a mean error of just 1.2 at. %—a close to 3x improvement compared to APEX 2.5 EDS Advanced. The incredible accuracy of standardless normalized analysis using eZAF correction in APEX 3.0 EDS Advanced makes standardless analysis a practical option for larger samples.

Figure 1. A plot of the error in quantitative evaluation of materials of certified composition using APEX 3.0 EDS Advanced using a Bremsstrahlung background model and correction for the thickness of the carbon coating. All composition values are provided in atomic percent. Image Credit: Gatan, Inc.

Figure 2 indicates a plot of the absolute error per element compared to the variations in atomic number (calculated from DZ = Zmax—Zmin). It shows that the analytical results error is greatest for samples of widely different elements.

This is mainly because of the large—and potentially uncertain—correction factors for secondary fluorescence and absorptions in materials with wide differences.

A large correction can affect any calculation of the net counts, which means that correction factors or uncertainty in the net counts gravitate toward an analytical result with bigger errors. That said, even in the examples of lanthanum hexaboride (DZ = 52) and thallium bromide iodide (DZ = 76), the error per element was just 0.3 and 2.1 at. %, respectively

Figure 2. A plot of the error in quantitative evaluation of materials of certified composition as a function of the maximum difference in atomic number. They were analyzed using APEX 3.0 EDS Advanced software using a Bremsstrahlung background model and correction for the thickness of the carbon coating. Image Credit: Gatan, Inc.

The results show notable improvements in the degree of accuracy of compositional analysis through normalized standardless analysis with the use of eZAF correction in EDAX APEX 3.0 Advanced software versions for nearly all samples, particularly for samples containing one or more light elements. A 3x improvement was seen in oxide samples analyzed and >5x for other light elements containing, for example, carbon, boron, or fluorine.

Conclusion

These results showcase that excellent accuracy is achievable for light and heavy elements using normalized standardless analysis with EDAX’s eZAF correction model using the APEX 3.0 EDS Advanced software. This means that standardless analysis becomes a practical option for a much larger array of materials than was previously thought.

References and Further Reading

1. EDAX. (2021). Evaluation of the accuracy of standardless EDS analysis in APEX EDS software. (online) Available at: https://www.edax.com/resources/application-notes/evaluation-of-the-accuracy-of-standardless-eds-analysis-in-apex-eds-software (Accessed 19 Feb. 2025).
2. Donovan, J.E., Pingitore, N.E. and Westphal, A.J. (2003). Compositional Averaging of Backscatter Intensities in Compounds. Microscopy and Microanalysis, 9(3), pp.202–215. https://doi.org/10.1017/s1431927603030137.

This information has been sourced, reviewed and adapted from materials provided by Gatan, Inc.

For more information on this source, please visit Gatan, Inc.