Leveraging Machine Learning for Advanced Nanoscale X-ray Analysis: Unmixing Multicomponent Signals and Enhancing Chemical Quantification
Source abstract
Abstract
Energy dispersive X-ray (EDX) spectroscopy in the transmission electron microscope is a key tool for nanomaterials analysis, providing a direct link between spatial and chemical information. However, using it for precisely determining chemical compositions presents challenges of noisy data from low X-ray yields and mixed signals from phases that overlap along the electron beam trajectory. Here, we introduce a novel method, non-negative matrix factorization based pan-sharpening (PSNMF), to address these limitations. Leveraging the Poisson nature of EDX spectral noise and binning operations, PSNMF retrieves high-quality phase spectral and spatial signatures via consecutive factorizations. After validating PSNMF with synthetic data sets of different noise levels, we illustrate its effectiveness on two distinct experimental cases: a nanomineralogical lamella, and supported catalytic nanoparticles. Not only does PSNMF obtain accurate phase signatures, but data sets reconstructed from the outputs have demonstrably lower noise and better fidelity than from the benchmark denoising method of principle component analysis.
Source: openalex_inverted_index · Quality: baseline
Research context
OpenAlex topics: Machine Learning in Materials Science · Advanced Electron Microscopy Techniques and Applications · Protein Structure and Dynamics
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
Electron Spectrometry and Microscopy Laboratory (LSME), Institute of Physics (IPHYS), École Polytechnique Fédérale de Lausanne (EPFL), 1015 Lausanne, Switzerland
References (38)
Show source-provided references
1. Williams, D. B.; Carter, C. B. X-ray Spectrometry. Transmission Electron Microscopy: A Textbook for Materials Science; Springer, 2009; pp 581–603.10.1007/978-0-387-76501-3_32 ↗
2. 10.1557/mrs.2011.33110.1557/mrs.2011.331 ↗
3. 10.1038/nenergy.2017.11110.1038/nenergy.2017.111 ↗
4. 10.1038/nenergy.2015.1210.1038/nenergy.2015.12 ↗
5. 10.1021/acsami.1c1025110.1021/acsami.1c10251 ↗
6. 10.1039/D0TA10415A10.1039/D0TA10415A ↗
7. 10.1021/acs.chemmater.0c0135210.1021/acs.chemmater.0c01352 ↗
8. Jolliffe I. T. · 2002
9. 10.1017/S143192762101393310.1017/S1431927621013933 ↗
10. 10.1186/s40679-019-0066-010.1186/s40679-019-0066-0 ↗
11. 10.1016/j.micron.2017.02.00110.1016/j.micron.2017.02.001 ↗
12. 10.1016/j.ultramic.2017.06.02310.1016/j.ultramic.2017.06.023 ↗
13. 10.1073/pnas.230202812010.1073/pnas.2302028120 ↗
14. 10.1016/S0893-6080(00)00026-510.1016/S0893-6080(00)00026-5 ↗
15. 10.1038/4456510.1038/44565 ↗
16. 10.1016/j.ultramic.2010.10.00110.1016/j.ultramic.2010.10.001 ↗
17. 10.1016/j.actamat.2016.01.04210.1016/j.actamat.2016.01.042 ↗
18. 10.1021/acs.nanolett.7b0178910.1021/acs.nanolett.7b01789 ↗
19. 10.1021/acsnano.2c0888410.1021/acsnano.2c08884 ↗
20. 10.1155/2020/373540310.1155/2020/3735403 ↗
21. 10.1093/jmicro/dfz03610.1093/jmicro/dfz036 ↗
22. 10.1016/j.rse.2011.03.00310.1016/j.rse.2011.03.003 ↗
23. 10.1109/TGRS.2006.88846610.1109/TGRS.2006.888466 ↗
24. 10.1109/JSTARS.2013.224225510.1109/JSTARS.2013.2242255 ↗
25. 10.1016/j.isprsjprs.2020.11.00110.1016/j.isprsjprs.2020.11.001 ↗
26. 10.1063/5.003132410.1063/5.0031324 ↗
27. 10.4324/978020348342810.4324/9780203483428 ↗
28. Hasinoff, S. W. Computer Vision: A Reference Guide; Springer, 2021; pp 980–982.10.1007/978-3-030-63416-2_482 ↗
29. 10.1029/2021GL09244610.1029/2021GL092446 ↗
30. 10.1126/science.aam856110.1126/science.aam8561 ↗
31. 10.1007/s00410-004-0641-810.1007/s00410-004-0641-8 ↗
32. 10.1016/j.gca.2004.06.04110.1016/j.gca.2004.06.041 ↗
33. 10.1016/j.gca.2004.03.01410.1016/j.gca.2004.03.014 ↗
34. 10.1016/j.ultramic.2024.11398110.1016/j.ultramic.2024.113981 ↗
35. 10.1016/j.ultramic.2023.11371910.1016/j.ultramic.2023.113719 ↗
36. 10.1007/s11831-020-09518-x10.1007/s11831-020-09518-x ↗
37. 10.1111/j.1365-2818.1975.tb03895.x10.1111/j.1365-2818.1975.tb03895.x ↗
38. 10.1039/D3EY00162H10.1039/D3EY00162H ↗