Model selection in spectroscopic ellipsometry data analysis: Combining an information criteria approach with screening sensitivity analysis
Source abstract
Abstract
Abstract In the field of optical metrology, the selection of the best model to fit experimental data is absolutely nontrivial problem. In practice, this is a very subjective and formidable task which highly depends on metrology expert opinion. In this paper, we propose a systematic approach to model selection in ellipsometric data analysis. We apply two well-established statistical methods for model selection, namely, the Akaike (AIC) and Bayesian (BIC) Information Criteria, to compare different dispersion models with various complexities and objectively determine the “best” one from a set of candidate models. The information criteria suggest the most optimal way to quantify the balance between goodness of fit and model complexity. In combination with screening-type parametric sensitivity analysis based on so-called “elementary effects” (the Morris method) this approach allows to compare and rate various models, identify key model parameters and significantly enhance process of ellipsometric measurements evaluation.
Source: openaire · Quality: candidate unverified · Source record ↗
Research context
OpenAlex topics: Spectroscopy and Chemometric Analyses · Advanced Statistical Methods and Models · Scientific Measurement and Uncertainty Evaluation
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
References (44)
Show source-provided references
1. Spectroscopic ellipsometry for inline process control in the semiconductor industry
2. Nanoscale characterization and metrology10.1116/1.4807116 ↗
3. Spectroscopic ellipsometry for active nano- and meta-materials10.1515/ntrev-2013-0043 ↗
4. JPhysics of optical metrology of silicon-based semiconductor devices
5. Optical physics of materials
6. Data analysis for spectroscopic ellipsometry
7. Parameterization of the dielectric function of semiconductor nanocrystals10.1016/j.physb.2014.03.065 ↗
8. Ellipsometric determination of optical constants for silicon and thermally grown silicon dioxide via a multi-sample multi-wavelength, multi-angle investigation10.1063/1.367101 ↗
9. Spectroscopic ellipsometry: a historical overview10.1016/S0040-6090(97)00762-1 ↗
10. A meeting with Enrico Fermi10.1038/427297a ↗
11. Drawing an elephant with four complex parameters10.1119/1.3254017 ↗
12. The problem of overfitting10.1021/ci0342472 ↗
13. H. Akaike, Information theory as an extension of the maximum likelihood principle, in: B.N. Petrov, F. Csaki (Eds.), Proceeding of the Second International Symposium on Information Theory, Akademiai Kiado, Budapest (1973), pp. 267–281; Reprinted in: S. Kotz, N.L. Johnson (Eds.), Breakthroughs in Statistics, vol. I, Foundations and Basic Theory, Springer, New York, NY, 1992, pp. 610–624 and E. Parzen, K., Tanabe, G. Kitagawa (Eds.), Selected Papers of Hirotugu Akaike, Springer New York NY, 1998, pp. 199–213.10.1007/978-1-4612-1694-0_15 ↗
14. A new look at the statistical model identification10.1109/TAC.1974.1100705 ↗
15. Estimating the dimension of a model10.1214/aos/1176344136 ↗
16. Burnham · 2002
17. Anderson · 2008
18. Banks · 2014
19. Box · 2005 · p440
20. A practitioner’s approach to evaluation strategy for ellipsometric measurements of multilayered and multiparametric thin-film structures10.1016/j.tsf.2015.10.078 ↗
21. Information criteria for efficient quantum state estimation10.1103/PhysRevA.83.062110 ↗
22. Optical determination of the film thicknesses in multilayer thin film structures10.1016/0040-6090(93)90762-E ↗
23. Sensitivity analysis as an ingredient of modeling10.1214/ss/1009213004 ↗
24. Efficient thin-film stack characterization using parametric sensitivity analysis for spectroscopic ellipsometry in semiconductor device fabrication10.1016/j.tsf.2015.05.049 ↗
25. How to avoid a perfunctory sensitivity analysis10.1016/j.envsoft.2010.04.012 ↗
26. When all models are wrong
27. Factorial sampling plans for preliminary computational experiments10.1080/00401706.1991.10484804 ↗
28. Saltelli · 2004
29. Saltelli · 2008
30. Enhancing the Morris method
31. An effective screening design for sensitivity analysis of large models10.1016/j.envsoft.2006.10.004 ↗
32. From screening to quantitative sensitivity analysis. A unified approach10.1016/j.cpc.2010.12.039 ↗
33. Bayesian model selection in social research10.2307/271063 ↗
34. Spectroscopic ellipsometry characterization of ultrathin silicon-on-insulator films10.1116/1.2213265 ↗
35. Observation of quantum confinement and quantum size effects10.1002/pssa.200777891 ↗
36. Advanced metrology for next generation transistors
37. Spectroscopic ellipsometry of nanoscale materials for semiconductor device applications
38. OLSA 1.3 User’s Guide, KLA-Tencor Corporation, Milpitas, CA, U.S.A., 1999.
39. Kramers-Kronig relations in optical data inversion10.1103/PhysRevB.44.8301 ↗
40. Structure and lattice dynamics of binary lead silicate glasses investigated by infrared spectroscopy10.1016/j.jnoncrysol.2006.02.004 ↗
41. Analytic representations of the dielectric functions of materials for device and structural modeling10.1016/S0040-6090(97)00799-2 ↗
42. Analytic representations of the dielectric functions of crystalline and amorphous Si and crystalline Ge for very large scale integrated device and structural modeling10.1116/1.581137 ↗
43. Regression calibration method for rotating element ellipsometers10.1016/0040-6090(93)90293-X ↗
44. A practical method for optical dispersion model selection and parameters variations in scatterometry analysis with variable n&k's10.1016/j.tsf.2014.03.082 ↗