Robust Spectral Signal Processing-Based Framework for Infrared Interferometry-Driven SiC Epitaxial Thickness Estimation

View at publisher Summarise with VIRAT
Authors: Yibo Zhang
Cited by: 0
References: 3

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

This paper addresses simulation-based parameter estimation for epitaxial-layer thickness from infrared reflectance spectra and proposes a hybrid computational framework that integrates a double-beam interference model, a robust signal-processing pipeline, and a Fabry–Perot regime detector. We design a data workflow with three-sigma outlier removal, moving-window smoothing, and an integer-order constraint derived from adjacent extrema to avoid direct order solving; when multi-beam interference is detected (FWHM30), we switch to a multi-beam fitting module to debias thickness estimates. On two silicon-carbide datasets measured at different incidence angles, the framework yields mean thicknesses of 18.5987 and 17.1213 micrometers with standard deviations of 0.4615 and 0.6202, achieving relative standard deviations below 5% and a cross-angle discrepancy of about 8.27%; for silicon samples the estimated means are 11.6348 and 11.5518 micrometers. After multi-beam correction, the silicon-carbide thicknesses adjust to 21.6914 and 21.1064 micrometers with cross-angle discrepancy reduced to 2.73%, demonstrating robustness and practical accuracy. The result is a deployable, generalizable computational solution for optical metrology in manufacturing settings.

Fields of Study & Subjects

Optical Polarization and Ellipsometry Spectroscopy and Chemometric Analyses Surface Roughness and Optical Measurements

Publication details

Journal / source
IEEE International Conference on Power, Electronics and Green Energy (ICPEGE)
Pages
793–797
Publication date
28 January 2026
Language
EN