Research On Parameter Characterization Of Nano-films Based On Ellipsometry Data Processing

With the continuous development of thin-film materials and manufacturing technologies,accurate and rapid characterization methods of optical constants and thickness of nano-scale thin films have attracted widespread attention.Elliptical polarization measurement is a high-precision,non-contact measurement method that obtains the optical constants and film thicknesses of film materials by analyzing changes in the polarization state of the polarized light before and after reflection on the surface of the film sample to be measured.Since elliptical polarization measurement is an indirect measurement method,its measurement results largely depend on the establishment of models and the application of intelligent methods of data fitting during data processing.In this paper,based on the basic principle of ellipsometry,the detailed description of the thin film material structure and optical characteristics modeling method and data fitting principle is given.For the thin film material structure characteristic modeling,a simplified three-phase modeling structure is proposed,which can simplify the modeling process Moreover,the fitting accuracy can be improved;secondly,the particle swarm algorithm is improved and the IPSO-BP algorithm is proposed for the shortcomings of the ellipsometric data fitting process,which particularly depends on experience and prior knowledge,and the poor repeatability of the fitting results.This data fitting algorithm is applied to the data processing of various thickness film samples.The results show that this method not only reduces the influence of human factors during modeling,but also ensures the consistency of the analysis results.Finally,the data processing method proposed in this paper is verified by using the film thickness standard samples produced by American VLSI Company,which is certified by the international authoritative measurement organization NIST,and the commercial ellipsometric data processing software.The results show that the data fitting algorithm proposed in this paper is more accurate than other algorithms in terms of structural and optical properties,and it can avoid the impact of periodicity when fitting large-size films,and has the only fitting result.specialty.The research results of this paper can lay a foundation for the research on the calibration and traceability of instruments.