Data Averaging Effect on Roughness Measurement using Phase-Shifting Interferometry for a Few-Ten-Angstrom and Sub-Angstrom Rough Surfaces

Abstract

The phase and intensity data averaging effect on roughness measurement using phase-shifting interferometry was investigated for a few-ten-Å and sub-Å rough surfaces. For a few-ten-Å rough surface, the roughness value did not depend on the data average. In contrast, the sub-Å rough surface determination was strongly influenced by the data average so that measurement error was significantly improved as the number of data averages increased. As an optimal data average condition that the measurement noise was minimized, 30 phase data and 20 intensity data averages were experimentally determined. Sub-Å roughness measurement under the optimal data average condition had high measurement accuracy of the absolute roughness and small measurement reproducibility error.