Immersion ellipsometry for the uncorrelated determination of ultrathin film thickness and index of refraction: Theory and examples

Abstract

Immersion ellipsometry can break the well-known correlation between optical constants and thicknesses of ultrathin (<5–10 nm) films, allowing both to be determined. In immersion ellipsometry, ellipsometric data is acquired in air and liquid ambients, and the data sets are combined in the analysis. The contrast in index between the liquid and film adds information to the analysis that breaks the correlation between the film thickness and refractive index that exists for air-only measurements. We describe the theory and practice of immersion ellipsometry. We also discuss the use of multiwavelength immersion ellipsometry to measure the thicknesses and optical constants of two thin films: native oxide on silicon and an alkyl monolayer on that native oxide. The average thicknesses of the native oxide and chloro(dimethyl)octadecylsilane (CDMOS) monolayer were 1.526 ± 0.027 nm and 1.968 ± 0.057 nm, and their average indices of refraction at 633 nm were 1.519 ± 0.005 and 1.471 ± 0.004, respectively. The native oxide and CDMOS monolayer were also characterized with x-ray photoelectron spectroscopy (XPS) and contact angle goniometry. Both the XPS C 1 s peak and the water contact angle increased substantially after monolayer deposition. While immersion ellipsometry has been known for decades, its use has been limited, maybe due to a lack of awareness of the technique and/or the need to immerse the sample surface in a liquid that could be destructive if the sample is not compatible with the liquid. As ultrathin films become widely used in science technology, immersion ellipsometry should increase in importance.