Basics of Ellipsometry

Abstract

Ellipsometry is a technique for the contact-less and non-destructive optical characterization of surfaces [1],[2],[3],[4]. It is based on the fact that a monochromatic electromagnetic wave changes its state of polarization if it strikes non-perpendicularly the interface between two dielectric media. In general, any arbitrary monochromatic transversal wave can be considered composed of two orthogonal coherent waves with a fixed phase relation, e.g., of two linearly polarized waves whose electric field vectors lie within two perpendicular planes (Fig. 1). The field vector E_res which results from a vector addition of the two components E_x and E_y lies within a plane if and only if the two constituent waves are in phase, or out of phase by a multiple of π (which corresponds to an inversion of one of them); the resulting wave is linearly polarized in this case (Fig. 1 (a)). Otherwise, the field vector performs a screw-like motion around the direction of its propagation (the z-direction); its projection onto a plane perpendicular to its direction of propagation describes, in general, an ellipse (Fig. 1 (b)). This ellipse degenerates to a line if the phase difference between the two constituent waves is a multiple of π; it becomes a circle if their amplitudes are equal, and the phase difference between them is π/2 (90°).