Comparison of optical and tactile layer thickness measurements of polymers and metals on silicon or SiO2

Abstract

The thickness measurement of transparent layers with optical techniques is very problematic. The observed deviations can easily reach 100% of the layer thickness to be measured. In order to analyse these deviations, tactile reference measurements have been developed. The proposed method is based on contact mode measurements with low contact pressure. With stylus instruments, this can be realized either by using the recommended tip radius of 2 µm and very small probing forces in the micronewton range (and low scanning speeds of 50 µm s−1) or by using the recommended probing force of 750 µN, but a large probing tip radius. Three metal layers on silicon or silicon dioxide and two polymer resist materials on a thin chromium adhesive layer on silicon are used as artefacts. The comparison of the optical measurements with the tactile reference values disclosed deviations of the optical measurements of up to 195% of the layer thickness. Layer thicknesses were between 200 nm and 4 µm. This paper analyses the deviations of two white light interference microscopes, one phase shift interference microscope, one confocal microscope, one autofocus sensor, a chromatic sensor, an interferometric film thickness sensor and one spectroscopic ellipsometer. Simple and well-known expressions for the description of the observed deviations are presented and discussed. The order of magnitude of the observed deviations can be described well by these correction formulas but further investigations are necessary in order to better understand the systematic deviations of optical surface measuring instruments on non-cooperating surfaces.