Quantitative detection of organic thin films on opaque surfaces utilizing Coherent anti-Stokes Raman Scattering Imaging

Abstract

Thin film contamination on critical surfaces is a significant cause of failure in different industries, including semiconductor, optical and medical manufacturing. Especially surfactants can alter surface properties that subsequent production steps and product quality are impacted. The contactless, non-destructive and fast detection of those trace organic residues remains a substantial metrological frontier. In this paper we establish epi-CARS Microscopy as a quantitative tool for mapping those contaminations. Using CARS in reflection on opaque surfaces poses certain obstacles, like non-resonant substrate background concealing the contaminants chemical signal and non-linear relationship between signal intensity and contaminant thickness. We describe the coherent interplay between the CARS signal arising from contaminant and substrate. By analyzing that signal, we decouple vibrational response from optical artifacts, thereby quantifying spatially resolved contaminant thickness. We validate this by analyzing silicon wafer samples contaminated with dimethylpolysiloxane in combination with AFM thickness verification. The epi-CARS detection limit for filmic contamination is determined, consequently more time-consuming and extensive techniques can be substituted. The ability to create quantitative maps of thin silicone films within seconds, achieving nanometer thickness sensitivity, is demonstrated. Quantitative epi-CARS is established as a fast, reliable, high-resolution and large-area analytical tool for surface cleanliness validation and contamination identification.