ABSTRACT
The characterization of thin coatings on polymers such as poly(ethylene terephthalate) (PET) is required in order to study chemical composition and coating continuity. Two different methods of applying Raman spectroscopy for this purpose are compared in this paper. Using confocal Raman microscopy, thick coatings (> 10 microns) are relatively easily identified; however, the Raman scattering from the acrylic coatings commonly used is much weaker than that of PET and consequently, there is a background due to the substrate. Thin acrylic coatings (< 1 micron) usually cannot be detected. Surface-enhanced Raman scattering (SERS) of uncoated PET gives intense signals and if the spectra are taken from the metal-coated side, there is no evidence of the underlying Raman scattering from the bulk. Acrylic coatings do not give sufficiently strong or reproducible SERS to be reliably identified, but even thin (20 nm) coatings completely block the SERS from the substrate. Only where gaps appear in the coating is the SERS of the underlying PET seen. To detect a positive signal from the coating, SERS active labels were incorporated into the acrylic at low concentrations either as a physical mixture or as reactive co-monomers. This uniquely labels the coating and allows detection and, in principle, mapping of the coverage. Thus, for thick (> > 1 micron) coatings, normal Raman spectroscopy is an effective technique for detecting the presence of the surface coating. However, it is ineffective with thin (< 1 micron) coatings, and SERS alone only indicates where the coating is incomplete or defective. However, when a SERS label is added, spectra can be detected from very thin coatings (20 nm). The concentration of the labels is sufficiently low for the coating to remain colorless.