This paper considers the quantitative implications of out-of-focus regions on the lateral and depth resolution of Raman microscopy, with special regard for the surface specificity of the technique. It builds on work that has recently appeared in the literature which shows that with transparent samples, signals can originate throughout a large extended illumination volume, even though most of this region is out of focus with regard to the confocal aperture. This gives rise to weak but readily detectable spectral contributions from regions that are tens of micrometers from the point of tightest focus, an effect that is easily demonstrated if the laser is focused far above the sample surface. When we integrate the signals arising throughout this extended volume, the resulting total signal can be significant with respect to the Raman signals originating from the point of focus; this has obvious implications for surface specificity and depth resolution. Furthermore, as one moves the focal point through and above a sample surface, signals from thick transparent samples decay relatively slowly compared with thin or opaque ones, where the extended focal volume is irrelevant. This means that on moving above the surface of a thinly coated thick substrate during a confocal axial scan, the substrate-to-coating signal ratio increases dramatically, contrary to intuition. Consequently, confusing spectral artifacts arise if one focuses above the sample surface, either inadvertently when mapping an uneven sample, or deliberately in an attempt to improve surface specificity. In this work we show how a simple analytical model can predict the surface/substrate signal ratio as a function of distance above the surface. The model is validated using experimental data from monofilms and coated films. Furthermore, we show how this effect is not limited to the confocal axial profiling geometry. Similar effects are obtained when one scans laterally beyond the edge of mechanically prepared cross-sections due to an extended, out-of-focus laser field that can sample lateral regions far to the side of the optimum focus. This effect can lead to very confusing results, such as spectra from the substrate increasing in absolute intensity as one moves beyond the edge of the coating into the air. These observations, which as far as we are aware have not previously been reported, are rationalized using a simple ray-tracing description, which shows the potential for coupling light into the cross-section, which acts as a waveguide. These effects have a completely different origin than the well-known anomalies that are introduced by refraction and spherical aberration; even with a perfect, aberration-free system, the extended focal volume may cause significant degradation in depth resolution. Although the effects have been demonstrated with simple film systems, they have the potential to impact the results from Raman mapping and imaging of any samples that contain significant refractive index discontinuities, which can potentially cause refraction and waveguiding, or that have compositional depth gradients and an uneven sample surface.
Read full abstract