Abstract:
The main focus of this thesis was the physical interpretation of the pumping cross-section. This was achieved by performing a statistical analysis of single molecule vibrational pumping events in which both the SERS and pumping cross-sections could be measured simultaneously. Samples were constructed in which small aggregates of silver colloids were evenly distributed on a dry surface.
The sample was then cooled to 77K so that the main mechanism for creating a vibrational population was through Stokes scattering. Spatial mappings were then performed which measured how the SERS spectrum varied with position
on the sample and the single molecule events were identified. The SERS cross-sections were determined from the Stokes intensity while the pumping cross-sections were determined from the ratio of the anti-Stokes and Stokes peaks. It was observed that the pumping cross-section was often significantly larger than the SERS cross-section, as much as four orders of magnitude in some cases. Several attempts were made to explain this discrepancy including the possibility of
the surface plasmon resonance favouring anti-Stokes scattering, underestimated lifetimes for the vibrational modes, and additional pumping from fluorescence. However, the most likely candidate was non-radiative Stokes scattering by the observed molecule which would increase the vibrational population but would not increase the Stokes intensity. To estimate the proportion of scattered light that is radiative or non-radiative,
single molecule measurements were performed under both surface-enhanced and unmodified conditions. By comparing the fluorescence and Raman intensities under these scenarios, it was possible to estimate the radiative and non-radiative enhancement factors. It was found that the non-radiative SERS cross-section was typically much larger than the radiative cross-section for samples consisting of aggregated silver colloids. The discrepancy between the pumping and SERS cross-section (which is the radiative cross-section) could therefore be explained by non-radiative scattering dominating the creation of the vibrational population, along with an additional contribution due to the plasmon resonance favouring certain vibrational modes. Furthermore, the lifetime of a molecule after it has been excited to the first electronic state was estimated to be as short as 25 fs. It would be impossible to measure lifetimes of this order of magnitude in single molecules using time-resolved techniques. Furthermore, to the very best of our knowledge, this is the first time that an experimental determination of the non-radiative SERS cross-section has been made.
