Controlled synthesis of silver nanostructures for use in surface-enhanced Raman spectroscopy : a thesis presented in partial fulfilment of the requirements for the degree of Master of Science in Nanoscience, School of Fundamental Sciences, Massey University, Palmerston North, New Zealand

Abstract

Surface-enhanced Raman scattering (SERS) has been extensively researched in the past few years, with further research being done into the development of noble metal substrates that have been shown to drastically increase the enhancement provided by SERS. Noble metals are exceptionally good at enhancing Raman signals stemming from their innate optical properties. The innate enhancement factor of these noble metals has been essential for SERS research, as without them, the appearance of SERS peaks will be indistinguishable from the background noise. Research has shown that the Raman enhancement provided by noble metals substrates can be improved upon by using controlled, uniform nanostructures with sharp corners and edges instead of standard substrates or heterogeneous mixtures of random nanoparticles. Both silver and gold have been shown to adopt several unique nanostructure shapes ranging from nanocubes to complex nanoflowers. This thesis discusses the synthesis and characterization of three novel silver nanostructures along with the results of using the nanostructures in a variety of Raman spectroscopy experiments. The nanostructures produced were nanocubes, nanoplates and nanowires with all three being used as substrates in several solution based Raman spectroscopy experiments, which included conventional SERS, single molecule SERS, microfluidic SERS and Raman tweezers. The primary hypothesis of the project is that different regions of the substrate will provide different degrees of enhancements due to a difference in localized surface plasmon resonance (LSPR) intensity. In the case of the nanostructures used in this project, there should be a difference in LSPR between the face, edges and corners of the three structures, as each region will contain different degrees of LSPR interaction. To test this hypothesis the data collected from the SERS experiments was processed using several statistical analysis techniques, including Euclidean distance mapping, principal component analysis (PCA) and self-organizing mapping, to test the validity of the hypothesis. For this hypothesis to be correct the data should fit into a discrete number of groups that represent the different regions of the substrate. The experimental data did not appear to support the original hypothesis, indicating that a more nuanced explanation may be required to describe the generation of SERS signal from controlled substrates.