Results were obtained which illustrated the use of tip-enhance Raman spectroscopy (TERS) and imaging in a novel top illumination geometry. A radially polarized beam was used to generate an electric field component in the direction of beam propagation, normal to the surface, resulting in a five times enhancement compared to a linearly polarized beam. This multiplicative enhancement facilitates a discrimination of the near-field signal from the far-field Raman background. The top illumination configuration allows the application of TERS for investigating molecules on a variety of surfaces, such as gold, glass and silicon. The near field Raman spectra of a variety of materials, including single wall carbon nanotubes, were examined. Sufficient enhancement was obtained to permit the obtaining of a sub-diffraction resolution hyperspectral Raman imaging cube of the surface distribution of large bundles of carbon nanotubes of varying diameter. To further facilitate nanoparticle enhanced Raman imaging of complex biological specimens, the use of higher laser order modes with radial and azimuthal polarizations focused onto a gold nanoparticle AFM tip was explored using a back-scattering microscope configuration. In particular, the radial polarized configuration provided enhanced spatial resolution. This is a consequence of the direction of the induced electron oscillation in the metal nanoparticle, which is oriented by the electromagnetic field at the laser focus. Specifically, the directionality of the electromagnetic enhancement is critical such that the molecule under observation experiences the greatest enhanced field at the apex of the nanostructure tip. These results indicate the potential for TERS to provide insights in molecular distributions on the nanoscale, as, for example, those systems representative of diseased states. Significant effort has been expended in optimizing experimental parameters associated with TERS. Although tip-enhancement was clearly observed on carbon nanotubes using gold-coated AFM glass tips with 647nm laser excitation, a stronger enhancement factor is necessarily desired for the Raman spectroscopic elucidation of biological samples. For example, glass micropipettes with gold particle balls attached were found to provide greater enhancement factors than gold-coated AFM glass tips. Additionally, proper discrimination of the near-field Raman scattering from the predominant far-field Raman scattering is necessary for the observation and validation of TERS effect. In the case of model lipid membranes, far-field Raman scattering is minimized by preparing lipid membranes in controlled numbers of layers using the Langmuir-Blogett (LB) film deposition technique. Further, LB film deposition provides a mean to maintain the tip-substrate gap required for tip enhancement. Similarly, thin layer deposition of lipid membranes can also be prepared using vesicle fusion method. Vesicle fusion deposition is particularly suitable for incorporating proteins that are relatively insoluble into lipid membranes. While quartz and mica substrates provide flat surfaces for lipid deposition and for topography visualization. Greater tip-enhancement effect is achieved by choosing substrates that favor the localization of surface plasmons upon laser excitation at the tip apex. With a gold particle ball micropipette, tip-enhanced Raman scattering in the 1800-200 cm-1 region is observed in as little as 1 second acquisition time on a single lipid monolayer deposited on a gold substrate. To obtain proper atomic force microscopy (AFM) topography of the sample, the gold substrate must be annealed prior to sample deposition to reduce the surface roughness from >10A0 to 3-4A0. Alternatively, gold deposits stripped from an ultraflat surface, such as silicon or mica, also provides consistent topography. We note the effect of a gold substrate on lipid membrane formation as lipids appear to form bilayers differently on gold coated substrates than mica substrates. Although infrared (IR) spectroscopy has been widely used for chemical characterization, a true confocal application of IR spectroscopy for studying the variation in chemistry at different distances from the sample surface has not been demonstrated. (Confocal characterization, even though it is essential in biomedical studies, has been mostly limited to structural analyses using fluorescence and Raman spectroscopic techniques.) Thus, we have recently outlined a new design for IR spectroscopic measurements with a theoretical spatial resolution of 1nm in the z-direction despite diffraction limit considerations. Our approach involves the modification of a time-scanned IR-microscope with the addition of a high-speed, high-precision nano-positioner.. The distance between the objective and the sample is modulated by the nano-positioner at a fixed periodicity in order to modulate the spectral signal. This represents a new application of a Fourier-transform IR difference spectroscopic technique. Experiments are initiated with samples of alternating atomic layers of HfO2 and Al2O3, useful samples for their simple chemical composition. Protein membranes will be assembled, layer-by-layer, to study the spectroscopic variations within individual bilayers.
