Surface- and tip-enhanced raman spectroscopy of silicon

Strongly enhanced optical fields generated by nanoparticles due to Local Surface Plasmonic Resonance (LSPR) enable Surface-Enhanced Raman Spectroscopy (SERS) to perform chemical analysis with extremely high sensitivity. This technique, however, suffers from the disadvantage of low spatial resolution, i.e. 500 nm, due to the optical diffraction limit imposed by conventional optics. On the other hand, a metallic tip irradiated by a laser beam, by which a highly localized and significantly enhanced optical field can be induced, allows Raman spectroscopy to have a spatial resolution of 10 nm. This technique is named as Tip-Enhanced Raman Spectroscopy (TERS). In this study, we took advantages of both SERS and TERS techniques to achieve both high resolution and high sensitivity at the same time. SERS-active silicon substrates with metallic nanostructures were fabricated by Nanosphere Lithography (NSL) technique. Silver (Ag)-coated tungsten tips and gold (Au) tips were prepared by the electrochemical etching method. In order to making use of the evanescent optical field underneath the tip apex, the tip was controlled to position above the substrate surface with a gap distance of 1 nm by a Scanning Tunneling Microscope (STM). The tip was precisely aligned in the region where two adjacent nanostructures (dipole) were located by a nanopositioner. The contrast ratio of Raman intensity of silicon substrates was much higher with the presence of the nanostructures. It is suggested that this extraordinary enhancement is attributed to the interplay between the enhanced optical fields from the nanostructure dipole and from the tip. Numerical simulation was performed to verify the suggestion using the finite-domain-time-difference (FDTD) algorithm based upon the Lorentz-Drude model. A spatial resolution well below 100 nm was confirmed by mapping the metallic nanostructures. This technique has a great potential for single molecule detection, disease diagnosis and therapy control, and nanodevice analysis.