Scanning tunneling microscopy (STM) is an attractive tool to be complemented by optical spectroscopy because the molecule under optical interrogation can be directly visualized with the otherwise unattainable angstrom resolution. The simplest ensuing technique from that combination is tip-enhanced Raman spectroscopy (TERS). TERS utilizes a localized surface plasmon (LSP) at the tip apex to enhance the Raman signal. It is the gap-mode LSP that is sensitive to the gap distance, thereby provides the spatial resolution better than the diffraction limit. However, the gap-mode LSP at the tunneling junction strongly depends on details of the tip apex such as the apex diameter and surface smoothness, which are difficult to tailor for realizing TERS with the single-molecule sensitivity.

Along with efforts to overcome the difficulties mentioned above, the Apkarian group has investigated  [4-(phenylazo)phenoxy]hexane-1-thiol (azobenzene thiol),[1] meso-tetrakis(3,5- di-tertiarybutylphenyl)-porphyrin (H2TBPP), and Fe(1,10-phenanthroline)₂(NCS)₂ molecules using TERS. Contrary to literature in which static TER spectra are usually justified by comparison to bulk Raman spectra, the TER spectra we obtained are quite different from standard Raman spectra. The vibrational peaks are often shifted, meander, and fluctuate. Also we often observe the Raman signal increases as the gap widens. I will address these unconventional results with the help of computational results, along with our efforts to produce reliable tips for TERS experiments.