Extensive effort and research are currently channeled towards the implementation of SERS (Surface Enhanced Raman Spectroscopy) as a standard analytical tool as it has undisputedly demonstrated a great potential for trace detection of various analytes. Novel and improved substrates are continuously reported in this regard. It is generally believed that plasmonic nanostructures with plasmon resonances close to the excitation wavelength (on-resonance) generate stronger SERS enhancements, but this finding is still under debate. In the current paper, we compared off-resonance gold nanobones (GNBs) with on-resonance GNBs and gold nanorods (GNRs) in both colloidal dispersion and as close-packed films self-assembled at liquid-liquid interface. Rhodamine 6G (R6G) was used as a Raman reporter in order to evaluate SERS performances. A 17-, 18-, and 55-fold increase in the Raman signal was observed for nanostructures (off-resonance GNBs, on-resonance GNBs, and on-resonance GNRs, respectively) assembled at liquid-liquid interface compared to the same nanostructures in colloidal dispersion. SERS performances of off-resonance GNBs were superior to on-resonance nanostructures in both cases. Furthermore, when off-resonance GNBs were assembled at the liquid interface, a relative standard deviation of 4.56% of the recorded signal intensity and a limit of detection (LOD) of 5 × 10-9 M could be obtained for R6G, rendering this substrate suitable for analytical applications.
Keywords: SERS; liquid interface; nanobones; nanorods; resonance; self-assembly.