In this perspective we discuss the roles of hot spots in surface-enhanced Raman spectroscopy (SERS). After giving background and defining the hot spot, we evaluate a variety of SERS substrates which often contain hot spots. We compare and discuss the differentiating properties of each substrate. We then provide a thorough analysis of the hot spot contribution to the observed SERS signal both in ensemble-averaged and single-molecule conditions. We also enumerate rules for determining the SERS enhancement factor (EF) to clarify the use of this common metric. Finally, we present a forward-looking overview of applications and uses of hot spots for controlling chemistry on the nanoscale. Although not exhaustive, this perspective is a review of some of the most interesting and promising methodologies for creating, controlling, and using hot spots for electromagnetic amplification.

Single-molecule surface-enhanced Raman spectroscopy (SMSERS) of crystal violet (CV) has been reported since 1997, yet others have offered alternative explanations that do not necessarily imply SMSERS. Recently, the isotopologue approach, a statistically significant method to establish SMSERS, has been implemented for members of the rhodamine dye family. We provide the first demonstration of SMSERS of a triphenylmethane dye using the isotopologue approach. Two isotopologues of CV are employed to create chemically identical yet vibrationally distinct probe molecules. Experimental spectra were compared extensively with computational simulations to assign changes in mode frequencies upon deuteration. More than 90 silver nanoparticle clusters dosed with a 50:50 mixture of CV isotopologues were spectroscopically characterized, and the vibrational signature of only deuterated or undeuterated CV was observed 79 times, demonstrating that the isotopologue approach for proving SMSERS is applicable to both the CV and the rhodamine systems. The use of CV, a minimally fluorescent dye, allowed direct evaluation of enhancement factors (EF), which are reported herein. Through experiment and theory, we show that molecular electronic resonance Raman (RR) and surface-enhanced Raman effects combine synergistically in SMSERS. Excluding RR effects, the EF(SERS) is ∼10(9). Variations and relationships between substrate morphology and optical properties are further characterized by correlated SMSERS-localized surface plasmon resonance (LSPR)-high-resolution transmission electron microscopy (HRTEM) studies. We did not observe SMSERS from individual nanoparticles; further, SMSERS-supporting dimers are heterodimers of two disparately sized particles, with no subnanometer gaps. We present the largest collection to date of HRTEM images of SMSERS-supporting nanoparticle assemblies.

Determining the existence of any direct spectral relationship between the far-field scattering properties and the near-field Raman-enhancing properties of surface-enhanced Raman spectroscopy (SERS) substrates has been a challenging task with only a few significant results to date. Here, we prove that hot spot dominated systems show little dependence on the far-field scattering properties because of differences between near- and far-field localized surface plasmon resonance (LSPR) effects as well as excitation of new plasmon modes via a localized emitter. We directly probe the relationship between the near- and far-field light interactions using a correlated LSPR-transmission electron microscopy (TEM) surface-enhanced Raman excitation spectroscopy (SERES) technique. Fourteen individual SERS nanoantennas, Au nanoparticle aggregates ranging from dimers to undecamers, coated in a reporter molecule and encased in a protective silica shell, were excited using eight laser wavelengths. We observed no correlation between the spectral position of the LSPR maxima and the maximum enhancement factor (EF). The single nanoantenna data reveal EFs ranging from (2.5 ± 0.6) × 10(4) to (4.5 ± 0.6) × 10(8) with maximum enhancement for excitation wavelengths of 785 nm and lower energy. The magnitude of maximum EF was not correlated to the number of cores in the nanoantenna or the spectral position of the LSPR, suggesting a separation between near-field SERS enhancement and far-field Rayleigh scattering. Computational electrodynamics confirms the decoupling of maximum SERS enhancement from the peak of the scattering spectrum. It also points to the importance of a localized emitter for radiating Raman photons to the far-field which, in nonsymmetric systems, allows for the excitation of radiative plasmon modes that are difficult to excite with plane waves. Once these effects are considered, we are able to fully explain the hot spot dominated SERS response of the nanoantennas.

The surface-enhanced Raman excitation profiles (REPs) of rhodamine 6G (R6G) on Ag surfaces are studied using a tunable optical parametric oscillator excitation source and versatile detection scheme. These experiments afford the ability to finely tune the excitation wavelength near the molecular resonance of R6G (i.e., approximately 500-575 nm) and perform wavelength-scanned surface-enhanced Raman excitation measurements of a single molecule. The ensemble-averaged surface-enhanced REPs are measured for collections of molecules on Ag island films. The relative contributions of the 0-0 and 0-1 vibronic transitions to the surface-enhanced REPs vary with vibrational frequency. These results highlight the role of excitation energy in determining the resonance Raman intensities for R6G on surface-enhancing nanostructures. Single-molecule measurements are obtained from individual molecules of R6G on Ag colloidal aggregates, where single-molecule junctions are located using the isotope-edited approach. Overall, single-molecule surface-enhanced REPs are narrow in comparison to the ensemble-averaged excitation profiles due to a reduction in inhomogeneous broadening. This work describes the first Raman excitation spectroscopy studies of a single molecule, revealing new information previously obscured by the ensemble.

Abstract