Raman Spectroscopy of Single Molecules

Abstract

Because of its high information content on chemical structure, Raman scattering is a very promising technique for single molecule spectroscopy, which allows establishing the structural identity of a single molecule based on its vibrational spectrum [1,2]. For experiments performed at room temperature and in solutions, SERS (surface-enhanced Raman scattering) is superior to broad and nonspecific fluorescence spectra obtained under similar conditions. Furthermore, non-fluorescent molecules might be detected and identified at the single-molecule level without need for fluorescence labeling. Raman scattering can be used under electronic “nonresonant” conditions, which avoids photobleaching. Relatively short vibrational lifetimes compared to fluorescence lifetimes allow more excitation emission cycles per time interval and therefore the maximum possible number of emitted Raman photons per time interval will be larger than the number of fluorescence photons [3]. Single molecule Raman spectroscopy is based on the strongly enhanced Raman scattering signal which occurs when the target molecule is attached to silver and gold nanostructures, called surface-enhanced Raman scattering (SERS) [1,2]. It is generally agreed that different effects contribute to the large effective Raman cross section observed in SERS experiments. The enhancement mechanisms are roughly divided into so-called electromagnetic and chemical or first layer effects [4-7]. The basic idea of chemical SERS enhancement is a metal electron mediated resonance Raman effect in the “system” molecule-metal. The magnitude of chemical enhancement has been discussed to reach not more than 102-103. The electromagnetic enhancement factor arises from enhanced local optical fields in the vicinity of the metallic nanostructures due to resonances with their surface plasmons. Particularly high field enhancement seems to exist for ensembles of metallic nanoparticles, such as silver or gold colloidal clusters formed by aggregation of colloidal particles or for island films of those metals. Plasmon resonances in such structures can result in a strong confinement of optical fields in very small areas, so-called “hot spots whose dimensions can be smaller than tenths of the wavelength [8]. The high local optical fields in the hot spots provide a rationale for non-resonant SERS enhancement up to 14 orders of magnitude, where field enhancement can contribute 12 orders of magnitude. This enhancement brings effective Raman cross sections to the level of effective fluorescence cross sections of good laser dyes and allows that a single molecule can be detected by means of its non-resonant surface-enhanced Raman spectrum [9-13]. Moreover, the strong lateral confinement of the field enhancement provides an additional opportunity for spectroscopically selecting a single species [14]. In SERS spectroscopy, the target species “feels” very high local optical field strength and, in particular, also very strong field gradients. These both effects might result in some special effects in single molecule SERS spectra, which do not occur in “normal” Raman spectroscopy [15,16]. The large field gradients on SERS-active substrates can result lowering of the symmetry of the resonance Raman scattering tensor compared to “normal” Raman scattering . In the large field strengths in the hot spots on a SERS-active substrate, the treatment of the molecular vibrations in the harmonic approximation might not be justified and vibrational modes can couple, which can result exchange in scattering power between two phonon modes. Surface-enhanced anti-Stokes Raman scattering originates from vibrational levels, which are populated by the very strong surface-enhanced Raman Stokes process: One photon populates the excited vibrational state, a second photon generates the anti Stokes scattering. Therefore, the anti-Stokes Raman scattering signal depends quadratically on the excitation laser intensity [9,17,18]. The two-photon process inherently confines the volume probed by surface-enhanced anti-Stokes Raman scattering compared to that probed by one-photon “normal” surface-enhanced Stokes scattering. Effective cross sections for two-photon anti-Stokes scattering are on the order of 10-42 cm4s, which is at least seven orders of magnitude larger than typical two-photon fluorescence cross sections.