Plasmon resonances of silver colloids studied by surface enhanced Raman spectroscopy E.C. Le Ru * , M. Dalley, P.G. Etchegoin The MacDiarmid Institute for Advanced Materials and Nanotechnology, School of Chemical and Physical Sciences, Victoria University of Wellington, P.O. Box 600, Wellington, New Zealand Available online 27 December 2005 Abstract It is shown that surface enhanced Raman scattering (SERS) average signals and fluctuations can be used to characterize the spectral dependence of plasmon resonances, and are particularly sensitive to the local fields, as opposed to other optical techniques. SERS can be used to probe both the plasmon resonances and the mechanisms of optical amplification. Ó 2005 Elsevier B.V. All rights reserved. PACS: 78.67.n; 78.20.Bh; 78.67.Bf; 73.20.Mf Keywords: Surface enhanced Raman spectroscopy; Plasmon; Metallic surfaces; Local field 1. Introduction Surface enhanced Raman scattering (SERS) [1] can become an analytical tool with many applications. There is renewed interest in SERS for its potential to observe sin- gle molecules [2]. Despite this, the mechanisms responsible for SERS are still not fully understood. The large optical enhancements in SERS are believed to be due to single or collective plasmon resonances. The SERS signal intensity is given by I SERS ¼ M ðx L ÞM ðx R ÞI RS ; ð1Þ where I RS is the signal in non-SERS conditions, x L and x R are the frequencies of the exciting laser and of the emitted photon, respectively, and M(x) is the local field intensity enhancement due to plasmon resonances: M ðxÞ¼ E loc ðxÞ E 0         2 . ð2Þ Plasmon resonances of SERS active media are usually characterized by the optical extinction, although it is very difficult to extract from these measurements in the far-field information about local field enhancements. Near field techniques can be used to probe the local field [3] but are not always applicable. Here, we show that SERS can be used as a probe of plasmon resonances. In Eq. (1), the SERS signal gives an indirect measurement of the resonance profile. One advan- tage is that it directly probes the local field at the metallic surface, where the molecules are adsorbed. A potential problem is that the signal usually comes from several mol- ecules with different field enhancements, a situation avoided when single molecule SERS is possible. Also, the measurement is indirect since we measure a convolution at two frequencies. For a typical dye the Raman shifts of the main peaks range from 400 to 1800 cm 1 for Stokes (S) processes. To extend this, it is also possible to measure anti-Stokes processes (AS). Finally, it is also possible to look at overtones. Using lasers at 514 and 633 nm, most of the visible wavelengths can then be probed. In this work, we studied SERS from Rhodamine 6G (RH6G) in colloidal silver solutions. The fluctuations of the signal 1567-1739/$ - see front matter Ó 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.cap.2005.11.030 * Corresponding author. Fax: +64 44635237. E-mail addresses: eric.leru@vuw.ac.nz (E.C. Le Ru), Pablo.Etche- goin@vuw.ac.nz (P.G. Etchegoin). www.elsevier.com/locate/cap www.kps.or.kr Current Applied Physics 6 (2006) 411–414