The inorganic chemistry of surface enhanced Raman scattering (SERS)

Karen Faulds, Aaron Hernandez-Santana and W. Ewen Smith

DOI: 10.1039/9781849730853-00001

1. Introduction

Surface enhanced Raman scattering (SERS) has enormous potential as a detection technique. Used correctly, it can combine many of the advantages of Raman spectroscopy such as stand off detection and selective identification of a species in situ, with the additional advantages of high sensitivity and even greater specificity.

The measurement of SERS involves adding an analyte to the roughened surface of a suitable metal and interrogating the surface using a Raman spectrometer. The metals most commonly used for practical SERS measurements are gold or silver. The reasons for this are that the electronic properties of gold and silver are suitable for SERS in that they have surface plasmons which lie in the visible region of the electromagnetic spectrum which coincides with the commonly used Raman excitation wavelengths, they have good resistance to corrosion allowing the surfaces produced to be reasonably long lived, and a large number of analytes adsorb effectively on them.

Raman spectroscopy is a selective technique, both because of the molecularly specific nature of the pattern of peaks obtained and because of the wide variation in the Raman cross section of different analytes. For example, water gives very weak Raman scattering and organic molecules usually have much larger scattering cross sections thereby enabling Raman scattering to be recorded from organic molecules in aqueous solution and allowing analytes to be identified in situ. However, the detection limits are usually high since Raman is an inherently weak process and surface enhancement provides much greater sensitivity. Surface enhancement was originally discovered by Fleischman when unusually intense Raman scattering was observed from pyridine adsorbed on a roughened silver electrode. The enhancement observed was calculated to be a factor of 106 over that expected for normal Raman scattering. It was subsequently discovered that the magnitude of the enhancement obtained with SERS can be increased further if the analyte used is a coloured molecule. In this technique, surface enhanced resonance Raman scattering (SERRS), the enhancement is due to a combination of surface enhancement and resonant enhancement obtained from molecular resonance from the coloured analyte when the wavelength of an electronic transition coincides with the laser excitation wavelength. Only Raman bands originating from the chromophore of the analyte are enhanced, which can afford selective detection of a resonant analyte in the presence of non-resonant contaminants. The enhancement factors claimed can be extremely high and have been calculated for a suitable analyte to be between 1013 and 1015. This form of spectroscopy has been shown in practice to rival or surpass fluorescence in sensitivity. However, the major advantage of SERS over fluorescence is the fact that SERS is a molecularly specific technique producing spectra which have sharp peaks whereas fluorescence spectra are broad and overlapping and less specific for a particular molecule. This enables much higher numbers of analytes to be discriminated in the one vessel by SERS/SERRS than by fluorescence where it is difficult to discriminate between more than 3 or 4 analytes in a mixture.

Thus, SERS is a label-less technology that has the advantage that many substances can be identified in situ and at low concentration but, if even higher sensitivity is required, this can be achieved by adding SERRS active labels. It should be noted that the labels need not be fluorophores or dyes, but can be any molecule which gives a strong SERS response and consequently a wider range of labels are available for SERS than for fluorescence techniques. A good example of this is that many haem compounds give very effective SERRS. These include chromophores which are native to proteins such as cytochrome C and cytochrome P-450.

Modern Raman spectroscopy is a simple and effective technique and this ease of use permits the development of SERS/SERRS methods where the advantages of selectivity, sensitivity and multiple analyte determination in a single cuvette or micro-titreplate well are required. Two difficulties have inhibited the growth of the technique and both are now better understood and both can be overcome. Firstly, SERS was originally discovered experimentally and the effect was poorly understood. Following much debate and many theoretical and practical studies the theory is now much better understood. The second difficulty concerns the need to obtain reliable and reproducible enhancement. Most roughened silver and gold surfaces which contain nanoscale features will give SERS and some will give very large enhancements. However, if the roughness features cannot be reproduced, it is not possible to obtain the same degree of enhancement for each surface making reliable analysis methods difficult to develop. A greater understanding of the nature of the effect has led to the design of reproducible and reliable substrates and the development of methods for obtaining reproducible results, overcoming some of the early problems with the technique. This article will initially discuss the basic theory required for practical use and then the main types of substrate that can be used to obtain reliable and reproducible results.

2. Theory

Many papers on SERS refer to two different effects which contribute to the SERS enhancement mechanism. The first, and dominant, effect is electromagnetic enhancement and the second is chemical or charge transfer enhancement.

Electromagnetic enhancement is observed when the analyte is on, or very close to, a metal surface. The analyte interacts with the surface plasmon, which is essentially a wave of electrons on the metal surface created by the interaction of the laser excitation with electrons bound to the surface. A surface plasmon created on a smooth surface does not scatter light since it is confined to the metal surface. To create scattering, the surface...