LAUSR

Abstract In this contribution we compared the efficiency in generation of Raman radiation in various conditions by three commonly used Raman reporters: carboxytetramethylrhodamine (TAMRA), cyanine 3 (Cy3) and cyanine 5 (Cy5) conjugated to model ssDNA adsorbed on gold and silver nanoparticles. Although such experiments are described in the literature as surface-enhanced Raman scattering (SERS) measurements, we found that in many cases, it is significantly more important to match the frequency of the excitation radiation to the resonant frequency of the Raman reporter than to the resonant frequency of the plasmonic system, when strong SERS effect is observed. For example, the relative intensity of the Raman radiation of TAMRA-labelled DNA on 20 nm gold nanoparticles measured using green excitation laser line (in this case SERS enhancement is very low) is significantly higher than the relative intensity of the Raman radiation recorded in experiments in which red excitation radiation has been used (in this case strong SERS enhancement is observed). Specific interactions between the metal nanoparticles and the deposited labelled DNA are also very important, for example, when green excitation laser line is used, Raman radiation generated from TAMRA-labelled DNA on gold is much higher compared to the TAMRA-labelled DNA on silver although silver nanoparticles are highly SERS active in these conditions, whereas gold nanoparticles are practically SERS inactive. Also, structures formed by model ssDNA with attached Raman reporters on silver and gold surfaces are probably different. Because in SERS sensors containing immobilised DNA there is usually a possibility of changing of many construction details, our studies may give some useful hints how to prepare such sensors which would work more effectively. The values of the increase in the efficiency of Raman scattering due to the resonance effect in comparison to the non-resonance conditions for adsorbed TAMRA and Cy3 labelled DNA have been also estimated.