SERS active sites, facts, and open questions

Abstract

Surface enhanced Raman scattering (SERS) by transient charge transfer (TCT) is a widely accepted mechanism. This first layer effect is not active at smooth surfaces, as shown from experiments and from metallic electron-hole pair (ehp) excitations after TCT. The first layer effect is rather the average over a sub-monolayer of "SERS active sites" (SAS), where ehp excitations are less likely. We review and present experiments using various physical methods at electrodes and at ultra high vacuum samples to prove the existence of SAS, which are probably situated within fractal monoatomic steps. As the first adsorbates in ultrahigh vacuum migrate to the SAS, scanning tunnelling microscopy can reveal their atomic structure. CO on vicinal Au(211) first settles down at the kink sites. NO dissociates and CO2 is activated on a subgroup of SAS of cold-deposited Cu, these processes are only observed by SERS but not by infrared reflection-absorption spectroscopy (IRRAS). Apparently, SERS focuses on "chemical active sites", whereas IRRAS monitors the majority species, which did not react. Here we propose two possible explanations.