Collective electronic excitations of metal surfaces have been studied with high resolution electron energy loss spectroscopy for over three decades. This technique is particularly well suited for the investigation of reciprocal space thanks to the relatively small angular acceptance of the spectrometers, which translates into a small integration over reciprocal space when operated at grazing incidence or when the coupled with an SPA-LEED optics. These studies evidenced the negative (positive) dispersion of the ordinary surface plasmon on simple (noble) metal surfaces: Such phenomenon which could be nicely interpreted by theory in terms of the increasing confinement in the near surface region of the electric potential associated with the induced charge with exchanged momentum. The very sharp surface plasmon loss of Ag allowed to evidence an anisotropy of the dispersion coefficient with respect to crystallographic face as well as direction. The latter effect is described by Liebsch’s theory but the prediction is for a reversed sign of the anisotropy with respect to the experimental outcome. Further data were recorded for Pd(110) showing a strong negative dispersion of the plasmon. To my best knowledge, this effect has so far not been reproduced by theory. Additional, so far unexplained, excitations have been reported for both Ag and Pd.
More recently, theory predicted the existence of the acoustic surface plasmon (ASP) associated with the excitation of surface Shockley states. In this case experiment followed theory, confirming the latter for the closed packed surfaces of Be, Au and Cu. Against expectation, however, no ASP was observed experimentally for Ag(111) for which very different predictions are present in literature and no theory is available for surface Shockley states located at the border of the two-dimensional Brillouin zone. A loss with approximatively linear dispersion is apparent indeed for Ag(100), however, at unexpectedly large energy loss values.