“…Marked on the figure are the bulk and surface plasmon peak energies for Si(111)-7X7. 15 The spec trum is remarkably similar to previous EELS results obtained with a conventional electron gun (normal to the surface) and a cylindrical mirror analyzer with a large acceptance angle; 15,16 in this case the analyzer integrates over the paral lel component of the electron wave vector, q 1 , leading to a broadening of the loss peaks. In our case, the acceptance angle of the hemispherical analyzer is well defined (3°), but the analyzer integrates over a range of electron trajectories from the tip (differing trajectories become indistinguishable when far from the tip apex).…”