Abstract
Recently interest has arisen in the possibility of significant deviations of the optical properties of metals from their random-phase-approximation (RPA) values. Starting with the approximation that the crystal pseudopotential can be regarded as weak, we have carried out calculations beyond RPA of the optical absorption and primary photoelectron kinetic-energy distribution. Dynamic screening (as opposed to the usual RPA static screening) results in corrections of great importance to the phonon (or disorder) contribution to the optical absorption and makes qualitative changes in the phonon-induced photoelectron distribution for photons having energies above wp. A new feature is a marked peak in the phonon contribution to the photoelectron distribution at the Fermi energy for photons having energies in the region of plasmon energies. The supposition of a weak crystal potential prevents quantitative application of these results except to Na and K, but the qualitative effect of dynamic screening should be the same in more complicated systems. The possibility of the experimental existence of photoemission features inexplicable in RPA has been recently discussed by Phillips.