Reexamination of the Si(100) surface reconstruction

Arguments in favor of 2 x 2-type reconstructions and against a purely 2 x 1 reconstruction on the Si(100) surface are presented. These are based on (1) total-energy calculations, (2) the extra ((1/2), (1/2)) diffuse intensities observed in He and electron diffraction experiments, (3) a comparison of the experimental and calculated surface state band dispersions, and (4) the observation in angle resolved photoemission measurements of two surface states with equal binding energies but differing by a ((1/2), (1/2))-type wave vector.