The interpretation of photoemission-spectromicroscopy studies of Au on GaSe requires a revision of established ideas about this interface, which has long been considered a prototype of Schottky-like systems. We find that the interface-formation process involves strong substrate-overlayer interactions, the release of free Ga, and the formation of interface species, and leads to a barrier height in total disagreement with the Schottky model. Furthermore, the space-resolving capabilities of our instruments revealed lateral inhomogeneities of the local overlayer thickness and of the local band bending.

Photoelectron energy distribution spectra taken for the first time on micrometer-sized areas of cleaved GaAs(110) reveal rigid shifts from location to location in the photoemission core level peak energies, indicating band-bending changes on a microscopic scale.

No abstract

Scanning-photoemission-spectromicroscopy data revealed substantial inhomogeneities in the lineup of the electronic states at the interface between the two semiconductors GaSe and Ge. These inhomogeneities would lead to valence-band discontinuity changes from place to place, whose magnitude is approximately 0.4 eV.Scanning-photoemission-spectromicroscopy experiments with undulator synchrotron radiation revealed a lateral modification of the energy lineup of the electronic states between two semiconductors.We performed this test on Ge deposited on GaSe; the equivalent valenceband discontinuity, AE"reproduced previously published values' for a visually defect-free region, but was smaller by =0.4 eV, well beyond the experimental uncertainty, in another region. This result shows that the theory of heterojunctions must take into account the possibility of the coexistence of different energy lineups and, therefore, of different band discontinuities at the same interface.For many years, semiconductor heterojunction band lineups have been one of the central problems of condensed-matter physics, because of the technological importance of the resulting band discontinuities, but also because of their fundamental importance. The theoretical understanding of the lineups is indeed a complicated conceptual problem, which requires, in turn, a good understanding of the absolute energy scale for the two band structures, plus an equally good understanding of the interface electronic structure. This problem has been extensively investigated both with theory and with experiments. ' Most of the experiments have been based on photoemission spectroscopy.Conventional photoemission spectroscopy, however, cannot clarify the important issue of the possible lateral dependence of the band lineup:Are the band-edge discontinuities different from place to place?We decided to explore the possible existence of lineup inhomogeneities by using the technique of scanning photoemission spectromicroscopy, by means of the instrument MAXIMUM at the Wisconsin Synchrotron Radiation Center in Madison. MAXIMUM is a scanning soft-x-ray photoemission microscope that focuses undulator radiation using a multilayer coated Schwarzschild objective. This instrument can take both energy-resolved photoelectron intensity two-dimensional images and photoemission spectra on microscopic areas. The images are formed by scanning the sample and collecting the photoelectrons. The size of the microscopic areas from which the photoemission spectra are taken depends on the experimental setup; in our experiment, it was 5 X 5 pm . The focusing concentrates approximately 35% of the total photon-beam intensity in the spot, and the rest is in the diffuse background. The density of photons in the primary beam is of the order of 10' photons/cm /s, whereas that of the scattered light is nearly five orders of magnitude smaller. The focus of the analyzer (a doublepass cylindrical mirror analyzer) is on the primary spot and only 4% of the total signal produced by the scattered light reach...

No abstract