In temperature dependent studies, it was possible to determine conditions under which nucleation sites are correlated so that high-quality, single-crystal epitaxial layers will result. GaAs, InP, and the As/InP growth surface have been studied by Fuoss and collabo

rators.

Toshio Takahashi of the Institute for Solid State Physics, University of Tokyo, reported studies of the growth of several metallic systems on Si(111), using similar techniques. Detailed structures were reported for partial monolayers, full monolayers, and multiple monolayers of B, Al, In, Sn, Ag, Sb, and Bi. By studying the angular symmetry of the x-ray scattering about the Bragg angle, the vertical distance of the overlayer above the substrate is obtained, and the phase change of the x ray upon Bragg reflection is determined. A highly detailed structure of the metal overlayer is obtained, but the experiment requires 12 days of data taking! The large increase in intensity which will be available at Spring-8 will be welcome, indeed, if many such studies are to be made.

Another method for studying atomic location, and also obtaining the thermalvibration amplitudes of surface atoms, was presented by Jamshed Patel of AT&T Bell Labs. Patel uses the x-ray standing wave, as obtained from the xray fluorescence that occurs concurrently with XRD. As the sample is rocked slightly through the Bragg condition or the x-ray wavelength is scanned through the Bragg condition, the XRD and the x-ray fluorescence are monitored. The x-ray standing wave within the crystal surface layers moves in phase, and the fluorescence yield varies according to which layer of atoms lies at the peak of the exciting standing wave. Thermal vibration of the atoms determines the width of the angular or wavelength range of fluorescence excitation. The results are analyzed using the Debye model, which shows quite low Debye temperatures (<100 K) for surface atoms.

In his crystal ball, Patel sees great advantages for higher intensity x-ray sources of the future. Experiments on not-so-perfect crystals, experiments nearer 90° Bragg angles, use of higher monochromator resolution, and studies of low-Z adsorbates will all become possible and offer exciting scientific opportunities.

An excellent presentation of x-ray absorption fine-structure (XAFS) studies of organic adsorbents on Ni(100) was presented by Toshiaki Ohta, currently of Hiroshima University but soon to take the professorship of the retiring Haruo Kuroda at Tokyo University. Ohta used both the x-ray absorption near-edge spectra (often referred to as XANES or NEXAFS) and the extended XAFS (EXAFS), along with temperature variation from cryogenic to 1,000 K, in his studies of thiophenol, thiophene, and CS, adsorbates on Ni(100).

These experiments were carried out near the K absorption edge of sulphur near 2.5 keV, a soft x-ray spectral region that has been difficult to access in the past, due to the need for large d-spacing monochromator crystals and a vacuum environment. Ohta's results illustrate the danger in extrapolation of behavior of one molecular adsorbate to another, even similar adsorbate. The surface structure, effective coordination, dissociation, and desorption of thiophene and thiophenol on the Ni surface behave quite differently as a function of temperature.

diffraction (PED) holography, and photoelectron microscopy in a threepart presentation.

In ultra-ESCA, Tonner pointed out that intrinsic line widths for low-Z elements are as low as 10 meV, which can be observed using very high resolution x-ray excitation instead of the customary electron beam excitation.

PED holography, usable on metal, insulator, or semiconductor surfaces, has the capability of imaging the surface with a resolution of 0.05 to 0.5 Å, which is competitive with scanning tunneling microscopy. The technique detects the angular dependence of the photoelectrons emitted from the surface as a hologram of the surface. Holographic reconstruction of the surface produces a high-resolution, threedimensional image. This image is compared with computed holographic images to obtain quantitative atomic positions and to eliminate artifacts.

As examples of PED holography, Tonner showed results of studies of the Cu(111), Ir(111), and Cu/Ir(111) surfaces. These results indicate the promise of x-ray-generated holographic techniques and illustrated the importance of increased x-ray flux if these techniques are to be used extensively in surface science.

Both photoemission microscopy and x-ray absorption microscopy were discussed by Tonner. In the former case, Tonner showed studies of a patterned layer of Al, 50 nm thick, on GaAs. As an example of x-ray absorption microscopy, Tonner showed results of studies of yttrium-barium-copper oxide using the Ba 4d x-ray absorption edge. The ability to select x-ray wavelengths above or below core-absorption edges, or at peaks of near-edge x-ray absorption spectra, and to microscopically image that data will help answer many detailed chemical and physical questions about surfaces of importance technologically.

Masaharu Oshima, self-styled "Samurai Spectroscopist" of the NTT Interdisciplinary Research Laboratory,

Tokyo, showed his application of photoemission spectroscopy to three interfacial systems of technological importance: CaF/GaAs, metals on sulphurpassivated GaAs, and InAs/SF EuBa,CuO, Using photoemission near the Ca(3p) absorption edge at 114 eV, Oshima studied the surface construction versus growth temperature during epitaxial growth of CaF2on GaAs(111)A & B, and GaAs(100). Ultimately, Oshima obtained an MS (metal-semiconductor) structure that was free of pinning on the GaAs(111)B surface.

Additional studies of the noise performance of this MS field effect transistor (FET) were made using x-ray standing wave excited fluorescence spectra of the S(K) absorption edge. The locations of the S atoms at the interface were determined for GaAs(100), GaAs(111)A, and GaAs(111)B surfaces. Binding-energy shifts of the S-Ga bonds were obtained and correlated with 1/f noise associated with the interfacial region.

The effects of sulphur passivation on aluminum layers on GaAs were studied using similar techniques. It was found that metallic Ga is formed, along with Al-S bonds. These surfaces appear to have considerable promise in GaAs electronic devices.

Finally, Oshima discussed his exciting studies of superconducting transistor structures: GaAs/EBCO/InAs, where the GaAs will serve as the emitter, EBCO (EuBa2Cu,O,) as the base, and InAs as the collector. First, GaAs was oxidized and EBCO deposited. Then SrF, was deposited, and finally InAs was deposited. In depositing the InAs, only three monolayers were deposited

at the low temperature of 200 °C, followed by the In layer before heating to the deposition temperature of InAs. The SrF, interlayer prevented the formation of InO prior to the higher temperature InAs deposition.

In the future, a higher photon flux from Spring-8 will allow combination measurements, real-time analysis of growth situations, imaging of growing structures, and spatial resolution of characterization measurements.

Studies of surface and thin-film magnetism were discussed by Jürgen Kirschner of the Freie Universität Berlin. Kirschner used polarized and unpolarized VUV synchrotron radiation to excite spin-polarized photoelectrons from thin films of Fe(100), Cu(100), Co/Cu(100), and multiple layers of Cu and Co. These studies show that a few monolayers of Co on Cu induce polarization of the Cu(3d) electrons and the Cu s-p bands. Co films nine monolayers thick show square domains polarized along the <110> directions.

In the Fe films, the Fe(2p) levels at 680 and 695 eV are split by spin-orbit interaction into spin-up and spin-down components. However, spin splitting of S-core states is very small. Kirschner reported that the spin polarization of s- and p-levels is not well understood theoretically.

The question of a possible spinpolarized scanning tunneling microscope (STM) was brought up by a questioner. Kirschner indicated that, although he has thought about it, no ideas have occurred yet. He reported a rumor from Basel, Switzerland, that someone is working on such a device, but emphasized that the rumor has not been verified.

The final paper of the conference was presented by Yoshitada Murata, Institute of Solid State Physics, University of Tokyo. He summarized the study of surface and interface science. Up to the present, problems undertaken have been oriented toward the study of interface physical structure, electronic structure, and interface absorption spectra. A prime example is the Si(111) 7x7 surface, including the second layer studies by STM.

Murata projects that in the future problems undertaken will utilize angleresolved and time-resolved synchrotronradiation techniques such as ARUPS (angle-resolved ultraviolet photoemission spectroscopy), SXAFS (surface x-ray absorption spectroscopy), spinpolarized (SP) UPS, or SPARUPS for the study of complex structures such as O/Ni(100) or for the production of new materials utilizing surface properties. Dynamical processes, phase transitions, surface reconstruction dynamics, and absorption-induced restructuring will be important areas for research effort.

Absorption, diffusion, desorption, and reaction will be studied from the quantum-mechanical viewpoint. Nearly elementary processes, thermal reactions, and nonthermal processes will be important to study.

Photodesorption and photonstimulated desorption will be pursued in order to understand high-reactionrate processes with high selectivity. Murata suggested a strong need for exclusive SR beamline facilities or for compact, individual-laboratory SR sources for VUV experiments, especially for complex systems involving laser and SR beams and the like.