鈴木 哲

A vacuum ultra-violet beamline for in situ angle-resolved photoemission spectroscopy of MBE-grown high-T-c superconductors at low temperature has been constructed at beamline ABL-6B of the normal-conducting ring in the synchrotron radiation facility of the NTT Atsugi R&D Center. The constant-deviation-angle varied-line-spacing plane grating monochromator covers the energy range of 20-200 eV by using two gratings. A photon flux of the order of 10(11) s(-1) with a resolving power of 2000 or more was achieved in the whole energy range. The endstation is equipped with an angle-resolved photoelectron spectrometer, an rf-stimulated He discharge lamp, a custom-designed sample manipulator for low-temperature measurements and a loadlock system for in situ measurements. Total energy resolution of about 13 meV has been obtained for the photoemission spectrum of the Au Fermi edge, in spite of the bending-magnet light source of a second-generation synchrotron radiation ring. (c) 2005 Elsevier B.V. All rights reserved.

Work functions of individual single-walled carbon nanotubes (SWNTs) were studied by means of photoemission electron microscopy. Work function differences between the nanotubes were clearly observed in secondary electron images. The work functions of 93 SWNTs were found to range within 0.6 eV, but most distributed in a much narrower energy range of 0.2 eV. The work functions of single-walled nanotubes do not seem to have large structural dependence. (C) 2004 American Institute of Physics.

The substrate temperature dependence of the chemical state of Si substrate surface and metal catalysts for the growth of carbon nanotubes (CNTs), Co and Fe, and their morphology were investigated using soft X-ray photoelectron spectroscopy. We found that the temperature at which the chemical states of Co changes depends on the initial thickness of the underlying SiO2 layer. In the case of Fe thin film on a thin SiO2 layer, temperatures of transformation to metal and silicides are higher than those for the Co layer. In addition, Co silicides coexist with the SiO2 layer on the Si substrate at the substrate temperature of 590degreesC, but Fe silicides do not. These results indicate that, in the case of CNT growth by chemical vapor deposition, we have to contrive different ways to control the chemical states of metal catalysts for different substrate surfaces and metal catalysts. (C) 2004 Elsevier B.V. All rights reserved.

Single-walled carbon nanotube networks grown on SiO2 pillars were studied by means of scanning photoemission microscopy. The individual nanotubes or nanotube bundles growing from the pillar tops were observed in C 1s images. Band bending near catalytic Fe/nanotube contacts in an end-bonded configuration was studied by measuring C 1s spectra along the tube axes. Within our experimental resolution, no band bending was observed. This implies that the depletion width is less than the spatial resolution of the scanning photoemission microscope (90 nm) or that the amount of the band bending is less than 0.1 eV. (C) 2003 Elsevier Ltd. All rights reserved.

Individual single-walled carbon nanotubes (SWNTs) produce highly bright images of the insulator surface around them when observed by scanning electron microscopy at low primary-electron voltage. We found that the insulator surface near SWNTs emits more secondary electrons due to electrons supplied through SWNTs connecting to the outside area of the primary-electron beam scanning. SWNTs are thus highlighted as bright lines corresponding to the electron-beam-induced current range around them. This technique provides a useful and effective way to investigate lateral growth morphology of SWNTs on the substrate. (C) 2004 American Institute of Physics.

Low-acceleration-voltage electron irradiation effects on single-walled carbon nanotubes were studied by resonant Raman spectroscopy. The irradiation at acceleration voltages of 0.5 to 25 kV was found to extinguish the characteristic optical property of the nanotubes and reduce their tolerance against annealing in air, indicating that the nanotubes are inevitably damaged by ordinary scanning electron microscope observation. The acceleration voltage of around 1 kV caused the most extensive damage. Less defective SWNTs were found to have a higher tolerance against the irradiation damage.

We have performed in situ angle-resolved photoemission spectroscopy on the electron doped cuprate superconductor La2-xCexCuO4 (x = 0.1) using high-quality films grown by molecular beam epitaxy. The obtained photoemission spectra were free from "dirt" peaks. In addition, the time evolution of the spectra due to impurity adsorption is rather gradual, and the sample surface is stable even at ambient temperature, which is in contrast with previous reports on bulk Nd2-xCexCuO4 (NCCO) single crystals. The energy distribution curves show clear dispersions near the Fermi energy (E-F) with some E-F crossings. The resultant Fermi surface is hole-like and centered around (pi, pi), which essentially agrees with that predicted for NCCO by band calculations and also with the experimentally determined Fermi surface of optimally doped NCCO (x = 0.15). (C) 2004 Published by Elsevier B.V.

We adopt a multiwalled carbon nanotube target to increase the efficiency of water-window and K-alpha X-ray pulse conversion from femtosecond-laser-produced plasma. The diameter of the carbon nanotubes is around 30 nm and the length is about 12-mum. The X-ray fluence enhancement in the water-window region is sevenfold compared with a conventional carbon plate target. Further enhancement can be expected by optimizing the size of the carbon nanotubes. Soft X-ray pulse duration is 26 ps. It is also found that the K-alpha X-ray line emission from the Si substrate of the carbon nanotube target was enhanced. This result indicates that by covering various solid materials with carbon nanotubes, enhanced short K-alpha X-ray pulses with the corresponding wavelength can be obtained. These results show that carbon nanotubes are very attractive as a target for femtosecond laser-produced-plasma X-ray sources in single-shot X-ray microscopy and time-resolved X-ray diffraction.

Electronic structural change of single-walled carbon nanotubes induced by potassium encapsulation was studied by photoemission spectroscopy. The potassium encapsulation caused a shift of the overall valence-band spectrum toward the higher binding-energy side by about 0.5 eV, which is basically understood by the simple rigid band shift model. However, the spectral intensity increase observed near the Fermi level was much larger than that expected by assuming the simple pi(*) band filling, indicating that, in addition to the pi(*) band, a part of the density of states initially located in the unoccupied states dips below the Fermi level by the potassium encapsulation. The result is qualitatively consistent with a recent band-structure calculation, which predicted that the nearly free-electron state hybridized with the K 4s state crosses the Fermi level. The potassium encapsulation also decreased the work function by 1.4 eV.

Electronic structural change of single-walled carbon nanotubes induced by potassium encapsulation was studied by photoemission spectroscopy. The potassium encapsulation caused a shift of the overall valence-band spectrum toward the higher binding-energy side by about 0.5 eV, which is basically understood by the simple rigid band shift model. However, the spectral intensity increase observed near the Fermi level was much larger than that expected by assuming the simple (formula presented) band filling, indicating that, in addition to the (formula presented) band, a part of the density of states initially located in the unoccupied states dips below the Fermi level by the potassium encapsulation. The result is qualitatively consistent with a recent band-structure calculation, which predicted that the nearly free-electron state hybridized with the (formula presented) state crosses the Fermi level. The potassium encapsulation also decreased the work function by 1.4 eV. © 2003 The American Physical Society.

We adopt a carbon nanotube target to increase the efficiency of water-window X-ray pulse conversion from femtosecond-laser-produced plasma. The target is an array of vertically aligned multiwalled carbon nanotubes, each 30 nm in diameter and about 12-mum long. Center-to-center nanotube distance is around 150 nm. The X-ray fluence enhancement in the water-window region is seven-fold compared with a conventional carbon plate target. Further enhancement can be expected by optimizing the size of the carbon nanotubes. X-ray pulse duration is 26 ps. The results show that carbon nanotubes are very attractive as a target for femtosecond laser-produced-plasma X-ray sources in single-shot X-ray microscopy.

Electronic structural change of single-walled carbon nanotubes induced by potassium encapsulation was studied by photoemission spectroscopy. The potassium encapsulation caused a shift of the overall valence-band spectrum toward the higher binding-energy side by about 0.5 eV, which is basically understood by the simple rigid band shift model. However, the spectral intensity increase observed near the Fermi level was much larger than that expected by assuming the simple pi(*) band filling, indicating that, in addition to the pi(*) band, a part of the density of states initially located in the unoccupied states dips below the Fermi level by the potassium encapsulation. The result is qualitatively consistent with a recent band-structure calculation, which predicted that the nearly free-electron state hybridized with the K 4s state crosses the Fermi level. The potassium encapsulation also decreased the work function by 1.4 eV.

Passivation-mediated growth was used to control the interface crystalline order and the magnetic continuity in epitaxial Co films on several GaAs(001) surfaces. Using Se, S, and O passivated layers on GaAs before Co deposition, we observed a variety of interface structures depending on the substrate condition. With the help of transmission electron microscopy (TEM) images, we show that a perfect and well-ordered crystalline growth at the initial stage of deposition for Co/Se/GaAs ensures a better growth for thicker film. The magnetic properties of Co overlayers on all the substrates were studied by magnetic linear dichroism in angular distribution (MLDAD) in photoemission and found to be different depending on the film crystallinity. By combining the TEM and MLDAD results, we conclude that Co film on Se/GaAs, which has an ordered interface, is a good candidate for use as a spin filter from the viewpoint of spin-injection mechanism. (C) 2002 American Institute of Physics.

The Cs intercalation process in multiwalled carbon nanotubes (MWNTs) was studied by cross-sectional scanning photoemission microscopy. Cs atoms initially deposited on the tips of aligned nanotubes diffused toward their roots. The Cs diffusion constant for the MWNTs at room temperature was evaluated from the Cs distribution measured along the axes of the tubes. The value of 2x10(-12) cm(2)/s obtained is seven orders of magnitude smaller than that in graphite, although the local atomic structure of an intercalated MWNT is very similar to that of intercalated graphite. (C) 2002 American Institute of Physics.

The electronic structure of multiwalled carbon nanotubes aligned perpendicularly on a Si substrate was studied by means of photoelectron spectromicroscopy. The valence band and the C 1s spectra, measured systematically from spatially selected regions along the tube axes, were the fingerprint for lateral variations in the electron density of states and in the band bending, respectively. It was found that the tips have a larger density of states near the Fermi level than the sidewalls, whereas band bending, which would explain such a spectral difference, was not observed. It is suggested that the different density of states near the Fermi level is due to a larger dangling bond density at the tips.

Electronic structural changes in multiwalled carbon nanotubes caused by adsorbates and Cs deposition were studied by photoemission spectroscopy. The adsorbates formed by air exposure slightly increased the work function. The increased work function is considered to be due to the surface dipole moment induced by the negatively charged adsorbates. On the other hand, the Cs deposition followed by the intercalation drastically decreased the work function. The results suggest that field emission current enhancement induced by adsorbates and by Cs deposition are based on essentially different mechanisms; that is, the former is associated with the formation of localized electronic states and the latter is mainly due to the work function decrease.

The final goal of nanostructure integration based on self-assembly is full-wafer design of whole atomic-level structures. Toward this goal, we must first be able to control atomic steps, reconstructed domains, surface strain, and atomic species. Atomic steps can be rearranged artificially in a large area using lithographic technique and we are now close to achieving complete control of step positions. Patterns of reconstructed domain regions can be ordered by self-organization. In nanostructure self-assembly, such as the coherently grown Ge quantum nanostructures on Si(0 0 1) and Si(1 1 3) surfaces, strain engineering is important for controlling position, shape, and distribution. Ordered Ge-island chains on Si(0 0 1) show that artificial strain distribution design is a powerful tool for nanostructure integration. Surface composition on SiGe mixed surfaces can be reversibly changed by hydrogen adsorption and desorption. These approaches to designing surface structures show that the bottom-up approach is a promising alternative in semiconductor integration technology. (C) 2002 Elsevier Science B.V. All rights reserved.

Photoemission spectra were measured from the tips and the sidewalls of multi-walled carbon nanotubes grown on Si substrates. The results show that the Fermi level is located slightly inside the conduction band at the tips. We suggest that the Fermi level pinning is due to a large defect density at the tips of the multi-walled carbon nanotubes.

Using atomic force microscopy and scanning electron microscopy, we demonstrate that a Co epitaxial film surface grown on S-passivated GaAs(001) is smoother than one grown on bare GaAs(001) and establish that the surfactant nature of sulfur plays a vital role in the formation of a smooth surface. Synchrotron radiation photoemission spectroscopy results confirm that S passivation greatly reduces the segregation of substrate atoms during film growth on a S-passivated surface. It was also found that methanol rinsing after chemical S passivation provides an even smoother surface. (C) 2001 American Vacuum Society.

The electronic structures and the work functions of pristine and Cs-intercalated single-walled carbon nanotube bundles were investigated by C 1s and valence band photoemission spectroscopy. The C 1s spectrum of the pristine material showed a Doniach-Sunjic type asymmetry, indicating the existence of metallic tubes. The work function of the pristine bundles was found to be 4.8 eV, which is about 0.2 eV larger than that of graphite. A drastic decrease of the work function to about 2.0 eV was observed in the Cs-intercalated sample. The Cs intercalation also caused a nearly two-order increase in the spectral intensity at the Fermi level. (C) 2001 Elsevier Science B.V. All rights reserved.

Ga 3d two-hole states were observed from GaAs(lll)A-(2X2), (001)-c(4x4), (001)-(2X4), and (110)-(1X1) surfaces using Ga 3p --> 4s resonant photoemission spectroscopy. The Slater integrals F-2 and F-4 Of the Ga 3d orbitals were determined to be 14.4 and 6.0 from the relative energy positions of multiplet terms (1)G, D-1, and F-3. Surface components of the two-hole states were observed in the spectra of the (111)A-(2X2), (001)-(2X4) and (110)-(1X1) surfaces involving Ga atoms at the surface. The intensities of the surface components of the two-hole states were found to be considerably smaller than those of the d(9) states. This is ascribed to the reduced absorption probability of the 3p electrons at the surface caused by surface relaxation. (C) 2001 Elsevier Science B.V. All rights reserved.

Due to the S passivation, a modified growth of Co on GaAs(001) has been found. Using reflection high-energy electron diffraction and transmission electron microscopy, we observed the formation of a hcp Co overlayer of approximately 5 nm thickness on S/GaAs(001). In contrast, a similar 5 nm Co film on GaAs(001) shows a bcc structure. The metal-semiconductor interfaces in both systems were found to be different, where Co/S/GaAs(001) showed a relatively more abrupt interface. This epitaxial modification is explained on the basis of the morphology of the initial substrate surface, chemical compositions, the nature of the chemical reaction between adatoms and substrate atoms, and the effect of atomic segregation. (C) 2001 American Institute of Physics.

The valence band and C 1s photoemission spectra were measured from the tips and the sidewalls of multiwalled carbon nantoubes grown on Si substrates. The results show an overall spectral shift to the higher-binding-energy side, a larger density of states at the Fermi level, and a lower work function at the tips. These results can be explained by assuming that the Fermi level is slightly shifted upward and located inside the conduction band at the tips. We suggest that the electronic structure at the tips is considerably affected by defects.

The electronic structures and the work functions of pristine and Cs-intercalated single-walled carbon nanotube bundles were investigated using ultraviolet photoemission spectroscopy. The valence bands of the pristine bundles were considerably altered from those of graphite. A spectral shift to the higher binding energy side was observed in the Cs-intercalated sample. The work function of the pristine bundles was found to be 4.8 eV, which is 0.1-0.2 eV larger than that of graphite. A drastic decrease of the work function to about 2.4 eV was observed in the Cs-intercalated sample. (C) 2000 American Institute of Physics. [S0003-6951(00)05326-2].

The Ga 3d two-hole states were observed from GaAs(001)-c(4 x 4), -2 x 4, and (110)-1 x 1 surfaces using resonant photoemission spectroscopy. Multiplet terms (1)G, D-1, and F-3 were clearly distinguished at the photon energy around the 3(p) absorption threshold. The surface components of the two-hole states were observed in the spectra of the (001)-2 x 4 and (110)-1 x 1 surfaces having Ga atoms at the surfaces. The curve-fitting analysis revealed that the intensities of the surface component of the two-hole states were considerably smaller than those of the d(9) states. This is ascribed to the reduced absorption probabilities of the 3p electrons at the surfaces caused by the surface relaxation.

The electronic structure and chemical states of the overlayer and substrate in epitaxially grown Co films on Se-treated GaAs(001) 2 x 1 are investigated by photoemission spectroscopy (PES) using synchrotron radiation. Adatom (Co)-substrate (Ga) intermixing is found at the initial stages of Co deposition. The intermixed layer is then buried and epitaxial growth is achieved following further deposition. During the intermixing, Se is floated over the surface of Co overlayers. A small amount of As-segregation is also observed in the thicker Co films.

Effects of a Se treatment on the work function of epitaxially grown GaAs surfaces were measured by photoemission spectroscopy. Although the Se treatments reduced the band bending of n-type GaAs, the work function increased. This result indicates that the surface component of the work function increased and that the surface dipole moment, whose direction was from the surface to the bulk (outer layer: -, inner layer: +), was formed at the Se-treated GaAs surface. On the other hand, qualitative analysis based on the electron counting model revealed that the direction of the surface dipole in the previously proposed structural model of Se/GaAs(001) was completely opposite to that obtained in our experimental results. It is suggested that this discrepancy is due to the vacancies in the internal Ga layer in the structural model.

The electronic structure and the charge balance mechanism of the Li7-xMnN4 system were investigated by core-level electron energy loss spectroscopy. The spectra are strongly affected by configuration interactions and reveal the existence of a considerable amount of nitrogen 2p holes in the ground state. Based on the results, we conclude that the Li7MnN4 system can be classified into the charge transfer-type compounds. The systematic variation of the operating voltages versus Li/Li+ of the Li7MnN4 system and the other transition metal nitrides and oxides can be qualitatively explained by classification into the Mott-Hubbard and the charge transfer regimes. (C) 1999 Elsevier Science B.V. All rights reserved.

Single-walled carbon nanotubes (SWNTs) were synthesized by ablating a graphite target mixed with metal catalysts with a pulsed Nd:YAG laser. The quality and nature of the SWNTs produced depended sensitively on the ablation conditions. The average nanotube diameter was found to shift with the ablation laser frequency and the gas how rate. Carbon nanotube/polymer composites were fabricated by solution casting. A method was developed to align the nanotubes inside the polymer matrix with controllable orientation and degree of alignment. SWNTs were intercalated with alkali metals and HNO3 molecules. Intercalation and in-situ TEM/EELS measurements were also performed on individual nanotube bundles. The results showed that guest species can be reversibly intercalated to the interstitial sites between the nanotubes.

Single-walled carbon nanotubes (SWNTs) were synthesized by ablating graphite targets with either the primary (1064 nm) or the second-harmonic (532 nm) beam of a pulsed Nd:YAG laser at high temperature. The structure and the morphology of the raw materials were studied by high-resolution transmission microscopy (HRTEM), X-ray diffraction, and micro-Raman techniques. The diameter distribution of the SWNTs was found to vary with the laser frequency used for ablation. The raw materials were reacted with alkali metal (K, Cs) by vapor transport method. The saturation composition was found to be MC(8)s (M = K or Cs). No crystalline structure was observed in the reacted materials by X-ray diffraction. In situ metal deposition, TEM, and electron energy loss spectroscopy (EELS) measurements were performed on individual SWNT bundles at 300 K. The results showed that alkali metals can be reversibly intercalated into the SWNT bundles. Although intercalation induced structural disorder, individual nanotubes and to a large extent the bundles maintained their structural integrity after intercalation and de-intercalation.

Cesium (Cs) or potassium (K) was deposited on single-walled carbon nanotube bundles in vacuum at room temperature. The deposited bundles were analyzed in-situ by transmission electron microscopy and electron energy loss spectroscopy techniques, The results indicate that both Cs and K can be reversibly intercalated with the bundles. The intercalants reside in-between the individual nanotubes within the bundles. Intercalation caused structural disorder to the two-dimensional lattice of the pristine nanotube bundles. The chemical compositions of the nanotube bundles intercalated with K and Cs were found to be about KC24 and CsC24 to CsC8. (C) 1998 Elsevier Science B.V.

Charge distributions of lithium cobalt nitrides Li2.6Co0.4N and Li1.0Co0.4N were investigated using core-level electron energy loss spectroscopy. The N-K spectra of Li2.6Co0.4N and reference sample Li3N Showed that both lithium and cobalt are monovalent in Li2.6Co0.4N The N-K spectrum of electrochemically lithium-extracted Li1.0Co0.4N indicated that a large amount of holes are introduced into nitrogen 2p orbitals by the lithium extraction and that Li1.0Co0.4N reveals a strongly covalent character. The branching ratio of the Co-L spectrum of Li1.0Co0.4N was almost intermediate between those of typical oxides having high-and low-spin ground states. This is attributed to the mixing of several electron configurations due to the covalency between cobalt and nitrogen. (C) 1997 Elsevier Science Ltd.

The core level electron energy loss spectra of nominally tetravalent CrO2 and trivalent Cr2O3 were measured. The Cr-L multiplet structure of CrO2 did not show a distinctly different spectrum from that of trivalent Cr2O3, although transition metal L spectra generally vary strongly with the oxidation states. The difference of the oxidation states was clearly observed in the O-K spectra in contrast to the Cr-L spectra. The O-K energy loss near edge structures revealed that the lowest conduction band of CrO2 has a large oxygen 2p component. It is concluded that the weak valence dependence of the Cr-L spectra is due to the relatively large d(3) (L) under bar component in the electron configuration of CrO2.

Electron energy loss spectroscopy installed in a transmission electron microscope was used to investigate the electronic states of LiMnVO4. The Mn-L multiplet structure of LiMnVO4 and the typical manganese oxides clearly revealed that the manganese in LiMnVO4 is divalent. The V-L spectrum showed a characteristic multiplet structure in tetrahedrally coordinated-d(0) oxides, revealing that vanadium is pentavalent. The O-K spectrum was compared to that of LiMnPO4, and showed that the lowest unoccupied states of LiMnVO4 have mainly vanadium 3d character and that the bonding between vanadium and oxygen in LiMnVO4 is more ionic than phosphorus and oxygen in LiMnPO4. (C) 1997 Elsevier Science Ltd.

Electron energy loss spectroscopy combined with a transmission electron microscope was applied to valence analysis of transition metal ions in LiMn2O4, LiMnTiO4, LiMnCrO4 and LiMnCoO4 powder samples. The transition metal 2p multiplet structures were compared with those of appropriate reference samples. The Mn, Ti and Co valences were clearly determined from the 2p spectra as LiMn3+Mn4+O4, LiMn3+Ti4+O4 and LiMn4+Co3+O4, respectively. The O 1s spectra were also effective for analysis of Cr in LiMnCrO4 and showed that the Cr ion was trivalent. The electron configuration of one of nominally tetravalent chromium compounds, CrO2 is also discussed.

Second-stage potassium-intercalated carbon nanotubes were synthesized in a specially designed ultrahigh vacuum analytical electron microscope and their valence-band excitation spectra in the region of the pi+sigma plasmon were measured by electron energy loss spectroscopy The carbon nanostructures consisted of graphene sheets. Potassium was deposited in an ultrahigh vacuum at room temperature. ks a result, a second stage of intercalated nanotubes was found to be formed close to the surface. The energy loss spectra of the intercalated nanotubes showed humps at about 16, 19, and 22 eV, in addition to those of unintercalated tubes,This suggests that intercalation modified the band structure of the interlayer bands and/or the sigma(sigma*) bands. (C) 1996 American Institute of Physics.

Electron energy loss spectroscopy has been used to investigate the B-K spectra of individual filaments grown by electron beam irradiation. We also measured the energy loss spectra of beta-rhombohedral boron and B4C for reference. High-resolution transmission electron microscope images, micro-electron diffraction patterns, and elemental analysis by energy loss spectra revealed that the filaments consist of amorphous boron. The similarity of B-K near-edge structures between the filaments and the reference forms of boron suggests that the filaments have B-12 icosahedra as their basic structural unit, which is consistent with previous structural studies of amorphous boron. The different peak sharpnesses just above the B-K edge may reflect differences in structural complication and/or disorder of B-12 icosahedron units.

It is experimentally demonstrated for the first time that an a-Si:H photodiode with reach-through structure can detect infrared light of 1.31 mum and 1.55 mum. A maximum gain-quantum efficiency product of 0.58 is obtained at a reverse bias of -10 V under 100 muW illumination at 1.31 mum. This value of gain-quantum efficiency product is comparable to the quantum efficiency of a non-gain-enhanced a-Si:H pin photodiode at visible wavelengths.

It was found by photoemission and inverse photoemission that an energy gap with a finite density of states at the bottom (a pseudo-gap) opens at the Fermi level for alkali-doped solid C60 at the composition A3C60 (A = alkali metal). Strong electron correlation and lattice distortions (Jahn-Teller effect) are discussed as possible origins for the pseudo-gap.

Photoemission and inverse photoemission spectroscopy has been used on C60 and K3C60 to study the change of the electronic structure upon alkali doping and its relation to the occurrence of superconductivity. The experimental results show that alkali doping does not lead to a rigid filling of the lowest unoccupied molecular-orbital band by the doped electrons but causes a transfer of electronic states from that band to an additional band produced in the band gap. A comparison of the photoemission and inverse photoemission spectra indicates that a pseudo-gap of about 0.5 eV opens at the Fermi level in K3C60. Possible origins for the pseudo-gap and its relation to superconductivity are discussed.

A comparative photoemission study on RbxC60, RbxC70 and RbC8 (graphite intercalation compound) is reported. It was found that the Fermi level of Rb3C60 is located at the bottom of a valley of the density of states, while that of RbC8 is at a maximum point of the occupied states. This experimental result is explained in terms of the existence of a pseudo-gap at the Fermi level in the fulleride. This indicates that a rigid-band picture is not a good starting point for understanding the change of the electronic structure upon the alkaline-doping. It was also found that the density of states at the Fermi level is much smaller in RbxC70 than in RbxC60. This accounts for the less conductive nature and the absence of superconductivity in RbxC70. © 1992.

High-energy-resolved oxygen K absorption spectra of Bi2Sr2Ca1-xYxCu2O8 (x = 0.0 and 0.6) single crystals and YBa2Cu3O7-delta (delta = 0.0-0.7) ceramics were reported. The experimental results show that (1) the electronic states at the Fermi level in superconductor has a dominant O2px,y symmetry, (2) the lowest unoccupied states in nonsuperconductor is located about 1 eV above the Fermi level and also has a strong in-plane symmetry, and (3) hole-doping causes a transfer of weight of the electronic states from the lowest unoccupied states (upper Hubbard band) to the mid-gap states produced by hole-doping at the Fermi level.

Angle-resolved inverse photoemission spectroscopy was performed on a Bi2Sr2CaCu2O8 single crystal to study the electronic structure in the vicinity of the Fermi level. Inverse photoemission intensity shows two maximal about 0.2 GAMMA-X and 0.6-GAMMA-X from the GAMMA point along the GAMMA-X direction, suggesting that there are two bands crossing the Fermi level. The present experimental result was compared with angle-resolved photoemission results and band structure calculations.

The polarization dependence of the Cu L(III) absorption spectrum was measured on a Bi2Sr2CaCu2O8 single crystal with linearly polarized synchrotron radiation. The Cu L(III) absorption peak shows no energy difference between the z (3d3z2-r2) and the x,y (3d(x2-y2)) component, which is consistent with the electron-energy-loss study by Nucker et al., but is different from the previous x-ray-absorption studies by Bianconi et al. and Abbate et al. The relative intensity of the z component to the x,y component was estimated to be less than 5%, suggesting a very small contribution of the Cu 3d3z2-r2 orbital to the electronic structure in the vicinity of the Fermi level.

Photoemission spectroscopy was performed for solid C60 and its alkali-metal (K and Rb) compounds produced by successive vapor-deposition of alkali metal (K or Rb) on an evaporated C60 film. A series of photoemission spectra show a unique variation against the alkali-metal content which is not explained in terms of a simple band-filling picture with alkali s electrons. The correlation between the photoemission spectra and the electrical resistivity of the film was also studied.

Inverse photoemission spectroscopy has been performed on LaSrCuO, YBaCuO, and BiSrCaCuO to study the electronic structure of unoccupied region. It was found that the spectrum of Bi2Sr2CaCu2O8 has a clear Fermi edge, thereby giving a direct evidence for the existence of a Fermi surface (Fermi-liquid states), while those of YBa2 Cu3O7 and La1·85Sr0·15CuO4 exhibit little intensity at the Fermi level. A remarkable and prompt reduction of the Fermi-edge intensity was observed for the La-system. The reduction and absence of the Fermi-edge structure is ascribed to the surface degradation probably due to desorption of oxygen from the surface. © 1989.

An inverse photoemission spectrum of Bi2Sr2CaCu2O8 is presented. The spectrum shows a clear Fermi-edge structure in contrast to those reported for YBa2Cu3O7 and La1.8Sr0.2CuO4, giving direct evidence for the existence of the Fermi-liquid states in Bi2Sr2CaCu2O8. © 1989 The American Physical Society.