We have systematically investigated the structural property and electrical structures of the transition metal intercalated titanium disulfide compound Fe$_x$TiS$_2$ ($0\leq x\leq0.33$) from angle-resolved photoelectron spectroscopy with tunable polarized synchrotron radiation. The effect of intercalation on the energy bands and density of states of the host material around two high-symmetry points ($\Gamma, L$) is studied. Charge transfer from the Fe atoms to the dichalcogenide layers leads to the movement of the Ti 3d-derived electron pocket. Two hybridized states are observed around $\sim500$ meV and Fermi level. Knowledge of hybridization among the Fe $3d$, Ti $3d$, and S $3p$ states is very important to understand the physical of Fe$_x$TiS$_2$ system.
Oxygen 1s soft X-ray absorption and emission spectroscopy (XAS and SXES) experiments have been carried out for layered MoO3 single crystal with three crystallography nonequivalent oxygen atoms; O1, O2, and O3 bonding with the Mo atoms along the b, c, and a axes, respectively. The polarization-dependent XAS results show that the conduction-band minimum of MoO3 consists of the O2 2pb components. The valence-band minimum (VBM) of MoO3 is, on the other hand, formed by the O2 2pa orbitals from the angular-dependent SXES spectra measured at the pre-edge of the O 1s absorption peak. The other SXES spectra have revealed that the O3 2pc states also contribute to the VBM and O1 2pb states locate at the bottom of the valence bands.
The temperature-dependent electronic states of FeSi have been studied by using high-resolution angle-resolved photoemission spectroscopy (ARPES) and using low-energy tunable photons. At low temperatures, a peak indicating the valence-band maximum (VBM) exists at a binding energy of similar to 20 meV along the Gamma R direction. The observed dispersional width of the energy band!; is narrower than that given by the band-structure calculation, and the width of the ARPES peak near the VBM rapidly broadens as the binding energy increases. Analysis of a model self-energy reveals the importance of electron correlation, especially near the VBM. We observed an unusual temperature dependence of the ARPES spectral features near the Fermi level (E(F)): Below similar to 100 K, the peak at the VBM and the energy gap structures are almost unchanged, while at similar to 100-350 K, the peak gradually moves toward E(F) and the gap is filled. The present results indicate that FeSi is a strongly correlated semiconductor, with a renormalized band near E(F) being responsible for the rapid collapse of the peak and the coherent energy gap upon heating.
Electronic structures of a layered semiconductor 1T-TiS2 and a Ni intercalation compound Ni1/3TiS2 are studied by means of Ti 2p and Ni 2p x-ray photoemission (XPS), 2p x-ray absorption (XAS), and 2p core excited resonance x-ray photoemission spectroscopies. It is found that the charge transfer from the guest Ni atom as well as from the S atom to the Ti atom is playing dominant roles in various spectra of Ni1/3TiS2. The origin of photoemission structures just below the Fermi level is interpreted. The importance of the electron correlation and hybridization effects is confirmed through the satellite features of the XPS and XAS spectra.
Monte Carlo simulations of atomic distribution of intercalated guest atoms in the layered lT-TiS2 have been performed by taking into account attractive or repulsive pairinteractions between the neighboring guest atoms in the a-axis plane (Val and Va2) and along the c-axis of the crystal lattice (Vcl and V2c); lattice size for computation: 18 × 18 × 6. X-ray diffraction patterns for FexTiS2 are calculated from the atomic distributions obtained using these four pair-interactions, in qualitative agreement with the experimental data of the 2a x 2a x 2c short-range ordered structure forズ= 0.15, 2寸3a x 2a x 2c superlattice for jc = 0.25, and 寸3a x J3a x 2c superlattice for x 0.333 with fractional site occupancy. From the calculated atomic distributions, we have evaluated the number of neighboring guest atoms, the formation of clusters, and percolation cluster, as well as its dimension, all of which are responsible for the dynamical relaxation behaviors of the thermoremanent magnetization observed in the spinand cluster-glass phases of Fe;cTiS2. In particular, oneand two-dimensional percolation clusters are formed by the third nearest neighbors in the a-axis plane for x = 0.15 and 0.25, respectively, and a two-dimensional one by the second nearest neighbors for x = 0.333, which corresponds to the magnetic phase diagram [spin-glass (x = 0.15) and cluster-glass (ズ= 0.25 and 0.333)]. Using the present results and EXAFS data, we have further discussed on the change in the local structures near the host sulfur atoms by intercalation of the guest atoms in FeJCTiS2 and CoxTiS2. Similar calculations have been made for another type of intercalation compound of stage-2 AgJCTiS2 (x = 0.15) that shows a phase transition from the ordered J3a x J3a x 2c to disordered structure around 250-300 K. The formation of stage-2 structure can be reasonably simulated by additional incorporation of repulsive pair-interaction, Vc2, along the c-axis direction, and by taking into account the considerably weak pair-interaction strengths, compared to those of Fe;cTiS2, in satisfactory accord with the observed X-ray patterns. Furthermore, the order-disorder phase transition can be understood by considering the temperature dependence of atomic distributions, the thermal displacements of Ti, S, and Ag atoms, and entropy change due to the phase transition. The essential difference between stage1 Fe^TiSj and stage-2 Ago 15TiS2 is discussed based on a local lattice deformation produced by intercalation. The size of the guest atom plays a crucial role to the formation of stage-1 or 0 2 compound; guest atoms with crystal radius larger than 1.0 A are found to produce preferentially the stage-2 structure in the host TiS2. Such a large deformation may be the origin to exert repulsive pair-interaction on the second nearest neighboring sites in the next layers.
Ln 4f partial densities of states (DOSs) in layered oxysulfide semiconductors (LnO)CuS (Ln=Ce, Pr, Nd) have been deduced by means of the Ln 4d–4f resonant photoemission spectroscopy. The Ce 4f DOS contributes just below the valence-band maximum (VBM) of (CeO)CuS, and on going from Ln=Ce to Pr, and to Nd, the 4f DOS shifts toward the deeper binding-energy side, leading no 4f contribution near VBM for Ln=Pr and Nd. The O 2p partial DOSs derived from O Kα emission spectra are almost unchanged with the Ln elements.
A high-resolution angle-resolved photoemission study of kish graphite using synchrotron radiation (hν=32 eV) has been conducted. We have directly observed a small Fermi surface (kF<∼0.05 Å−1) centered at the K point, and determined the π band group velocities along the K–M and K–Γ directions.
We have examined core-level photoemission spectra of 1T-TiSe2 and 1T-TaS2 to study chemical shifts associated with unusual superlattice formation. While large temperature-dependent core-level splitting was observed for TaS2, temperature-dependent chemical shift was not observed for TiSe2. On the other hand, angle-resolved valence-band photoemission spectra of TiSe2 showed appearance of a new branch (band-folding) at the L points in the ordered phase at 50 K. These results suggest smaller lattice distortion in TiSe2 than that in TaS2.
We present electrical resistivity (rho) measurements for the intercalation compound FexTiSe2 (0 <= x < 0.16) over the temperature range from 4.2 to 300 K, and angle-resolved photoemission spectra for x=0, 0.05, and 0.14 at 50 and 250 K (or 280 K). At 250 K, TiSe2 is a semimetal having hole pockets centered at the Gamma point and electron pockets around the L points of the Brillouin zone. Upon intercalation, Fe-derived flat bands appear just below the Fermi energy, and the Se 4p derived bands forming hole pockets at the Gamma point are lowered. At 50 K, band folding due to 2ax2ax2c superlattice is observed clearly near the L point for the host and x=0.05, while it vanishes for x=0.14, consistent with the rho-T data. The critical concentration for the suppression of the superstructure (0.05 < x(c)<= 0.075) can be explained reasonably well by the percolation threshold of a two-dimensional-trianglar lattice consisting of seven Ti atoms, which is estimated to 1/14 (=0.0714).
Soft x-ray photoemission spectroscopy for layered oxysulfides (LnO)CuS (Ln=Ce, Pr, Nd) has been curried out in the Ln 3d-4f absorption region, to deduce the Ln 4f partial densities of states (DOSs). Only the Ce 4f DOS contributes to the valence-band maximum of (CeO)CuS, and on going from Ln=Ce to Pr, and to Nd, the 4f DOS shifts toward the deeper binding-energy side due to the lanthanide contraction. The Ce 3d photoemission spectrum of (CeO)CuS shows the mixing of Ce3+ and Ce4+ in the ground state.
We measured the photoemission spectra of the IBr graphite intercalation compounds (IBr–GIC) with stage-2 and stage-4 structures at 16 K with incident photon energies hν=40–200 eV. The peak positions of the I 4d and Br 3d core-levels are unchanged for the stage-2 and stage-4 IBr–GICs. Partial density-of-states of the I 5p and Br 4p states in the valence bands have been evaluated by resonant photoemission spectroscopy. These spectra indicate a significant hybridization between the host and the guest IBr in the van der Waals gap.
Angle resolved photoemission spectra (ARPES) of host HOPG and IBr-GIC have been measured at 16K using hν=122eV. Both HOPG and IBr-GIC show clear dispersions of upper π band and σ bands derived from C 2s and 2p electrons. The intensity plot of ARPES spectra shows the overlap of the dispersion curves along the ΓM and ΓK directions due to the in-plane mosaic structure of HOPG. Based on the Johnson–Dresselhaus band model, the dispersion curves of the π bands of HOPG and IBr-GIC are reproduced over the full Brillouin zone. We discuss the important role of the interactions between C–C atoms in the neighboring layers in the stage-2 structure.
The electronic band structures of 1T-TiTe2 and Fe0.25TiTe2 have been clarified by angle-resolved resonant-photoemission spectroscopy. Upon Fe intercalation, while most of the band structure is unchanged, two new flat bands appear at the Fermi level and at E∼−0.5 eV. Constant-initial-state spectra indicate that these states are derived from both Ti 3d and Fe 3d states. While the Fermi wave number of the electron pocket around M point is reduced by Fe intercalation, those of hole pockets around Γ point are almost unchanged, suggesting that the additional electrons are accommodated by the new flat band at −0.5 eV.
A structural study of orthorhombic molybdenum trioxide MoO3 with a layered structure has been performed using synchrotron-radiation powder diffraction from 110 to 1000 K. MoO3 exhibits an anisotropic thermal expansion. With increasing temperature, the lattice expands remarkably along the a axis of the stacking direction, while it expands a little along the c axis but contracts slightly along the b axis. Based on the quasiharmonic approximation, the temperature variations in the lattice constants and thermal expansion coefficient are analyzed and they are reproduced quantitatively using three stretching mode frequencies for Mo-O bonds, 1000, 820, and 670 cm(-1), as well as two low frequencies, 250 and 100 cm-1. The large thermal expansion along the a axis is attributed to the coexistence of high frequency vibration due to the covalent Mo-O bond and high compressibility due to the van der Waals gaps in MoO3.
Angle-resolved photoemission spectra of 1T-TaS2 have been measured to reveal the electronic structure of valence bands in the incommensurate (IC) and commensurate (C) charge density wave (CDW) phases. In the ICCDW phase at 370 K, Ta 5d states are observed as a single branch along the Γ–M direction over the binding energy range EB=0–1.2 eV. In the CCDW phase at 17 K, on the other hand, Ta 5d states are divided into three sub-band manifolds centered at EB=0.2, 0.5 and 1.0 eV. The former sub-band at EB=0.2 eV is less dispersive, while latter two show clear dispersive nature with the bandwidths of ∼0.5 and 0.3 eV. Furthermore, the spectral intensities are concentrated around the original Ta 5d band.
Two-dimensional oxide η-Mo4O11 exhibits almost isotropic thermal expansion in the normal phase, while anisotropic negative thermal expansion (NTE) in the charge density wave (CDW) phase below Tc1=105K, where a remarkable anomalous expansion occurs along the a-axis. The incommensurate nesting vector along the b-axis is independent of temperature, q1=(0.000(1)0.2335(1)0.0000(5)). We propose that the CDW-induced NTE occurs as a result of the structural relaxation of the MoO6 octahedra layers along the stacking direction with the aid of the open spaces around the MoO4 tetrahedra.
The accurate electron charge density distributions of MoO3 have been investigated by analyzing the synchrotron radiation X-ray powder diffraction data using the MEM (Maximum Entropy Method)/Rietveld method. The covalent bonding is observed only on the five Mo-O bonds, while six O atoms surround the Mo atom. Turning attention to the covalent bonding network, the crystal structure of MoO3 is regarded as a two-dimensional mono-layer structure consisting of the distorted Mo-O-5 pyramids. The ionic state of the Mo atom is estimated to be +3.0(1) and those of three O atoms are -1.6(1), -0.7(1), -0.8(1) by counting number of electrons around these atoms, which coincide with the results of theoretical studies.
We have measured electrical resistivity, Hall coefficient, thermoelectric power, and magnetization for charge-density wave (CDW) material 1T-TaS2 single crystals grown by varying the excess sulfur content xes. We have revealed that a small mid-gap state is formed inside the Mott gap and that anomalous low temperature transport is not governed by the Mott gap state itself but by the mid-gap state. The electric properties of the mid-gap state are modified by increasing xes (or hole doping), and we have found the insulator-metal transition occurs by hole doping below 60 K.
Long-length scale interaction model proposed by Littlewood has been extended to study dipole and screening effects on sliding motion of one-dimensional charge-density-wave (CDW). The system in the extended model consists of a site-by-site main-system fixed at pinning centers and the continuous sub-systems in each pinning site. For strong pinning case, we have demonstrated these effects on the CDW phase, CDW current, and narrow band noise spectrum. The dipole effect suppresses the pinning potential and gives rise to a change from deformable "plastic flow" like to rigid "moving soil" like CDW condenstates, while the screening effect acts to recover the former.