Evidence for the constancy of U in the Mott transition of V2O3

Physical review. B, Condensed matter (Impact Factor: 3.66). 08/2009; DOI: 10.1103/PhysRevB.84.075117
Source: arXiv

ABSTRACT We have performed high-resolution hard X-ray photoemission spectroscopy for the metal-insulator transition (MIT) system (V(1-x)Cr(x))2O3 in the paramagnetic metal, paramagnetic insulator and antiferromagentic insulator phases. The quality of the spectra enables us to conclude that the on-site Coulomb energy U does not change through the MIT, which eliminate all but one theoretical MIT scenario in this paradigm material. Comment: 4 pages, 3 figures

  • [Show abstract] [Hide abstract]
    ABSTRACT: The electronic structure and the metal–insulator transition (MIT) of V2O3 are investigated in the framework of density functional theory and GGA+U. It is found that, both the insulating and metallic phases can be realized in rhombohedral structure by varying the on-site Coulomb interaction, and the MIT in V2O3 can take place without any structure phase transition. Our calculated energy gap (0.63eV) agrees with experimental result very well. The metallic phase exhibits high spin (S=1) character, but it becomes S=1/2 in insulating phase. According to our analysis, the Mott–Hubbard and the charge-transfer induce the MIT together, and it supports the mechanism postulated by Tanaka (2002) [11].
    Journal of Physics and Chemistry of Solids 09/2011; 72(9):1085-1089. · 1.59 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Highly bulk-sensitive hard x-ray photoemission has been performed for the Magneacuteli phase V(5)O(9) and V(6)O(11), which are composed of V(2)O(3) and VO(2) from chemical point of view. The valence-band spectra near the Fermi level are rather similar between V(5)O(9) and V(6)O(11) in both metallic and insulating phases. Although the core-level spectra in the metallic phase are similar between V(5)O(9) and V(6)O(11), the core-level spectra in the insulating phase are noticeably different. Our results are understood by considering the contribution of the V-V dimerization effects, which are strong in V(6)O(11) and negligible in V(5)O(9). Comparison of the core-level spectra of V(5)O(9) and V(6)O(11) with those of VO(2) has revealed that the mechanism of the metal-insulator transitions for the former two compounds is different from that for VO(2).
    Physical Review B 01/2010; 81. · 3.66 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: The tunability of bonding character in transition-metal compounds controls phase transitions and their fascinating properties such as high-temperature superconductivity, colossal magnetoresistance, spin-charge ordering, etc. However, separating out and quantifying the roles of covalency and metallicity derived from the same set of transition-metal d and ligand p electrons remains a fundamental challenge. In this study, we use bulk-sensitive photoelectron spectroscopy and configuration-interaction calculations for quantifying the covalency and metallicity in correlated compounds. The method is applied to study the first-order temperature- (T-) dependent metal-insulator transitions (MITs) in the cubic pyrochlore ruthenates Tl2Ru2O7 and Hg2Ru2O7. Core-level spectroscopy shows drastic T-dependent modifications which are well explained by including ligand-screening and metallic-screening channels. The core-level metallic-origin features get quenched upon gap formation in valence band spectra, while ionic and covalent components remain intact across the MIT. The results establish temperature-driven Mott-Hubbard MITs in three-dimensional ruthenates and reveal three energy scales: (a) 4d electronic changes occur on the largest (∼eV) energy scale, (b) the band-gap energies/charge gaps (Eg∼160–200 meV) are intermediate, and (c) the lowest-energy scale corresponds to the transition temperature TMIT (∼10 meV), which is also the spin gap energy of Tl2Ru2O7 and the magnetic-ordering temperature of Hg2Ru2O7. The method is general for doping- and T-induced transitions and is valid for V2O3, CrN, La1−xSrxMnO3, La2−xSrxCuO4, etc. The obtained transition-metal–ligand (d–p) bonding energies (V∼45–90 kcal/mol) are consistent with thermochemical data, and with energies of typical heteronuclear covalent bonds such as C-H, C-O, C-N, etc. In contrast, the metallic-screening energies of correlated compounds form a weaker class (V*∼10–40 kcal/mol) but are still stronger than van der Waals and hydrogen bonding. The results identify and quantify the roles of covalency and metallicity in 3d and 4d correlated compounds undergoing metal-insulator transitions.
    Physical review. B, Condensed matter 01/2013; 87(4). · 3.66 Impact Factor

Full-text (2 Sources)

Available from
Jun 3, 2014