d−p model and spin-orbital order in vanadium perovskites

Abstract

Using the multi-band d−p model and unrestricted Hartree-Fock approximation we investigate the electronic structure and spin-orbital order in three-dimensional VO3 lattice. The main aim of this investigation is testing if simple d − p model, with partly filled 3d orbitals (at vanadium ions) and 2p orbitals (at oxygen ions), is capable of reproducing correctly nontrivial coexisting spin-orbital order observed in the vanadium perovskites. We point out that the multi-band d − p model has to include partly filled eg orbitals at vanadium ions. The results suggest weak self-doping as an important correction beyond the ionic model and reproduce the possible ground states with broken spin-orbital symmetry on vanadium ions: either C-type alternating orbital order accompanied by G-type antiferromagnetic spin order, or G-type alternating orbital order accompanied by C-type antiferromagnetic spin order. Both states are experimentally observed and compete with each other in YVO3 while only the latter was observed in LaVO3. Orbital order is induced and stabilized by particular patterns of oxygen distortions arising from the Jahn-Teller effect. In contrast to timeconsuming ab-initio calculations, the computations using d − p model are very quick and should be regarded as very useful in solid state physics, provided the parameters are selected carefully. z

Extended Falicov–Kimball model: Hartree–Fock vs DMFT approach

“…In the first case, Eqs. (12) and (13) of Ref. [34] are in a coincidence with our results represented by Eqs.…”

“…For example, these systems exhibit quite complex phase diagrams with, e.g., metal-insulator transitions and competition between different ordering such as spin-, charge-, orbital-order as well as superconductivity, e.g., Refs. [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17]. Knowing and understanding their properties is important not only in the context of condensed matter physics, but also for physics of ultra-cold quantum gases, where intensive experimental development occurs in the recent years (for a review see, e.g., Refs.…”

Spin-orbital order in LaMnO3 : d−p model study

“…The splitting value should be ∼ 0.1 eV what is an educated guess (compare Ref. [35]) or even smaller [16]. What concerns the occupied e g levelsthey are also split.…”

“…However, instead of introducing this term with a quite complicated form [37], we shall model the JT interaction by a simple Ansatz which can be inserted into H diag as local potentials acting on e g electrons. The cluster geometry and precise forms of different terms are standard; these terms were introduced in the previous realizations of the d − p model devoted to transition metal oxides [32][33][34][35].…”

“…Note that model studies done on LaMnO 3 before usually neglected these Hamiltonian components. In addition, we would like to account for the possibility of non-trivial self-doping x = 0 (just as found before in ruthenium, iridium, titanium, and vanadium oxides [32][33][34][35]).…”