Electronic Structure of the Heavy Fermion MetalLiV2O4

Abstract

The electronic structure of the first reported heavy fermion compound without f electrons, LiV 2 O 4 , was studied by an ab initio calculation method. As a result of the trigonal splitting and d-d Coulomb interaction, one electron of the d 1.5 configuration of the V ion is localized and the rest partially fills a relatively broad conduction band. The effective Anderson impurity model was solved by the noncrossing-approximation method, leading to an estimation for the single-site Kondo energy scale T K . Then, we show how the so-called exhaustion phenomenon of Nozières for the Kondo lattice leads to a remarkable decrease of the heavy-fermion (or coherence) energy scale T coh ϵ T 2 K ͞D (D is the typical bandwidth), comparable to the experimental result.

Orbital state and magnetic properties ofLiV2O4

“…The orbital occupation, which confirms the earlier conjecture based on the LDA+U approach, 33 can be readily understood from the previous analysis of the LDA band structure. There we found that the center of mass of the a 1g orbital is 0.26 eV lower than the corresponding value for the e π g orbitals.…”

“…The value and the sign of the trigonal splitting will determine the orbital in which the V 3d electrons should be localized when a strong Coulomb interaction, which is larger than the bandwidth, is taken into account. 33 We found that this trigonal splitting value is very sensitive to the accuracy of the band structure calculations. In order to increase the accuracy, the overlap between atomic spheres in our LMTO calculation was set to zero and more empty spheres were introduced.…”

“…The main goal of this study is to further explore the idea of a separation of the d-electrons in LiV 2 O 4 into two subsets, localized and itinerant as proposed in Ref. 33. The rest of the paper is divided into two parts: In the first we discuss standard DFT/LDA results (section II) and then present our LDA+DMFT(QMC) results in the second (section III).…”

“…A novel attempt to provide a microscopic model based on material properties of LiV 2 O 4 , which does not evoke the idea of frustration was suggested by Anisimov et al. 33 Their basic idea is a separation of the electrons on partially filled t 2g orbitals into localized ones, forming local moments, and delocalized ones, producing a partially filled metallic band. The hybridization between those two subsets of electrons, like in f -electrons materials, can give rise to heavy-fermion effects.…”

“…17,26 From a theoretical point of view, LiV 2 O 4 has been intensively studied by standard band structure calculations by various groups using different implementations of density functional theory 31 within the local density approximation (DFT/LDA). 32 Anisimov et al 33 investigated the possibility of localization of the d-states in LiV 2 O 4 within the linearized muffin-tin orbitals (LMTO) basis 34 supplemented by the LDA+U method. 35 The electronic structure was furthermore studied by Eyert et al 36 with the scalar-relativistic augmented spherical wave (ASW) basis, 37 while Matsuno et al 38 used a full-potential, scalar-relativistic implementation 39 of the linear augmented plane wave (LAPW) approach 40 for the band structure calculation and furthermore applied a simple tight-binding (TB) model.…”