Magnetic phase diagram of the anisotropic multi‐band Hubbard model

Abstract

Using quantum Monte Carlo (QMC) simulations we determine the magnetic phase diagram of the anisotropic two-band Hubbard model within the dynamical mean-field theory (DMFT) in the important intermediate-coupling regime. We compare the QMC predictions with exact results from second-order weakand strong-coupling perturbation theory. We find that the orbital-selective Mott transition (OSMT), which occurs in the fully frustrated case, is completely hidden in the antiferromagnetic (AF) ground state of the model. On the basis of our results, we discuss possible mechanisms of frustration. We also demonstrate the close relationship of the physics of the two-band Hubbard model in the orbital-selective Mott (OSM) phase to the Falicov -Kimball model.

Orbital-selective Mott transitions in a doped two-band Hubbard model

“…(1) as the J z -model. 28,29,30 A general discussion of the Hund exchange would have to include spin-flip and pair hopping terms, However, such terms are not essential for OSMTs (Ref. 28) and change their character only in the SU (2) symmetric limit J z = J ⊥ ≡ J (i.e., in the Heisenberg limit).…”

“…( 1) as the J z -model. 28,29,30 A general discussion of the Hund exchange would have to include spin-flip and pair hopping terms,…”

“…In this paper, we extend high-precision 35 quantum Monte Carlo (QMC) calculations 28,30,36 within DMFT for two-band model (1) to general doping levels. Of central interest is the phase diagram, which is derived from orbital-dependent fillings, double occupancies, and quasiparticle weights.…”

Orbital-selective Mott transitions in a doped two-band Hubbard model with crystal field splitting

“…For these new particles, we then define annihilation operators of double occupancies d iσ =c i2σci1σ (with σ =↑, ↓) and d i0 = These results are interesting, because they show that T c is dominated in both models by the largest hopping amplitude, which is that of the broad band (t * 2 ≡ t 2 √ Z). We conclude that the broad band primarily determines the energy scale at strong coupling; from previous work 35 we know that the energy scales of antiferromagnetism at weak coupling are primarily determined by the narrow band (t *…”

“…These results are interesting, because they show that T c is dominated in both models by the largest hopping amplitude, which is that of the broad band (t * 2 ≡ t 2 √ Z). We conclude that the broad band primarily determines the energy scale at strong coupling; from previous work 35 we know that the energy scales of antiferromagnetism at weak coupling are primarily determined by the narrow band (t * 1 ≡ t 1 √ Z). Note that, since the effective strongcoupling Hamiltonians are ∆-independent (within their range of validity), the same holds for the critical temperatures.…”