No Arabic abstract
We study kinks in the electronic dispersion of a generic strongly correlated system by dynamic mean-field theory (DMFT). The focus is on doped systems away from particle-hole symmetry where valence fluctuations matter potentially. Three different algorithms are compared to asses their strengths and weaknesses, as well as to clearly distinguish physical features from algorithmic artifacts. Our findings extend a view previously established for half-filled systems where kinks reflect the coupling of the fermionic quasiparticles to emergent collective modes, which are identified here as spin fluctuations. Kinks are observed when strong spin fluctuations are present and, additionally, a separation of energy scales for spin and charge excitations exists. Both criteria are met by strongly correlated systems close to a Mott-insulator transition. The energies of the kinks and their doping dependence fit well to the kinks in the cuprates, which is surprising in view of the spatial correlations neglected by DMFT.
We consider the optical conductivity $sigma_1(omega)$ in the metallic phase of the one-dimensional Hubbard model. Our results focus on the vicinity of half filling and the frequency regime around the optical gap in the Mott insulating phase. By means of a density-matrix renormalization group implementation of the correction-vector approach, $sigma_1(omega)$ is computed for a range of interaction strengths and dopings. We identify an energy scale $E_{rm opt}$ above which the optical conductivity shows a rapid increase. We then use a mobile impurity model in combination with exact results to determine the behavior of $sigma_1(omega)$ for frequencies just above $E_{rm opt}$ which is in agreement with our numerical data. As a main result, we find that this onset behavior is not described by a power law.
We propose using ultracold fermionic atoms trapped in a periodically shaken optical lattice as a quantum simulator of the t-J Hamiltonian, which describes the dynamics in doped antiferromagnets and is thought to be relevant to the problem of high-temperature superconductivity in the cuprates. We show analytically that the effective Hamiltonian describing this system for off-resonant driving is the t-J model with additional pair hopping terms, whose parameters can all be controlled by the drive. We then demonstrate numerically using tensor network methods for a 1D lattice that a slow modification of the driving strength allows near-adiabatic transfer of the system from the ground state of the underlying Hubbard model to the ground state of the effective t-J Hamiltonian. Finally, we report exact diagonalization calculations illustrating the control achievable on the dynamics of spin-singlet pairs in 2D lattices utilising this technique with current cold-atom quantum-simulation technology. These results open new routes to explore the interplay between density and spin in strongly-correlated fermionic systems through their out-of-equilibrium dynamics.
We have investigated the half-filling two-orbital Hubbard model on a triangular lattice by means of the dynamical mean-field theory (DMFT). The densities of states and optical conductivity clearly show the occurence of metal-insulating transition (MIT) at U$_{c}$, U$_{c}$=18.2, 16.8, 6.12 and 5.85 for J=0, 0.01U, U/4 and U/3, respectively. The distinct continuities of double occupation of electrons, local square moments and local susceptibility of the charge, the spin and the orbital at J > 0 suggest that the MIT is the first-order; however at J=0, the MIT is the second-order in the half-filling two-orbital Hubbard model on triangular lattices. We attribute the first-order nature of the MIT to the low symmetry of the systems with finite Hunds coupling J.
Two very different methods -- exact diagonalization on finite chains and a variational method -- are used to study the possibility of a metal-insulator transition in the symmetric half-filled periodic Anderson-Hubbard model. With this aim we calculate the density of doubly occupied $d$ sites as a function of various parameters. In the absence of on-site Coulomb interaction ($U_f$) between $f$ electrons, the two methods yield similar results. The double occupancy of $d$ levels remains always finite just as in the one-dimensional Hubbard model. Exact diagonalization on finite chains gives the same result for finite $U_f$, while the Gutzwiller method leads to a Brinkman-Rice transition at a critical value ($U_d^c$), which depends on $U_f$ and $V$.
The properties of condensed matter are determined by single-particle and collective excitations and their interactions. These quantum-mechanical excitations are characterized by an energy E and a momentum hbar k which are related through their dispersion E_k. The coupling of two excitations may lead to abrupt changes (kinks) in the slope of the dispersion. Such kinks thus carry important information about interactions in a many-body system. For example, kinks detected at 40-70 meV below the Fermi level in the electronic dispersion of high-temperature superconductors are taken as evidence for phonon or spin-fluctuation based pairing mechanisms. Kinks in the electronic dispersion at binding energies ranging from 30 to 800 meV are also found in various other metals posing questions about their origins. Here we report a novel, purely electronic mechanism yielding kinks in the electron dispersions. It applies to strongly correlated metals whose spectral function shows well separated Hubbard subbands and central peak as, for example, in transition metal-oxides. The position of the kinks and the energy range of validity of Fermi-liquid (FL) theory is determined solely by the FL renormalization factor and the bare, uncorrelated band structure. Angle-resolved photoemission spectroscopy (ARPES) experiments at binding energies outside the FL regime can thus provide new, previously unexpected information about strongly correlated electronic systems.