Talk 1:
Two-site entanglement in the two-dimensional Hubbard model from parquet dynamical vertex approximation
Anna Kauch, TU Wien
The study of entanglement in strongly correlated electron systems typically requires knowledge of the reduced density matrix. We present an approach that can be used as a postprocessing step for calculating the two-site reduced density matrix and from it entanglement measures such as the mutual information and entanglement negativity. Input is only the one- and two-particle Green’s function, which is the output of numerous many-body methods. We then apply the parquet dynamical vertex approximation to study the two-site reduced density matrix at varying distance, in the Hubbard model at weak coupling. This allows us to investigate the spatial structure of entanglement in dependence of interaction strength, electron filling, and temperature. We compare results from different entanglement measures, and benchmark against quantum Monte Carlo.
Talk 2:
Finite-difference parquet method and the strong-coupling pseudogap
Fabian Kugler, Institute for Theoretical Physics, University of Cologne
We first present the finite-difference parquet method, a two-particle diagrammatic approach with nonperturbative input. It takes the fully irreducible two-particle vertex from a reference solution while requiring only its full vertex explicitly. Using dynamical mean-field theory (DMFT) as a reference, this yields a reformulation of the parquet dynamical vertex approximation circumventing ill-behaved two-particle irreducible vertices. We then use this method to investigate the pseudogap phase of the underdoped Hubard model. Our numerical results are consistent with diagrammatic Monte Carlo simulations and shed new light on the microscopic mechanism of the strong-coupling pseudogap: With dominant short-ranged antiferromagnetic spin fluctuations, we find an enhanced electron-paramagnon scattering amplitude crucial for the pseudogap opening. The form of this enhancement, reflected in the real part of the Hedin vertex, requires strong local correlations from DMFT as well as nonlocal correlations in multiple two-particle channels from solving the parquet equations.