Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe

TT: Fachverband Tiefe Temperaturen

TT 13: Correlated Electrons: Method Development I

TT 13.1: Vortrag

Montag, 9. März 2026, 15:00–15:15, HSZ/0101

Real-Frequency Dynamical Mean-Field Theory Without Broadening Inside the Self-Consistency Cycle — •Aleksandrs Zacinskis1, Frank T. Ebel2, Sina Shokri1, Lukas Hellman1, Fabian B. Kugler3, Karsten Held2, and Maurits W. Haverkort11Institute for Theoretical Physics, Heidelberg University — 2Institute of Solid State Physics, TU Wien — 3Institute for Theoretical Physics, University of Cologne

Dynamical mean-field theory has become one of the central frameworks for studying strongly correlated electron systems. Real-frequency implementations offer a major advantage by providing direct access to spectral properties without the need for analytic continuation. Most current real-frequency implementations work slightly off the real axis and evaluate G(ω+iΓ) instead of G(ω), which introduces an artificial broadening and can lead to systematic errors. We present a strictly real-frequency method with zero broadening inside the self-consistency cycle, addressing both the Dyson equation and the "improved estimator" self-energy [PRB 105, 245132]. Our formulation allows the self-energy to be extracted directly from the continued-fraction representation of the impurity Green’s function, as obtained from Krylov bases in Lanczos methods, Wilson chains in NRG, or Kernel Polynomial methods. This method completely removes the need for artificial broadening and is applicable to all real-frequency impurity solvers. We compare the zero-broadening self-energy to its finite-broadening counterpart obtained with Exact Diagonalization solver, demonstrating the impact of broadening on accuracy and various quantities.

Keywords: Dynamical mean-field theory; Anderson impurity model; Self-energy; Exact diagonalization; Broadening

100% | Bildschirmansicht | English Version | Kontakt/Impressum/Datenschutz
DPG-Physik > DPG-Verhandlungen > 2026 > Dresden