Effective Hamiltonians and Phase Diagrams for Tight-Binding Models
Nilanjana Datta (EPFL, Switzerland), Roberto Fernandez (Universidade, de Sao Paolo, Brazil), Juerg Froehlich (ETH, Switzerland)

TL;DR
This paper rigorously analyzes effective Hamiltonians and phase diagrams for various Hubbard model variants in the strong-coupling regime, extending previous results and providing algorithms for high-order computations.
Contribution
It introduces a mathematically controlled perturbation expansion and applies quantum Pirogov-Sinai theory to establish stable phase diagrams for the models, including small temperature and quantum perturbations.
Findings
Validated zero-temperature phase diagrams for the Falicov-Kimball model
Extended phase diagram results to small temperatures and quantum effects
Provided explicit expressions and algorithms for high-order effective interactions
Abstract
We present rigorous results for several variants of the Hubbard model in the strong-coupling regime. We establish a mathematically controlled perturbation expansion which shows how previously proposed effective interactions are, in fact, leading-order terms of well defined (volume-independent) unitarily equivalent interactions. In addition, in the very asymmetric (Falicov-Kimball) regime, we are able to apply recently developed phase-diagram technology (quantum Pirogov-Sinai theory) to conclude that the zero-temperature phase diagrams obtained for the leading classical part remain valid, except for thin excluded regions and small deformations, for the full-fledged quantum interaction at zero or small temperature. Moreover, the phase diagram is stable upon the addition of arbitrary, but sufficiently small, further quantum terms that do not break the ground-states symmetries. This…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
