Renormalized vs unrenormalized perturbation-theoretical approaches to the Mott transition

TL;DR

This paper investigates why unrenormalized perturbation theory often outperforms self-consistent methods in modeling the Mott transition using an exactly solvable toy model.

Contribution

It provides insights into the effectiveness of unrenormalized perturbation theory over self-consistent approaches for the Mott transition.

Findings

Unrenormalized perturbation theory better captures the transition

Self-consistent approaches may introduce inaccuracies

Toy model demonstrates key differences in methods

Abstract

I employ an exactly soluble toy model to investigate why unrenormalized perturbation theory works better than fully self-consistent approaches in describing the correlation-driven metal-insulator transition.