Grand canonical Gutzwiller approximation for magnetic inhomogeneous systems

TL;DR

This paper develops an improved grand canonical Gutzwiller approximation for inhomogeneous magnetic systems, incorporating higher-order correlations and different constraints, leading to better agreement with variational Monte Carlo results and insights into spectral asymmetries.

Contribution

It introduces a refined Gutzwiller approximation with spin-dependent constraints, higher-order corrections, and spectral analysis for inhomogeneous magnetic systems, advancing beyond conventional methods.

Findings

Improved GA agrees better with variational Monte Carlo.

Higher powers of intersite contractions cause spectral asymmetry.

Asymmetry varies with pairing symmetry and is generally weak near the Fermi level.

Abstract

The Gutzwiller approximation (GA) for Gutzwiller-projected grand canonical wave functions with fugacity factors is investigated in detail. Our systems in general contain inhomogeneity and local magnetic moments. In deriving renormalization formulae, we also derive or estimate terms of higher powers of intersite contractions neglected in the conventional GA. We examine several different constraints, i.e., local/global spin-dependent/independent particle-number conservation. Out of the four, the local spin-dependent constraint seems the most promising at present. An improved GA derived from it agrees with the variational Monte Carlo method better than the conventional GA does. The corrections to the conventional GA can be interpreted as two-site correlation including the phase difference of configurations. Furthermore, projected quasi-particle excited states are orthogonal to each other within the GA. Using these states, spectral weights are calculated. We show that asymmetry between electron addition and removal spectra can appear by taking into account the higher powers of the intersite contractions in the case of the d-wave superconductors and the Fermi sea; the addition is smaller than the removal. However, the asymmetry is quite weak especially near the Fermi level. In contrast, projected s-wave superconductors can have the opposite asymmetry (addition larger than removal) especially near the Fermi level. In addition, formulae from the other three constraints are also derived, which may be useful depending on purposes.