Flavor symmetry breaking in mixed-action QCD

TL;DR

This paper analyzes the phase structure of mixed-action QCD with Wilson sea quarks and chiral valence quarks, proving that certain mixed condensate phases are impossible, which constrains low-energy constants.

Contribution

It demonstrates, using properties of the chiral Dirac operator, that a theoretically possible mixed condensate phase cannot occur in mixed-action QCD.

Findings

No mixed condensate phase occurs, ensuring pure-sea correlation functions are unaffected by valence quark masses.

Bounds on low-energy constants are derived from the non-occurrence of the mixed phase.

The phase structure is clarified, supporting the consistency of mixed-action lattice simulations.

Abstract

We study the phase structure of mixed-action QCD with two Wilson sea quarks and two chiral valence quarks, starting from the chiral lagrangian. A priori, the effective theory allows for a rich phase structure, including a phase with a condensate made of sea and valence quarks. Because this would lead to mass eigenstates that are admixtures of sea and valence fields, pure-sea correlation functions would depend on valence quark masses, in contradiction with the actual setup of mixed-action simulations. Using properties of the chiral Dirac operator, we prove that such a phase does not occur, and that this leads to bounds on low-energy constants.