Nature of the chiral phase transition of two flavour QCD from imaginary chemical potential with HISQ fermions

TL;DR

This study investigates the chiral phase transition in two-flavor QCD using lattice simulations with HISQ fermions and imaginary chemical potential, revealing finite volume effects and the need for larger lattices to clarify the transition's nature.

Contribution

First lattice QCD study employing HISQ fermions at imaginary chemical potential to analyze the chiral phase transition, highlighting finite volume effects and the necessity for larger lattices.

Findings

Binder cumulant at intersection deviates from Z(2) universality class

Largest lattice volume results align with previous studies

Finite cut-off effects are reduced but larger lattices are needed

Abstract

The nature of the thermal phase transition of two flavor QCD in the chiral limit has an important implication for the QCD phase diagram. We carry out lattice QCD simulations in an attempt to address this problem. Simulations are conducted with a Symanzik-improved gauge action and the HISQ fermion action. Within the imaginary chemical potential formulation, five different quark masses, $am=0.020,\, 0.018, \, 0.015, \, 0.013,\, 0.010$, and four different lattice volumes $N_s=8, \, 12,\, 16, \, 20$ with temporal extent $N_t=4$ are used to explore the scaling behavior. At each of the quark masses, the Binder cumulants of the chiral condensate on different lattice volumes approximately intersect at one point. We find that at the intersection point, the Binder cumulant $B_4(am,a\mu_c) $ is around $3$ which deviates from the $Z(2)$ universality class value 1.604. However, based on the expectations of $Z(2)$ criticality, the fitting result only with the data from the largest lattice volume $N_s=20$ agrees well with earlier result [ Phys. Rev., D90, 074030(2014) ]\cite{Bonati:2014kpa}. This fact implies that, although the finite cut-off effects could be reduced with HISQ fermions even on $N_t=4$ lattices, larger lattices with spatial extent $N_s>=20$ for such studies are needed to control finite volume effects.