Abstract. In this work we revise the standard iterative procedure to find the solution of the gap equation in the Nambu-Jona-Lasinio model within the most popular regularization schemes available in literature in the super-strong coupling regime. We observe that whereas for the hard cut-off regularization schemes, the procedure smoothly converges to the physically relevant solution, Pauli-Villars and Proper-Time regularization schemes become chaotic in the sense of discrete dynamical systems. We call for the need of an appropriate interpretation of the non-convergence of this procedure to the solution of the gap equation.
IntroductionThe Nambu-Jona-Lasinio (NJL) model [1, 2] was one of the earliest attempts to describe strong interactions among protons and neutrons. Spontaneous breaking of chiral symmetry is presented in the model through an analogy with the phenomenon of superconductivity in condensed matter physics. The model has a single coupling constant for two kinds of 4-Fermi interactions and is non-renormalizable, which demands for a regulator. Nevertheless, up to date, the model continues being used as an effective description of quarks dynamics either as originally proposed, where fermion fields are now regarded as quark fields, or with slightly modifications [3, 4, 5] to account for confinement. NJL model has been widely used to sketch the quantum chromodynamics (QCD) phase diagram (see, for instance, Ref.[6] for a review) and its magnetized extension [7] from a complementary point of view of lattice [8] and other quantum field theoretical approaches [9,10]. Several regularization schemes have been used in literature within the context of NJL model. Perhaps the simplest regulator one can think of is to establish a hard cut-off in the four-momentum integrals, or over the spatial momentum integrals, which thus allows a straightforward incorporation of temperature and density effects within the Matsubara formalism [11]. Pauli-Villars regularization has also been used in NJL studies due to the advantages of the scheme. Nevertheless, external magnetic field effects are best included through the Schwinger Proper-Time representation of the fermion propagator. This observation has been exploited to address the problem of magnetic catalysis [12] and inverse magnetic catalysis observed in lattice simulations [13,14] and confirmed by several approaches [7,8,15]. Under this environment, the effective coupling of the model can reach very large values, and therefore,