Effects of isospin imbalance on the chiral phase transition within a Chiral Dual Partner Model

Abstract

The chiral symmetry is a property of the fundamental theory of the strong interaction that is relevant for the hadronic physics. It is expected that at sufficiently high temperature and matter density, this symmetry becomes manifest. Within the Chiral Dual Partner Model the chiral transformation is implemented in such a way that a fermion mass term is allowed if the parity partner of each baryon is included in the framework. This model is used here to study possible manifestations of the chiral symmetry in dense nuclear matter, assuming constant isospin fraction. It is found that the onset of the odd parity baryons is associated with two branches of thermodynamical instabilities, one of them leads to a first order phase transition for temperatures below $T_{c}\simeq 11$ MeV. An analysis of these instabilities in the phase space is given, and neutron star matter is considered as a special case.

I INTRODUCTION

The earlier studies of the manifestations of the chiral symmetry in a nuclear environment precede the formulation of a general theory of the strong interaction [1, 2, 3, 4, 5]. After the consolidation of QCD as the fundamental theory, these studies were developed by using effective models, due to the intricacies of this formulation. The nontrivial structure of the vacuum, the running coupling constant and the confinement mechanism turn almost impossible a description of dense hadronic matter in terms of quarks and gluons. These effective models try to extend the abstract symmetries of QCD to the hadronic physics, restricting their applicability to a range of phenomena characterized as the low energy domain. Within such domain the relevant symmetry is precisely the chiral symmetry.
One can find among the chiral effective models those that propose a scheme of progressive approximations, but with a limited applicability for the range of matter densities. The purpose of other models is to capture the right dynamics by judiciously choosing the degrees of freedom and their interactions. In such case they usually rely on self-consistent approaches, as for example the mean field approach (MFA) or similar, which allows to extend the validity of the description to the densest systems accessible to observation today. A reduced number of examples are given by [6, 7, 8, 9, 10, 11, 12].
The standard prescription for the chiral transformation require massless fermion fields, therefore the particles with a finite mass must obtain it dynamically. The coupling with a scalar meson field having a nonzero vacuum expectation value is the usual mechanism for this purpose. However, it has been pointed out [8, 9, 10] that the implementation of the chiral symmetry through an alternative prescription, known as mirror, allows the existence of a mass term for the fermion fields. This possibility was used to introduce a mass parameter $m_{0}$ which could be interpreted as a contribution from the quark structure that can not be accounted for solely by the hadronic interaction. It is expected that at high density/temperature the dynamical mass decrease to its minimum value $m_{0}$, this is the signal of the chiral restoration in a hadronic medium. The mirror assignment for the chiral symmetry is accomplished by introducing a complementary parity state for each baryonic degree of freedom, for this reason this framework is usually known as the Chiral Dual Partner Model (CDPM).
The CDPM has been extended from its early formulation to include a variety of issues, as for instance the phase diagram in an SU(2) treatment [13], axial vector mesons [14], neutron stars [15], hyperons [16, 17, 18, 19, 20, 21, 22, 23], the delta isobar [24, 25, 26], fluctuations of the conserved charges [27, 28, 29, 30], isospin imbalanced nuclear matter [31, 32], corrections by the method of functional renormalization group [33, 34], and Dirac sea effects [35]. Other approaches proposed an hybrid composition of the CDPM and quark interactions [36, 37, 38, 39].

The models aimed to describe the dense hadronic medium are subject to a variety of phenomenological constraints. In addition to the vacuum low lying mass spectrum, they should also give an acceptable explanation for the gross features of the atomic nuclei and to adjust the properties of the homogeneous nuclear matter at the saturation baryonic density $n_{0}=0.15$ fm^-3, such as binding energy, compressibility, symmetry energy and its slope parameter. The effective nucleon mass at the saturation density $M^{*}$ plays a special role, it has not been measured precisely but it is expected to vary within a restricted domain. The scalar interaction produces a decrease of this effective mass with the density, which in turn, affects the spin-orbit coupling for the atomic nuclei. It has been found that the larger $M^{*}$ the weaker spin-orbit term [40], therefore the reproduction of the energy levels for the closed shell nuclei of medium mass imposes a restriction on $M^{*}$. Furthermore, the increasing variety and precision of the astronomical observational data provides useful information, such as the mass-radius relation for neutron stars and the tidal deformability of a binary system, which involves the high density equation of state and must be properly taken into account.

The isospin imbalance is a common feature of nuclear dense systems which manifests, for instance, in the predominance of the neutron over the proton fractions. In models of meson exchange the nuclear isospin density is the source for the isovector mesons $\rho(780)$ and $a_{0}(980)$, which modify the energy spectrum and the effective mass of the fermions, and consequently their chemical potentials. In the CDPM the chiral phase transition is driven by the onset of the parity partners of the proton and neutron, which is determined by their chemical potentials. Therefore, in a system with conserved isospin configuration it is expected that the coupling to the isovector mesons must have sizeable effects on the chiral transition. There is only a reduced number of publications that considered the inclusion of the $a_{0}(980)$ meson within the CDPM [38, 39], but they impose a transition to quark matter at a relatively low baryonic density $n_{B}/n_{0}=2$. Therefore it is possible that such effects were not detected.
Another interesting reason to study strong interacting systems with isospin imbalance is that this would be the bridge between the hadronic phenomenology and the results obtained by QCD lattice simulations [41]. These simulations have technical difficulties to treat finite baryonic density, although they are viable in the presence of a finite isospin chemical potential $\mu_{I}$ [42]. Under such conditions the ground state obtained for low $\mu_{I}$ contains a pion condensed phase, and this condensate modifies the dynamical properties of the fermions [43].

This work is devoted to the study of possible manifestations of the chiral symmetry in a dense hadronic environment with isospin imbalance. The SU(2) CDPM with the addition of a scalar isovector meson is used for this purpose. The remaining of this work is organized as follows, in the next section a brief description of the model and the procedure used for fix its parameters is given. The main results are presented and discussed in Sec. III, they correspond to homogeneous matter with a constant isospin fraction and to neutron star matter as a special case. The conclusions are drawn in Sec. IV.

II THE MODEL

In this work an extended version of the CDPM suited to study isospin asymmetric matter is used. Therefore, as in previous investigations [9, 36, 31, 21, 15, 16, 32, 22, 37, 44, 45, 39], I include the vector isovector $\rho(770)$ meson. Furthermore, an explicit coupling between the nucleons and the scalar isovector meson $a_{0}(980)$ is considered. Its effects have been usually disregarded, although a few investigations [38, 39] have studied its role in the star matter at zero temperature. Thus, with exception of this addition, the usual procedure for the CDPM is applied here. For the sake of completeness it is summarized in the following. The Lagrangian density for the fermions is written in terms of the isospinor fields $\Psi_{1}$ and its chiral partner $\Psi_{2}$, which have opposite parity

$${\cal L}_{N}=\bar{\Psi}_{1}i\not\!\!D\Psi_{1}-g_{1}\bar{\Psi}_{1}\Sigma\Psi_{1}+\bar{\Psi}_{2}i\not\!\!D\Psi_{2}-g_{2}\bar{\Psi}_{2}\Sigma^{\dagger}\Psi_{2}-m_{0}\left(\bar{\Psi}_{1}\gamma_{5}\Psi_{1}-\bar{\Psi}_{2}\gamma_{5}\Psi_{2}\right)$$ (1)

where $\Sigma=\sigma+\bm{\tau}\cdot\bm{\zeta}+i\gamma_{5}\bm{\tau}\cdot\bm{\pi}$ collects the scalar and pseudosclar vertices, and $D_{\mu}=\partial_{\mu}+i\,g_{w}\omega_{\mu}+i\,g_{r}\bm{\tau}\cdot\bm{\rho}_{\mu}$ includes the vector meson couplings. Despite the mass term, this construction is chiral invariant under the mirror prescription [8, 10],

$$\left[Q^{a}_{V},\Psi_{k}\right]=-t^{a}\,\Psi_{k},\;\left[Q^{a}_{A},\Psi_{k}\right]=-t^{a}\,\eta_{k}\,\Psi_{k}$$

for $Q_{V},\,Q_{A}$ the vector and axial charges, $t_{a}=\tau_{a}/2$, and $\eta_{k}=(-1)^{k+1}$.
The meson sector is decomposed as the sum of the kinetic and interaction terms

$${\cal L}_{\text{kin}}=\frac{1}{2}\partial_{\mu}\sigma\,\partial^{\mu}\sigma+\frac{1}{2}\partial_{\mu}\bm{\zeta}\cdot\partial^{\mu}\bm{\zeta}-\frac{1}{4}W_{\mu\nu}W^{\mu\nu}-\frac{1}{4}R_{\mu\nu}R^{\mu\nu}+\frac{1}{2}m_{w}^{2}\,\omega_{\mu}\omega^{\mu}+\frac{1}{2}m_{r}^{2}\,\bm{\rho}_{\mu}\cdot\bm{\rho}^{\mu}$$ (2)

$${\cal L}_{\text{int}} = \frac{C_{0}}{4}\,\text{Tr}\left(\Sigma^{\dagger}\Sigma\right)-\frac{C_{1}}{16}\,\left[\text{Tr}\left(\Sigma^{\dagger}\Sigma\right)\right]^{2}-\frac{C_{2}}{8}\,\text{Tr}\left(\Sigma^{\dagger}\Sigma\right)^{2}+h\left[\text{Tr}\left(\Sigma+\Sigma^{\dagger}\right)/2-f_{\pi}\right]$$ (3)

$$-\sum_{3}^{4}\frac{\alpha_{n}}{2^{n}n!}\left[\text{Tr}\left(\Sigma^{\dagger}\Sigma\right)/2-f_{\pi}^{2}\right]^{n}-\frac{\alpha_{0}}{8}\left(\omega^{4}+\rho^{4}+6\omega^{2}\rho^{2}\right)+C_{3}$$

The scalar mesons interaction, in the first line, resembles that of the Linear sigma model (LSM) [4, 46], but it is complemented here by polynomial terms of order six and eight, shown in the second line. Such completion has been a common practice for the use of nuclear chiral models [47, 48, 49, 50, 51], which has been extended to the CDPM [22]. In the vector meson sector, a self-interaction is included in Eq. 3, using the convention $\omega^{2n}=\left(\omega_{\mu}\omega^{\mu}\right)^{n}$ and $\rho^{2n}=\left(\bm{\rho}_{\mu}\cdot\bm{\rho}^{\mu}\right)^{n}$. Nuclear models have introduced this class of vertices to fulfill phenomenological requirements, particularly in relation to the isospin response. It has also been previously associated with the CDPM, as for instance in [22].
In the LSM the coefficients of the scalar potential are usually chosen to adjust the masses of the lightest mesons. This can be done by expanding the scalar meson fields around their vacuum expectation values, $\sigma=s_{0}+S$, $\zeta_{a}=z_{0}+Z_{a}$, $\pi_{a}=\pi_{0}+P_{a}$, with $s_{0}=f_{\pi}$, $z_{0}=\pi_{0}=0$ and defining

$$m_{F}^{2}=\frac{\delta^{2}{\cal L}}{\delta F\,\delta F^{\dagger}}(S=0,\bm{Z}=0,\bm{P}=0),\;\;F=S,\,Z_{a},\,P_{a}$$

Thus, the meson masses are known even if the calculations were performed at the tree level.
The same decomposition can be applied at finite density and temperature $\sigma=s+\delta\sigma$, $\zeta_{a}=z\delta_{3a}+\delta\zeta_{a}$, $\pi_{a}=\delta\pi_{a}$, $\omega_{\mu}=w\,g_{0\mu}+\delta\omega_{\mu}$, $\rho^{a}_{\mu}=r\,\delta_{a3}g_{0\mu}+\delta\rho^{a}_{\mu}$, replacing vacuum by in medium expectation values. Assuming homogeneous and isotropic matter these expectations values behave as uniform background fields. In this work the mean field approach (MFA) is used, which is obtained by neglecting the meson fluctuations. Within this approach the baryon mass matrix in the isotopic space of the chiral partners is given by

$${\cal M}=\left(\begin{array}[]{cc}g_{1}(s+I_{a}z)&m_{0}\gamma_{5}\\ -m_{0}\gamma_{5}&g_{2}(s+I_{a}z)\end{array}\right),$$

where $I_{a}=\tau^{3}_{aa}$. Furthermore the meson fields matrix reduces to $\Sigma=\text{diag}\left(s+z,s-z\right)$.
To obtain the mass eigenstates, which will be identified as the nucleons and their chiral partners, a linear transformation is proposed

$$\left(\begin{array}[]{c}\Psi_{1}^{a}\\ \Psi_{2}^{a}\end{array}\right)=\left(\begin{array}[]{cc}\cos\theta&-\gamma_{5}\sin\theta\\ \gamma_{5}\sin\theta&\cos\theta\end{array}\right)\,\left(\begin{array}[]{c}N_{1}^{a}\\ N_{2}^{a}\end{array}\right)$$

By choosing $\tan\theta=m_{0}/G_{1}(s+I_{a}z)$, the eigenvalues of the mass matrix

$$M_{i}=G_{2}(s+I_{i}z)\,J_{i}+\sqrt{m_{0}^{2}+G_{1}^{2}(s+I_{a}z)^{2}},$$ (14)

are obtained, where $G_{1}=(g_{1}+g_{2})/2$, $G_{2}=(g_{1}-g_{2})/2$, $I_{i}=(-1)^{i+1}$, and $J_{i}=1$ for $i=1,2$, $J_{i}=-1$ for $i=3,4$. The indexes are assigned by the convention $i=1,\,2$ for the proton ($p$) and neutron ($n$) respectively, while the odd parity partners $p^{*}$ and $n^{*}$ are identified by $i=3,\,4$ respectively.

Within the approach just described, the grand potential per unit volume of the system can be written as

$$\Omega(s,z,w,r,T,\mu_{k})=\sum_{i=1,4}\Omega_{i}(T,M_{i},\mu_{i})-U(s,z)-V(w,r),$$

the sum extends over isotopic and parity components, and

$$\Omega_{i}=-\frac{2}{\beta}\int\frac{d^{3}p}{(2\pi)^{3}}\left[\ln\left(1+e^{-\beta\alpha_{i}}\right)+\ln\left(1+e^{-\beta\bar{\alpha}_{i}}\right)\right],$$ (15)

where $\beta=1/T$, $\alpha_{i}=\sqrt{p^{2}+M_{i}^{2}}+g_{w}w+g_{r}I_{i}r-\mu_{i}$, and $\bar{\alpha}_{i}=\sqrt{p^{2}+M_{i}^{2}}-g_{w}w-g_{r}I_{i}r+\mu_{i}$. Since the conservation of both the baryon number and the isospin composition are analyzed in this work, one must consider two chemical potentials $\mu_{B}$ and $\mu_{I}$, respectively. The particle chemical potentials are related to these, i.e. $\mu_{1}=\mu_{3}=\mu_{B}-\mu_{I}$, $\mu_{2}=\mu_{4}=\mu_{B}+\mu_{I}$. Furthermore, the meson potentials in the MFA are given by

$$V=\frac{1}{2}\left[(m_{w}w)^{2}+(m_{r}r)^{2}\right]+\frac{\alpha_{0}}{4}\left(w^{4}+r^{4}+3w^{2}r^{2}\right)$$

$$U=\frac{C_{0}}{2}\left(s^{2}+z^{2}\right)-\frac{C_{1}}{4}\left(s^{2}+z^{2}\right)^{2}-\frac{C_{2}}{4}\left(s^{4}+6s^{2}z^{2}+z^{4}\right)+h(s-f_{\pi})-\sum_{n=3,4}\frac{\alpha_{n}}{2^{n}n!}\left(s^{2}+z^{2}-f_{\pi}^{2}\right)^{n}+C_{3}.$$

The meson expectation values are determined by the extremum condition for $\Omega$, leading to the equations

$$s\,\left[-C_{0}+C_{1}\left(s^{2}+z^{2}\right)+C_{2}\left(s^{2}+3z^{2}\right)+\sum_{n=2,3}\frac{\alpha_{n+1}}{2^{n}n!}\left(s^{2}+z^{2}-f_{\pi}\right)^{n}\right]-h+\sum_{i=1,4}N_{i}\frac{\partial M_{i}}{\partial s}=0,$$

$$z\,\left[-C_{0}+C_{1}\left(s^{2}+z^{2}\right)+C_{2}\left(z^{2}+3s^{2}\right)+\sum_{n=2,3}\frac{\alpha_{n+1}}{2^{n}n!}\left(s^{2}+z^{2}-f_{\pi}\right)^{n}\right]+\sum_{i=1,4}N_{i}\frac{\partial M_{i}}{\partial z}=0,$$

$$w\left[m_{w}^{2}+\alpha_{0}\left(w^{2}+3r^{2}\right)\right]-g_{w}n_{B}=0,$$

$$r\left(m_{r}^{2}+3\alpha_{0}w^{2}\right)-g_{r}n_{I}=0.$$

The total barionic density is the sum of the partial contributions $n_{B}=\sum_{i}n_{i}$, as well as the isospin asymmetry $n_{I}=\sum_{i}I_{i}n_{i}$, where

$$n_{i}=2\,\int\frac{d^{3}p}{(2\pi)^{3}}\left[n_{F}(T,\mu_{i})-\bar{n}_{F}(T,\mu_{i})\right],$$

$$N_{i}=2\,\int\frac{d^{3}p}{(2\pi)^{3}}\,\frac{M_{i}}{E_{i}}\left[n_{F}(T,\mu_{i})+\bar{n}_{F}(T,\mu_{i})\right],$$

and $n_{F},\,\bar{n}_{F}$ stand for the canonical Fermi distribution functions for particles and antiparticles, and $E_{i}=\sqrt{p^{2}+M_{i}^{2}}$. Finally, the energy per unit volume is obtained by the thermodynamical relation ${\cal E}=\Omega+T\,{\cal S}+\sum_{i}\mu_{i}n_{i}$, where the entropy density is obtained as usual by ${\cal S}=-\partial\Omega/\partial T$.

This model has several free parameters, some of them can easily be related to the meson phenomenology. As in the LSM the constants $C_{0},C_{1},C_{2}$ are related to the meson masses in the MFA as

$$C_{0}=\left(m_{\sigma}^{2}-3m_{\pi}^{2}\right)/2,\;C_{1}=\left(m_{\sigma}^{2}-m_{\zeta}^{2}\right)/2\,f_{\pi}^{2},\;C_{2}=\left(m_{\zeta}^{2}-m_{\pi}^{2}\right)/2\,f_{\pi}^{2}.$$

Furthermore $h=f_{\pi}m_{\pi}^{2}$ is chosen in order that the vacuum expectation value of the scalar fields be $s_{0}=f_{\pi}=93$ MeV, $z_{0}=0$. Whereas $C_{3}$ is determined by requiring that the pressure $P=-\Omega$ vanishes at zero temperature and density. On the other hand, taking the nucleon mass $M_{1}^{0}=939$ MeV and choosing the chiral partner as $M_{2}^{0}=1530$ MeV, one obtains $G_{2}=\left(M_{1}^{0}-M_{2}^{0}\right)/2\,f_{\pi}$.

The remaining constants are used to adjust the nuclear matter phenomenology at $T=0$ and normal density $n_{0}=0.15$ fm^-3, and to obtain compatibility with observational data on compact stars. The following data on symmetric nuclear matter is used,

$$P(n_{0})=0,\;\frac{{\cal E}(n_{0})}{n_{0}}=923\,\text{MeV,}\;E_{\text{sym}}=\frac{n_{0}}{2}\,\frac{\partial^{2}{\cal E}}{\partial n_{I}^{2}}=30\,\text{MeV,}$$ (16)

$$L_{\text{sym}}=3\,n_{0}\,\frac{dE_{\text{sym}}}{dn}=70\,\text{MeV,}\;K=9\,n_{0}\,\frac{d\mu_{B}}{dn}=260\,\text{MeV}.$$ (17)

corresponding to the condition of minimum energy per particle, binding energy, symmetry energy and its slope, and the compressibility, respectively. Some of the chosen values are of qualitative nature since they have not been precisely measured yet. Furthermore, there is some tension between the commonly accepted values and the recent analysis of the experimental data on the symmetry properties $E_{\text{sym}}$ and $L$ [52, 53, 54].
Similarly, the nucleon effective mass in a dense medium $M^{*}$ is poorly known. However, a possible range of variation has been established, mainly due to nuclear structure estimations. In this work the range $0.65<M_{1}^{*}/M_{1}^{0}<0.8$ is explored for the nucleon mass, Eq. 14, at the normal density. For the same reason, the sigma meson mass is varied within the range $200$ MeV $<m_{\sigma}<500$ MeV.

Figure 1: The mass-radius relation for neutron stars for different parametrization. The upper (lower) panel corresponds to the cases with an effective nucleon mass fixed at $M_{1}^{*}/M_{1}^{0}=0.65$ ($M_{1}^{*}/M_{1}^{0}=0.8$).

III RESULTS AND DISCUSSION

In first place I try to tune up the undetermined parameters in order to reproduce the properties of nuclear matter just mentioned. For this purpose the selected domain in the $M_{1}^{*}-m_{\sigma}$ plane is explored. It is found that for a given value of $M_{1}^{*}$, these conditions can not be adjusted for the lowest value of $m_{\sigma}$. On the other hand, the parameter $\alpha_{0}$ of the $\omega-\rho$ mixing decreases with $m_{\sigma}$, and it becomes negative for $m_{\sigma}\sim 400$ MeV. This in turn, breaks down the equation of the $\omega$ expectation value for sufficiently high densities. Aiming to the study of neutron stars, densities up to $n/n_{0}=10$ are considered here. Therefore a more restrictive condition $300\,\text{MeV,}\leq m_{\sigma}\leq 400$ MeV must be imposed. It must be mentioned that small variations of the numerical values taken in Eqs. 16-17, produce continuous small changes in the mentioned results.
Applying the criterion just stated, a reduced set of parameters is selected to study star matter, that is nuclear matter stable against beta decay, and electrically neutral. As explained in more detail in the next section, a cloud of free electrons must be considered to fulfill these conditions. The equations of state obtained are introduced in the Tolman-Oppenheimer-Volkoff equation for the structure of a non-rotating compact star. The results for the mass-radius relation of the star are shown in Fig. 1. In the upper panel the results for $M_{1}^{*}/M_{1}^{0}=0.65$ and two cases with different $\sigma$ meson masses $m_{\sigma}=300$ MeV (A1), and $m_{\sigma}=391.4$ MeV (A2), corresponding to the vector meson mixing $\alpha_{0}=1241.4$ and $0$, respectively. In the lower panel, the case $M_{1}^{*}/M_{1}^{0}=0.8$ is presented for three different values $m_{\sigma}=300$ MeV (A3), $m_{\sigma}=400$ MeV (A4), and $m_{\sigma}=431$ MeV (A5). They correspond to $\alpha_{0}=437.8,\,60.1$, and $0$ respectively. As a first conclusions, it is found that the greater the value of $M_{1}^{*}$, the wider the range of admissible $m_{\sigma}$. In addition, the maximum mass of the star corresponds to the higher $m_{\sigma}$. Since there is empirical evidence of compact stars with a gravitational mass greater than two solar masses, we keep only the parametrizations A2 and A5, which are specified in Table 1 with the new labels A and B respectively.

Set	$m_{0}$ [MeV]	$G_{1}$	$g_{w}$	$g_{r}$	$\alpha_{3}$ [MeV-2]	$\alpha_{4}$ [MeV-4]	$m_{\sigma}$ [MeV]
A	493.72	12.17	11.76	9.61	-0.11462	0.03184	391.4
B	622.69	11.46	8.27	7.39	-0.10342	0.0494	431.1

Table 1: The two sets of parameters used in this work. For both cases is $G_{2}$=-3.18.

III.1 Neutron Stars

Different versions of the CDPM were used previously to study neutron stars, as for instance in [9, 15, 55, 16, 22, 26, 25, 17, 56, 36, 37]. Some of these works consider the presence of hyperons [17, 56] or the transition to deconfined quark matter [36, 37] or both of them, hence they are not directly comparable to the present results. In other cases [15, 55, 16], the mean field scheme was unable to reconcile the constraint of a gravitational mass greater than two solar masses. However, the addition of different physical considerations allows to fulfill this condition. For instance, the inclusion of vacuum effects in [55], or a sixth order polynomial self-interaction of the scalar mesons in [26, 25]. It is remarkable that such polynomial self-interaction, with an order higher than four, seems to be a key ingredient to reconcile the low density nuclear phenomenology and the astronomical data.
In this simplified treatment certain complemental degrees of freedom, such as hyperons and free quarks, are not taken into account. Several publications using the CDPM are devoted to this specific issue [17, 16, 18, 19, 20, 21, 22, 51, 23, 45]. The presence of hyperons could have significant effects on the structure of neutron stars, giving rise to the effect known as hyperon puzzle, but this is not the case for the CDPM. In fact, some calculations [22] have shown a negligible effect on the mass-radius relation of neutron stars.

Figure 2: The population of different particles in the stellar environment, in terms of the total density. The upper (lower) panel corresponds to the parametrization case A (case B).

For the sets A and B is $\alpha_{0}=0$, so the mean values of the vector mesons are simply given by $w=g_{w}n_{B}/m_{w}^{2}$ and $r=g_{r}n_{I}/m_{r}^{2}$. At zero temperature, the results for the pressure and the energy density are given by

$${\cal E}=\frac{1}{8\pi^{2}}\sum_{i=1-4,e}\left[p_{Fi}E_{Fi}\left(M_{i}^{2}+2\,p_{Fi}^{2}\right)-M_{i}^{4}\ln\left(\frac{p_{Fi}+E_{Fi}}{M_{i}}\right)\right]+V(w,r)-U(s,z),$$

$$P=\sum_{i}\mu_{i}n_{i}+\,\mu_{e}n_{e}-{\cal E},$$

in terms of the Fermi momentum $p_{Fi}$ and energy $E_{Fi}=\sqrt{p_{Fi}^{2}+M_{i}^{2}}$. The conservation of the baryon number density and electric charge are reflected by the equations $n_{B}=\sum_{i=1-4}n_{i}$ and $0=n_{1}+n_{3}-n_{e}$ respectively. As usual the number density for a particle is given by $n_{i}=p_{Fi}^{3}/3\,\pi^{2}$. The associated chemical potentials are related to the particle chemical potentials by $\mu_{B}=\mu_{2}=E_{F2}+g_{w}w-g_{r}r$, $\mu_{Q}=-\mu_{e}=E_{Fe}$. The stability against beta decay imposes $\mu_{2}-\mu_{e}=\mu_{1}$, and $\mu_{2}=\mu_{4}\;\mu_{1}=\mu_{3}$ if the dual partners are present. The onset of the dual partners takes place at densities for which $\mu_{2}=M_{4},\;\mu_{1}=M_{3}$.
The assumption of homogeneous matter is broken at very low densities, giving place to the formation of finite nuclei. For this reason, within such domain, the results obtained are joined in a continuous way and replaced by the equation of state given in [57]. This is the composite input used for the Tolman-Oppenheimer-Volkoff equation

$$\frac{dP}{dr} = -\,\frac{1}{r}\,[{\cal E}(r)+P(r)]\,\frac{{\cal M}(r)+4\pi r^{3}P(r)}{r-2{\cal M}(r)}\,,$$

$${\cal M}(r) = \int_{0}^{r}4\pi\,{r^{\prime}}^{2}\,{\cal E}(r^{\prime})\,dr^{\prime}\,.$$

to obtain the results shown in the next figures. Units for which $c=1,\,G=1,\,\hbar=1$ have been used.
In Fig.2 the population of the different particles are shown for a wide range of densities. As the density is increased the parity partner $n^{*}$ emerges earlier than the $p^{*}$. This is a consequence of two facts, beta stability and a non-zero mean field value $z>0$. The first condition requires that $\mu_{1}<\mu_{2}$, and the second one leads to $M_{3}>M_{4}$. Since $\mu_{2}$ increases with the density and from the inequalities $M_{3}>M_{4}\geq\mu_{2}>\mu_{1}$, it follows that the threshold $\mu_{2}=M_{4}$ is reached before than $\mu_{1}=M_{3}$. For the set A this happens at a notably low density, slightly lesser than $n_{B}/n_{0}\simeq 2$, while for the set B this fact is delayed until $n_{B}/n_{0}\simeq 4$. This feature can be explained by the lower value of $m_{0}$ for the set A, see Table 1, which makes the baryonic dynamics more reactive to the in-medium effects.
The threshold of the $n^{*}$ is marked by a strong decrease of the neutron population. The proton and electron curves coincide for densities below the threshold of $p^{*}$, as required by the charge neutrality condition. At relatively large densities, the parity partners are almost equally populated. The combined analysis of Figs. 2 and 3, the latter exhibiting the gravitational mass of a star in terms of its central density, shows that the stars having the standard mass $M/M_{\odot}=1.4$ are composed solely by nuclear matter. The odd parity $n^{*}$ enters in the composition of the stars having masses around the maximum value, and for the set A a scarce fraction of $p^{*}$ can also be found.

Figure 3: The gravitational mass of neutron stars in terms of its central density.

The onset of the dual partners is associated to thermodynamical instabilities, as shown in Fig. 4. Near their activation points the pressure becomes locally non-monotonous, with the exception of $n/n_{0}\simeq 3.5$ for the set A where the only manifestation is a slight change in the slope of the pressure. In such case an abrupt change in the compressibility of the system is expected. The small insert in this figure, shows in detail the low density region. For the set B a monotonous increasing pressure is obtained, while for the set A a van der Waals-like equation of state undoubtedly indicates the liquid-gas phase transition. This transition is a common feature of the nuclear matter, however it is unusual in the neutron star matter.

At this point it is useful a contrast with the more recent analysis for the massive pulsar PSR J0740+6620 having a measured mass $M/M_{\odot}=2.08\pm 0.07$. Its radius has been estimated within the range $11.7<R\,[\text{ km}]<14.25$ [58]. The result A, with a prediction of $R\simeq 16.3$ km is clearly out of the tolerance range, while the value $R\simeq 14.1$ km given by the set B, is acceptable although greater than the preferred estimation $R\simeq 12.8$ km. Another interesting piece of evidence corresponds to the compact object PSR J0030+0451, which was first identified as a neutron star with mass $M/M_{\odot}=1.4$ and a radius $R\simeq 13$ km. A recent revision [59] has estimated its mass and radius as $1.3<M/M_{\odot}<1.6$, and $11.7<R\,[\text{ km}]<12.9$. Within such mass domain the predictions of the present model are $R\simeq 15.7$ km (A), and $R\simeq 14.4$ km (B). However, another interpretation of the same observational data [60] considers that $M/M_{\odot}=1.7$, $R\simeq 14.5$ km, which is compatible with the result given by the set B.
The parametrization set A admits heavy stars $2<M/M_{\odot}<2.8$, but having relatively large radii $R>16$ km, which is in tension with presently accepted values. For this reason the results corresponding to the set A will be postponed, and only the set B is considered for the remaining of this work.

Figure 4: The pressure inside a neutron star in terms of the total density. The arrows indicate the position of the $n^{*}$ threshold, and of the $p^{*}$ only for the set A.

III.2 THE CHIRAL TRANSITION IN ISOSPIN IMBALANCED NUCLEAR MATTER

Up to this point an effective lagrangian has been defined that takes account of the vacuum masses of nucleons and their parity partners, as well as the masses of the pion and a0(980) meson. For the $\sigma$ meson mass a value consistent with the known phenomenology has been chosen. In addition, it is able to adjust the following properties of dense matter: saturation density, binding energy, compressibility, symmetry energy and its slope parameter. The effective nucleon mass at the normal nuclear density is slightly high, but it falls within the range of acceptable values. Finally, the prediction for the mass-radius of a neutron star passes the test of the results usually taken as standards. Thus the fitting of masses of the low lying mesons has been incorporated to the general scheme of fixing parameters, this criterium is similar to that used in [13, 61]. This is a common practice within the LSM which is not adequately considered in the applications of the CDPM.

In this section the isospin composition of matter is taken into account by introducing the isospin fraction $\text{w}=(n_{2}+n_{4}-n_{1}-n_{3})/n_{B}$ as a parameter. If the isospin composition is conserved in addition to the baryon density $n_{B}$, then $w$ is also a constant.

As in the last section, it is found that low density matter is composed only by protons and neutrons. But as the density is increased, their parity partners become active as it is shown in Fig. 5. This figure corresponds to zero temperature and includes several isospin fractions. For isospin symmetric matter ($\text{w}=0$) protons and neutrons are indistinguishable, as well as $p^{*}$ and $n^{*}$ are. By increasing the asymmetry, a splitting of the partners population is found. In addition, the onset of the $n^{*}$ is shifted downward, while that of $p^{*}$ is raised to higher densities. Thus, for $\text{w}=0.6$ the $p^{*}$ does not appear within the range analyzed in this figure and, consequently, its partner $p$ shows a constant trend. For pure neutron matter ($\text{w}=1$), the $n$ and $n^{*}$ have a specular behavior and asymptotically tend to an equipartition. This is a manifestation of the chiral recovery, since as the masses of the partners tends to the common value $m_{0}$, they become statistically indistinguishable.

Figure 5: The population of different baryons in dense matter in terms of the total baryonic density for zero temperature and different isospin fractions,.

Figure 6: The pressure in terms of a limited range of densities and several isospin asymmetries at zero temperature.

The emergence of the dual partners leaves its imprints in the equation of state, as shown in Fig. 6. For each of the isospin compositions considered here, the pressure is monotonically increasing until the presence of the $n^{*}$ is strong enough to cause a first order phase transition. At this point the pressure shows a typical van der Waals behavior concentrated in a narrow range of density, with its unstable and metastable sections. It can be appreciated that a small change in the asymmetry drastically changes the location of the instability, for this reason only a limited range for w is shown in this figure. The location of the critical density progressively decreases with w, from $n_{B}/n_{0}\simeq 5.3$ for symmetric matter it falls to $n_{B}/n_{0}\simeq 2.7$ for pure neutron matter.

Figure 7: The pressure in terms of the density, limited to the range of the instability. Several isospin asymmetries and temperatures are considered.

To take account of the thermal effects on the pressure, the instability domain is shown for several asymmetries and temperatures in Fig. 7. For a given isospin fraction the pressure curves become flatter as the temperature increases, until a maximum value $T_{m}$ where they acquire a monotonously increasing trend. This figure shows that $T_{m}$ has a dependence on w, for instance for $\text{w}=0.2$ is $T_{m}\simeq 5$ MeV, then it increases for $\text{w}=0$ and $0.6$ within the range $8<T_{m}<10$, and finally it exceeds $T=10$ MeV for $\text{w}=1$.

Figure 8: The maximum temperature for the first order chiral transition in terms of the isospin asymmetry.

Two conclusions can be drawn from this figure, the critical temperature for this first order phase transition should take place for the maximum asymmetry, furthermore the temperature $T_{m}$ takes on a minimum value for nonzero but low asymmetries. The specific details are shown in Fig. 8, where $T_{m}$ is shown in terms of w. A sudden drop of almost $30\%$ is found for matter with a slight neutron excess $\text{w}=0.02$ as compared with the perfectly isospin symmetric case. Up from this point, this characteristic temperature grows uniformly and reaches its maximum value $T_{c}=11.2$ MeV for pure neutron matter.
The comparison with published results using the DCPM is adequate at this point, although the criterium for the selection of parameters is not strictly coincident. For instance the maximum temperature for the first order chiral phase transition at $\text{w}=0$ is near $T_{m}\sim 10$ MeV in [31], and $T_{m}=7$ MeV in [29]. The critical density under the same conditions, varies within $5.9<n_{B}/n_{0}<7.9$ when the parameter $m_{0}$ ranges from $m_{0}=600$ MeV to $m_{0}=700$ MeV. In the present work a slightly lower value $n_{B}/n_{0}=5.3$ is obtained for the set B, which has a parameter $m_{0}$ within the mentioned interval, see Table 1. A very different result $n_{B}/n_{0}=7.8$ is obtained in [29], notwithstanding it is consistent with the use of a relatively high $m_{0}=750$ MeV.

The phase transition is the consequence of the violation of the thermodynamical equilibrium criteria, which can be stated in terms of the change of convexity of the free energy ${\cal F}(T,n_{B},n_{I})$. The conditions

$$\frac{\partial^{2}{\cal F}}{\partial n_{B}^{2}}>0,\;\;\frac{\partial^{2}{\cal F}}{\partial n_{B}^{2}}\,\frac{\partial^{2}{\cal F}}{\partial n_{I}^{2}}-\left(\frac{\partial^{2}{\cal F}}{\partial n_{B}\partial n_{I}}\right)^{2}>0$$ (18)

guarantees the stability of a bi-component system [62, 63]. They are usually known as the mechanical and chemical stability conditions respectively. In Fig. 9 these relations have been used to map the instability region, which is bounded by the spinodal curve. In the main panel the case $T=5$ MeV is representative for temperatures below $T\simeq 6.5$ MeV. It is evident that unstable matter can be found for the whole range $\text{w}>0$, and temperatures below $T_{c}$.

Figure 9: The spinodal surface in the $n_{B}-\text{w}$ plane for two characteristic temperatures. In order to show its full extension, a scale is used for which the surface seems like a thin curve.

Furthermore, the phase transition is sharply focused in the phase space as implied by the fact that the spinodal surface seems like a thin curve for the scale used. There are two separate branches, the low density one spreads over the full range $0<\text{w}<1$, and corresponds to the violation of both mechanical and chemical equilibrium. It is associated with the emergence of the neutral partner $n^{*}$, since the spinodal coincides with the threshold where this particle reaches a concentration of about $5\%$. The charged partner $p^{*}$, instead, has a negligible presence for $0<w<0.3$ or it is missing at all for $\text{w}>0.3$. In contrast, the high density branch only breaks the chemical equilibrium but fulfills the mechanical condition. It marks the point where the particle $p^{*}$ becomes thermodynamical relevant. It extends only up to $w=0.3$ because at this temperature the incoming of $p^{*}$ is disfavored for greater asymmetries.
A slight variation of this situation is found as the temperature is raised, as shown in the inserted panel in the same figure. In this case the low density branch splits into separate parts, one of them exists for $\text{w}>0.22$ because $T_{m}<7$ MeV for $\text{w}<0.22$ (see Fig. 8). The other part is restricted to almost symmetric matter $\text{w}<0.003$ and in this figure it could be mistaken as the starting point of the high density branch. The gap in w separating both parts increases with the temperature.
The existence of two disjoint branches of unstable matter, spreading over very different domains of the phase space, remarks that the information provided by the equation of state of Fig. 6 is insufficient to sketch a panoramic view in the present model. In order to improve this issue, the behavior of the chemical potentials in terms of the isospin asymmetry is shown in Fig. 10. For this purpose two representative baryon densities are chosen, in the upper panel the case $n_{B}/n_{0}=4$ is affected by the low density branch, which is characterized by both mechanical instability and a non-monotonous increase of $\mu_{2}$ with the asymmetry. On the lower panel, instead, the feature of the high density branch is exhibited for $n_{B}/n_{0}=6$. In this case the loss of thermodynamical equilibrium manifests in the irregular decrease of $\mu_{1}$ with w, since the onset of $p^{*}$ leads this instability as mentioned earlier.

Figure 10: The chemical potential of the neutron (proton) in the upper (lower) panel in terms of the isospin asymmetry for two representative densities and several temperatures. The non-monotonous trend is a signal of the violation of the chemical equilibrium.

A complete study of asymmetric matter within the CDPM was given in [32], using an approach similar to the present one. The main differences are the use of the $a_{0}(980)$ meson and the lack of the Dirac sea effects in the present work. The regularization at zero temperature made in [32] comprises the tree level divergence of the grand potential in Eq. 15, and also the vertices $C_{0}$ and $C_{1}+C_{2}$, which are beyond that order. The comparison of the spinodal surface in the w-$n_{B}$ plane shows that the effect of the scalar isovector meson is really significant. Furthermore, the critical temperature in [32] is $T_{c}=8.5$ MeV and it takes place for symmetric matter. In contrast, a grater value $T_{c}=11.2$ is obtained here, and fundamentally it occurs for pure neutron matter.

Isospin symmetric matter as well as pure neutron matter play a special role, since in both instances there is only one chemical potential for all the particles. In the first case both branches of instability coincide, and in the latter case the proton and its chiral partner are excluded by definition, eliminating the high density branch. However, for both of them the initial and final points of the phase transition preserve the isospin asymmetry. It is interesting to study the energy balance under such conditions. It is well known that a first order phase transition gives rise to a latent heat, which for an isothermal process at constant chemical potential can be evaluated as

$$\ell=T\left(\;\;\left[\frac{{\cal S}}{n_{B}}\right]_{b}-\left[\frac{{\cal S}}{n_{B}}\right]_{a}\right),$$

the subindexes represent the extreme states of the transition, and by convention $a$ stands for the denser one. The results displayed in Fig. show the typical bell-shaped curves, with a scale of the order of keV, several orders of magnitude lesser than the results commonly obtained for the liquid-gas phase transition. This is in accordance with the fact that the spinodal has a considerably reduced extension in the phase space.

IV SUMMARY AND CONCLUSIONS

In this work the possible manifestations of the chiral symmetry in dense nuclear matter have been studied, using a model of hadronic fields which realise the chiral symmetry under the mirror assignment [8, 10]. For this purpose the CDPM introduces the parity partners $n,\,n^{*}$ and $p,\,p^{*}$. The parameters of the model have been been determined in the MFA, imposing conditions over a wide range of matter densities. The model adjusts the masses in vacuum of the nucleons and their parity partners as well as those of the pions and the low lying scalar mesons $\sigma$ and $a_{0}(980)$. It also describes the key features of nuclear matter at the saturation density and zero temperature. The highest densities have been included through the requirement of predicting a mass-radius relation for neutron stars that is compatible with the recent analysis [58, 59, 60].
The focus has been posed on isospin asymmetric matter at finite termperature, therefore the model includes the interaction with the vector isovector $\rho$ and scalar isovector $\zeta$ mesons, the latter identified as the $a_{0}(980)$. Initially, a coupling between vector mesons was proposed (Eq. 3), including a $\omega-\rho$ mixing vertex. However, it has been found that the reproduction of the phenomenological constraints does not favor its inclusion.
The emergence of the parity partners to the Fermi sea produces thermodynamical instabilities which lead to first order phase transition. These instabilities are described by two separate branches, the low density one has the strongest effects and manifests by a non-monotonous behavior of the equation of state. It has a critical temperature $T_{c}=11.2$ MeV. The other branch is generated by diffusive instabilities associated with the emergence of $p^{*}$, and takes place for higher densities. The spinodal extends over the full range of isospin asymmetries $0\leq\text{w}\leq 1$, although it is extremely concentrated. Instabilities of the same characteristics have been found even for neutron star matter. The coupling to the $\zeta$ meson is the main cause of such drastic behavior, since its coupling to the nucleons distinguish the isospin structure. Therefore a splitting of the effective nucleon masses occurs, giving a different dynamical response for each member of the isospin duplet. This effect reinforces the difference in the energy spectrum due to the coupling to the vector isovector meson. However, the influence of the $\rho$ meson must be of lesser magnitude since this phenomenon has not been reported in previous investigations which only include the vector meson in their conceptual framework to describe isospin imbalanced dense matter. Thus, it must be concluded that a complete treatment of the chiral transition under such conditions should not do without the coupling of the nucleons to the scalar isovector meson.
The specific latent heat, evaluated for symmetric nuclear matter and pure neutron matter, shows that the energy released by a particle participating in the chiral phase transition is of the order of the keV. This result is significantly lower than other characteristic transitions, as for instance the liquid-gas conversion at low density.

There are interesting issues, not considered in this work, which will be object of further investigation such as the effect of vacuum contributions or the possible consequences of the chiral phase transition on the dynamics of neutron stars.