betaFeedback

WorksPhysics and Astronomy

Chiral phase transition of a dense, magnetized and rotating quark matter

Abstract

We investigate the chiral symmetry restoration/breaking of a dense, magnetized and rotating quark matter within the Nambu Jona-Lasinio model including Nf=2N_f=2 and Nc=3N_c=3 numbers of flavors and colors, respectively. Imposing the spectral boundary conditions, as well as the positiveness of energy levels, lead to a correlation between the magnetic and rotation fields such that strongly magnetized plasma can not rotate anymore. We solve the gap equation at zero and finite temperature. At finite temperature and baryon chemical potential μBμ_B, we sketch the phase diagrams Tc(μB)T_c(μ_B) and Tc(RΩ)T_c(RΩ) in different cases. As a result, we always observe inverse-rotational catalysis mean to decrease TcT_c by increasing RΩRΩ. But the magnetic field has a more complex structure in the phase diagram. For slowly rotating plasma, we find that TcT_c decreases by increasing eBeB, while in the fast rotating plasma we see that TcT_c increases by increasing eBeB. Also, we locate exactly the position of Critical End Point by solving the equations of first and second derivatives of effective action with respect to the order parameters, simultaneously.

Introduction

Quark Matter (QM) experiences phase transitions under extreme external conditions [1]. These external conditions could be temperature, chemical potential, magnetic or rotation fields, and so on. Examining the phase transition of QM has revealed much valuable information including the existence of the Quark-Gluon Plasma (QGP) phase that contains deconfined quarks and gluons at very high temperatures [2, 3]. At very dense systems, the quarks bind to each other to break the color symmetry and produce the so-called color-superconductivity phase [4]. It may arise weird phases under special conditions which depend on the considered symmetry [5]. The phase transition properties can serve as Equation of State (EoS) data or inputs to the dynamical evolutions of QM at Large-Hadron Colliders (LHC) and that is why they are so important.

Magnetic field eBeB and rotational fields Ω\Omega have much relevance to the QM studies. In a non-central collisions of heavy ions the spectator particles create a strong magnetic field of order eB∼0.1GeV\sqrt{eB} \sim0.1GeV for the RHIC and eB∼0.5GeV\sqrt{eB} \sim0.5GeV for the LHC [6]. Also, the non-central collision may produce a large amount of angular momenta which results in the most vortical fluid ever seen [7]. The strength of the rotational field in LHC is about 1022s−110^{22}s^{-1} which is very larger than any observed vortical fluids in nature. Needless to say that these huge magnetic and rotational fields can potentially affect the QM properties.

There is much more known about the properties of QM under the external magnetic field. Most of the low-energy models predicted that chiral symmetry breaking is enhanced as a function of eBeB, see the review article [8]. Magnetic fields have a potential to induce variety of exciting effects in the thermodynamics of QCD. Some effective model calculations have shown that the transition temperature between the hadron matter and QM increases with eBeB [9, 10]. This conclusion was proved in the Sakai-Sugimoto model [11] and within the holographic approach [12]. Furthermore, splitting between the deconfinement and chiral transitions was predicted to take place for large eBeBs. Also, the strength of the transition was seen to increase and results in a first-order phase transition [13]. However, it was observed the opposite effect of decreasing transition temperature with increasing eBeB. This observation was confirmed in the lattice QCD [14, 15, 16], in the holographic models [17] and in the quark-hadron phase transition [18]. The Strong magnetic field can also reveal the non-trivial structures of gauge-field topology as a chiral magnetic effect [19, 20].

In recent years, studies of QM under the strong rotation have got lots of interests. Here, we refer to some of them. The uncharged rotating fermionic matter phase diagrams were studied within the Nambu-Jona Lasinio (NJL) model [21, 22]. A two-flavor NJL model was used to study the phase diagrams of charged quarks under a rotation [23]. Also, the chiral phase transition of a two-flavor NJL model in a rotating sphere was investigated by considering the finite-size effects [24]. Lattice QCD techniques have been developing to study the phase diagrams of rotating QM and they need more attention [25]. The influence of helicity imbalance on the phase diagrams of dense QM was already discussed [26]. Another interesting topic is the possibility of charged pion condensation in external magnetic and rotation fields and it takes much attention [27, 28, 29]. There is a good review on the QCD phase diagram under the rotation and magnetic field that states the basics of this topic [30]. In the paper [31], the authors study the uncharged fermions phase diagrams within the one-flavor NJL model in presence of constant magnetic and rotation fields and a global boundary condition.

A missing chain of the QM phase diagram studies in magnetic and rotation fields is to consider situations that could be so close to the real world experiments. Motivated by this in the current paper, we investigate the chiral phase transition of a bounded two-flavor and three-colors NJL model at finite chemical potential and temperature in presence of constant magnetic and rotation fields.

A part of our work is devoted to developing a general framework to obtain the effective action of chiral order parameter σ\sigma starting from the NJL model in curved space. It has used the Ritus method as a general approach that can solve the Dirac equation in many cases [32]. The obtained master formula for the in-medium effective action is characterized by the Ritus eigenvalues and we benefit from it to study the phase diagrams.

We also try to solve the Dirac equation for quarks with different charges in a constantly rotating and magnetized plasma. This will be done by using the Ritus method, similar to the paper [31]. Eigenfunctions are written in terms of the hypergeometric functions and eigenvalues (generalized Landau levels) are determined by applying appropriate boundary condition. We select the spectral boundary condition on which QM is confined in a cylinder with radius R and RΩ≤1R\Omega\leq1 because of the causality. Furthermore, the net flux going through the boundary Rα=1ΩR\alpha=\frac{1}{\Omega} is zero. Because of this boundary condition and the positiveness condition of energy eigenvalues, the magnetic field strength ¹ and the rotation field RΩR\Omega turn out to be correlated. We find that if α≤αc=7\alpha\leq\alpha_c=7 then 0≤RΩ≤10 \leq R\Omega\leq1, while for α≥αc\alpha\geq\alpha_c we obtain 0≤RΩ≤RΩMax0 \leq R\Omega\leq R\Omega_{\mathrm{Max}} where this maximum value RΩMaxR\Omega_{\mathrm{Max}} decreases by increasing α\alpha. It is such that for α∼30\alpha\sim30 and beyond, the only allowed values of RΩMax∼0R\Omega_{\mathrm{Max}}\sim0. Therefore, strongly magnetized plasma can not rotate anymore. To the best of our knowledge, this is a non-trivial and novel effect that is rooted in the applied boundary condition.

The gap equation is solved numerically at zero and finite temperature for the QM with Nc=3N_c=3 and Nf=2N_f=2. Some features of the phase diagram in our model are in common with the normal NJL model (no rotation or magnetic field) such as increasing of order parameter by increasing the coupling constant or momentum cut-off. A consistency occurs for the needed coupling constant to provide a non-zero dynamical mass at small magnetic fields with the normal NJL coupling threshold, i.e. GcΛ2∼2π2NcNfG_c\Lambda^2 \sim\frac{2\pi^2}{N_cN_f}. The strength of this threshold coupling decreases by increasing α\alpha which is a clear sign of magnetic-catalysis. Near the border x∼αx\sim\alpha, ² the dynamical mass starts to fall because of the violation of slowly varying field assumption, ∂rσ≪σ2\partial_r\sigma\ll\sigma^2. We observe an interesting effect near the boundary, namely the surface magnetic catalysis that is due to the mode accumulation [33]. Since vorticity does not play a role at zero temperature, we don not observe the rotational-magnetic inhibition [34]. The profile σ(α)\sigma(\alpha) seems to be very oscillatory at small magnetic fields and no assertive statement can be made on the magneto-catalysis or inverse magneto-catalysis.

At finite temperature, the phase diagram can be viewed as Tc(μ)T_c(\mu) or Tc(RΩ)T_c(R\Omega) planes.³ Below these curves we get the chirally broken phase, while above them the chiral symmetry is restored. We fit the transition points of the diagrams to a polynomial function in order to compare them better with the Lattice results [35]. It will be found that the magnetic and rotation fields increase the curvature of the phase diagram compared to the normal model. The important feature is the interplay among the magnetic and rotation fields that has enriched the physics of the phase diagram. As a consequence, by fixing α\alpha and μ\mu, TcT_c always decreases with RΩR\Omega which is a sign of inverse-rotational catalysis. But fixing RΩR\Omega gives complicated diagrams that in slowly rotating systems we have inverse-magneto catalysis, while in fast rotating matters the dominant pattern of the phase diagram is magneto-catalysis. To the best of our knowledge, this is a novel situation that has not been seen so far. This feature comes from the boundary condition and the relation between magnetic and rotation fields. We will present a new technique to find the location of Critical End Point(CEP) by solving the gap equation and its derivatives with respect to the σ\sigma field, simultaneously. Another finding is that the more rotational the system is, the smaller the CEP is. This is a clear sign of inverse-rotational catalysis. Also, the location of CEP and its profile in the (RΩ,Tc)(R\Omega,T_c) plane depend on α\alpha. Slope of the curve TCEP(RΩ)T_{\mathrm{CEP}}(R\Omega) or dTCEPdRΩ\frac{dT_{\mathrm{CEP}}}{dR\Omega} at each fixed μ\mu for smaller α\alpha is bigger than larger α\alpha. It is such that at very large α\alpha, this slope closes to zero. As already mentioned, at large α\alpha the RΩMax→0R\Omega_{\mathrm{Max}}\to0 and thus (RΩ−Tc)(R\Omega-T_c) diagram for large α\alpha is a point on the vertical axis.

The Organization of the paper is as follows. First in section 2, we construct our general framework to obtain the effective action of the “σ\sigma” field in curved space by utilizing the Ritus method. Then in section 3, we try to solve the Dirac equation for different quark flavors and this task is done similarly to the procedures of the paper [31]. We derive the energy levels as well as the generalized Landau levels and impose the spectral boundary condition as well as the positiveness condition of energy levels to set a proper quantization scheme. Next in section 4, by using the master effective action formula we solve numerically the gap equation at zero and finite temperature by keeping (α,RΩ)(\alpha,R\Omega) fixed. Many plots are sketched and their physical consequences are discussed separately. Finally in section 5, we close our paper with a brief conclusion and outlook to future works.

Effective action

In this section, we aim to derive a concrete expression for the effective action of static auxiliary fields (σ,πi)(\sigma,\pi_i) in the NJL model with Nf=2N_f=2 and NcN_c number of flavors and colors, respectively. We can assume the presence of external fields in this general framework which could be either constant magnetic fields or rotating frames or even both of them. Although this subject has been known in detail, we shall try to give a general formula for the effective action. After that, we describe its variants by including those fields.

The quantum field theory of spin 12\frac{1}{2} particles in curved spaces needs the use of vierbein tensors eμae_\mu^a. They link the global curved space with metric gμνg_{\mu\nu} to the local flat space with metric ηab\eta_{ab} via the identity gμν=eμaeνbηabg_{\mu\nu}=e_\mu^a e_\nu^b\eta_{ab}. Therefore, a quantity like AμA_\mu connects to AaA_a or vice versa through the relation 6

Aμ=eμaAa,Aa=eaμAμ.(1)A_\mu=e_\mu^a A_a,\qquad A_a=e_a^\mu A_\mu. \tag*{(1)}

We use the orthogonality relations ebμeμa=δbae_b^\mu e_\mu^a=\delta_b^a or eaμeνa=δνμe_a^\mu e_\nu^a=\delta_\nu^\mu for transformation between the local and global coordinates. Consider the Nf=2N_f=2 local NJL action in the massless limit in a curved space 7

SNJL=∫x−gLNJL,(2)S_{\mathrm{NJL}}=\int_x\sqrt{-g}\mathcal{L}_{\mathrm{NJL}}, \tag*{(2)}
LNJL=∑f=u,d(ψˉf(i̸D(f)+μBγ0)ψf+G2((ψˉfψf)2+(ψˉfiγ5τiψf)2)).\mathcal{L}_{\mathrm{NJL}}=\sum_{f=u,d}\left(\bar{\psi}_f\left(i\not{D}^{(f)}+\mu_B\gamma^0\right)\psi_f+\frac{G}{2}\left((\bar{\psi}_f\psi_f)^2+(\bar{\psi}_f i\gamma_5\tau_i\psi_f)^2\right)\right).

Here, “gg” denotes the determinant of metric, ψf=(ψuψd)\psi_f=\begin{pmatrix}\psi_u\\\psi_d\end{pmatrix} represents the fermion doublet, GG is the strength of interaction, μB\mu_B is the baryonic chemical potential and τi(i=1,2,3)\tau_i(i=1,2,3) are the Pauli matrices. We use the convention ∫x≡∫d4x\int_x\equiv\int d^4x and ̸D(f)=γμDμ(f)=γaDa(f)\not{D}^{(f)}=\gamma^\mu D_\mu^{(f)}=\gamma^aD_a^{(f)}. Generally, the kernel Dμ(f)D_\mu^{(f)} is a matrix in Dirac space and includes external fields contribution for each flavor as well as the free partial term

Dμ(f)=∂μ+Dμ(f).(3)D_\mu^{(f)}=\partial_\mu+D_\mu^{(f)}. \tag*{(3)}

Introducing the auxiliary fields (σ,πi)(\sigma,\pi_i) would enable us to derive the effective action Veff\mathcal{V}_{\mathrm{eff}} of those fields in the Eq. (II.2) via the bosonization procedure

Z=∫∏i=13Dσ Dπi∏f=u,dDψfDψˉf ei∫x−gLB=∫∏i=13Dσ Dπi ei∫x−gVeff,(4)Z=\int\prod_{i=1}^{3}D\sigma\,D\pi_i\prod_{f=u,d}D\psi_fD\bar{\psi}_f\,e^{i\int_x\sqrt{-g}\mathcal{L}_B} =\int\prod_{i=1}^{3}D\sigma\,D\pi_i\,e^{i\int_x\sqrt{-g}\mathcal{V}_{\mathrm{eff}}}, \tag*{(4)}

with the following action

LB=−12G(σ2+πiπi)+∑f=u,dψˉf(i̸D(f)+μBγ0−σ−iγ5τiπi)ψf.(5)\mathcal{L}_B=-\frac{1}{2G}(\sigma^2+\pi_i\pi_i)+\sum_{f=u,d}\bar{\psi}_f\left(i\not{D}^{(f)}+\mu_B\gamma^0-\sigma-i\gamma_5\tau_i\pi_i\right)\psi_f. \tag*{(5)}

This action equals to its counterpart given by the Eq. (II.2) in the level of equations of motion

σ=−G∑f=u,d⟨ψˉfψf⟩,πi=−G∑f=u,d⟨ψˉfiγ5τiψf⟩.\sigma=-G\sum_{f=u,d}\langle\bar{\psi}_f\psi_f\rangle,\qquad\pi_i=-G\sum_{f=u,d}\langle\bar{\psi}_f i\gamma_5\tau_i\psi_f\rangle.

The equivalence between the action (II.2) and (II.5) can also be seen from the partition function point of view, which by integrating over the (σ,πi)(\sigma,\pi_i) fields, we obtain the original action. Throughout this paper, we set πi=0\pi_i=0. Thus, Veff\mathcal{V}_{\mathrm{eff}} can be given as follows

Veff=−12Gσ2+∑f=u,dlog⁡det⁡(i̸D(f)+μBγ0−σ).(6)\mathcal{V}_{\mathrm{eff}}=-\frac{1}{2G}\sigma^2+\sum_{f=u,d}\log\det\left(i\not{D}^{(f)}+\mu_B\gamma^0-\sigma\right). \tag*{(6)}

Our interesting systems let us to simplify the second term of the Eq. (II.6) due to the property [γ5,Da(f)]=0[\gamma_5,D_a^{(f)}]=0

log⁡det⁡(i̸D(f)+μBγ0−σ)=Tr⁡log⁡(i̸D(f)+μBγ0−σ)=12Tr⁡log⁡(i̸D(f)+μBγ0−σ)+12Tr⁡log⁡(γ5(i̸D(f)+μBγ0−σ)γ5)=12Tr⁡log⁡((̸D(f))2−iμB{̸D(f),γ0}−i[̸D(f),σ]−μB2+σ2).(7)\begin{aligned} \log\det\left(i\not{D}^{(f)}+\mu_B\gamma^0-\sigma\right)=\operatorname{Tr}\log\left(i\not{D}^{(f)}+\mu_B\gamma^0-\sigma\right) \\ &=\frac{1}{2}\operatorname{Tr}\log\left(i\not{D}^{(f)}+\mu_B\gamma^0-\sigma\right)+\frac{1}{2}\operatorname{Tr}\log\left(\gamma_5\left(i\not{D}^{(f)}+\mu_B\gamma^0-\sigma\right)\gamma_5\right) \\ &=\frac{1}{2}\operatorname{Tr}\log\left((\not{D}^{(f)})^2-i\mu_B\left\{\not{D}^{(f)},\gamma^0\right\}-i\left[\not{D}^{(f)},\sigma\right]-\mu_B^2+\sigma^2\right). \tag*{(7)} \end{aligned}

The third term of the last line expression can be ignored because of the slowly varying field assumption, i.e. ∂rσ≪σ2\partial_r\sigma\ll\sigma^2 and the second term goes to {̸D(f),γ0}=2D0(f)\{\not{D}^{(f)},\gamma^0\}=2D_0^{(f)}. Therefore, the effective potential results in

Veff=−12Gσ2+12∑f=u,dTr⁡log⁡((̸D(f))2−2iμBD0(f)−μB2+σ2).(8)\mathcal{V}_{\mathrm{eff}}=-\frac{1}{2G}\sigma^2+\frac{1}{2}\sum_{f=u,d}\operatorname{Tr}\log\left((\not{D}^{(f)})^2-2i\mu_BD_0^{(f)}-\mu_B^2+\sigma^2\right). \tag*{(8)}

We utilize the Ritus method to obtain the contribution of the second term of Eq. (II.8). It can help us to solve the Dirac equation and obtain the eigenvalues and eigenstates [36, 37]

iD(f)ψM,κ(f)(pˉ)=κψM,κ(f)(pˉ)γap~a(f),i\mathcal{D}^{(f)}\psi^{(f)}_{M,\kappa}(\bar{p})=\kappa\psi^{(f)}_{M,\kappa}(\bar{p})\gamma^{a}\tilde{p}^{(f)}_{a},
iD0(f)ψM,κ(f)(pˉ)=κp~0(f)ψM,κ(f)(pˉ).(9)i\mathcal{D}^{(f)}_{0}\psi^{(f)}_{M,\kappa}(\bar{p})=\kappa\tilde{p}^{(f)}_{0}\psi^{(f)}_{M,\kappa}(\bar{p}). \tag*{(9)}

Here, κ=±1\kappa=\pm1 refers to the positive and negative frequency eigenmodes and ψM,κ(f)(pˉ)\psi^{(f)}_{M,\kappa}(\bar{p}) stands for eigenstates of the Dirac equation. Also, the corresponding eigen-four-momentum p~a(f)=(p~0(f),p~1(f),p~2(f),p~3(f))\tilde{p}^{(f)}_{a}=(\tilde{p}^{(f)}_{0},\tilde{p}^{(f)}_{1},\tilde{p}^{(f)}_{2},\tilde{p}^{(f)}_{3}) is a model-dependent quantity and the collection of quantum numbers which identifies the eigenstates is called by MM. The center-four-momentum pˉa=(pˉ0,pˉ1,pˉ2,pˉ3)\bar{p}_{a}=(\bar{p}_{0},\bar{p}_{1},\bar{p}_{2},\bar{p}_{3}) is the eigenvalues of center members of free kernels ∂ˉc(f)\bar{\partial}^{(f)}_{c}, those which commute with other kernels Dk(f)D^{(f)}_{k}. Therefore, this center-momentum appears in the Fourier expansion

ψM,κ(f)(pˉ)=eiκpˉ⋅xˉψM,κ(f)(x⊥).(10)\psi^{(f)}_{M,\kappa}(\bar{p})=e^{i\kappa\bar{p}\cdot\bar{x}}\psi^{(f)}_{M,\kappa}(x_{\perp}). \tag*{(10)}

where x⊥x_{\perp} is the perpendicular subspace to the xˉ\bar{x} space. Usually, the center-momentum space includes energy and one of the original momenta of particles. Due to the independence of the trace from the chosen bases, we select the eigenstates of the Eq. (II.9) to compute the second term of the Eq. (II.8). We sandwich the operator of the Eq. (II.8) between the eigenstates with the same quantum numbers given by the Ritus ansatz

Veff=−12Gσ2+12∑f,M∫dpˉ(2π)dTr⁡(ψM,+1(f)†(pˉ)log⁡((D(f))2−2iμBD0(f)−μB2+σ2)ψM,+1(f)(pˉ))=−12Gσ2+Nc2∑f,M∫dpˉ(2π)dTr⁡(ψM,+1(f)†(pˉ)ψM,+1(f)(pˉ))log⁡(−(p~0(f)+μB)2+E(f)2).(11)\begin{aligned} \mathcal{V}_{\mathrm{eff}}=-\frac{1}{2G}\sigma^{2}+\frac{1}{2}\sum_{f,M}\int\frac{d\bar{p}}{(2\pi)^{d}}\operatorname{Tr}\left(\psi^{(f)\dagger}_{M,+1}(\bar{p})\log\left((\mathcal{D}^{(f)})^{2}-2i\mu_{B}\mathcal{D}^{(f)}_{0}-\mu_{B}^{2}+\sigma^{2}\right)\psi^{(f)}_{M,+1}(\bar{p})\right) \\ &=-\frac{1}{2G}\sigma^{2}+\frac{N_{c}}{2}\sum_{f,M}\int\frac{d\bar{p}}{(2\pi)^{d}}\operatorname{Tr}\left(\psi^{(f)\dagger}_{M,+1}(\bar{p})\psi^{(f)}_{M,+1}(\bar{p})\right)\log\left(-\left(\tilde{p}^{(f)}_{0}+\mu_{B}\right)^{2}+\mathcal{E}^{(f)2}\right). \tag*{(11)} \end{aligned}

where E(f)≡σ2+∑i=13(p~i(f))2\mathcal{E}^{(f)}\equiv\sqrt{\sigma^{2}+\sum_{i=1}^{3}(\tilde{p}^{(f)}_{i})^{2}} and dd is the number of non-zero elements in pˉa\bar{p}_{a} space. In the last line trace acts on the Ritus bases. The Eq. (II.11) is our general formula for the effective action in the zero temperature limit (T=0T=0) and in curved space irrespective of external fields. In some cases we could simplify further the effective action such as integrating over the pˉ\bar{p} space. Likewise, the Ritus functions are chosen to satisfy the orthogonality condition as follows

∫xψM,κ(f)†(pˉ′)ψM′,κ′(f)(pˉ)=δM,M′δκ,κ′δ(pˉ′−pˉ).(12)\int_{x}\psi^{(f)\dagger}_{M,\kappa}(\bar{p}^{\prime})\psi^{(f)}_{M^{\prime},\kappa^{\prime}}(\bar{p})=\delta_{M,M^{\prime}}\delta_{\kappa,\kappa^{\prime}}\delta(\bar{p}^{\prime}-\bar{p}). \tag*{(12)}

Generalization of the Eq. (II.11) to T≠0T\ne0 is straightforward. To this purpose, we have to notice that

p~0(f)=pˉ0+LM(f),(13)\tilde{p}^{(f)}_{0}=\bar{p}_{0}+L^{(f)}_{M}, \tag*{(13)}

where LM(f)L^{(f)}_{M} is a generalized and model-dependent angular-momentum contribution. This new part arises due to the curvature of space via the spin-connection term. It means that in general p~0(f)≠pˉ0(f)\tilde{p}^{(f)}_{0}\ne\bar{p}_{0}^{(f)}, but in flat space LM(f)=0L^{(f)}_{M}=0. To obtain the effective action at finite temperature in the imaginary time formalism the following replacements should be made

pˉ0→iωn=i(2n+1)πT,∫dpˉ02π→iT∑n=−∞∞.\bar{p}_{0}\to i\omega_{n}=i(2n+1)\pi T,\qquad\int\frac{d\bar{p}_{0}}{2\pi}\to iT\sum_{n=-\infty}^{\infty}.

We follow the normal procedure to sum over the Matsubara frequencies of the Eq. (II.11) [38] and then we

values of center members of free kernels ∂ˉc(f)\bar{\partial}^{(f)}_{c}, those that commute with other kernels Dk(f)D^{(f)}_{k}. Therefore, this center-momentum appears in the Fourier expansion

ψM,κ(f)(pˉ)=eiκpˉ⋅xˉψM,κ(f)(x⊥).(14)\psi^{(f)}_{M,\kappa}(\bar{p})=e^{i\kappa\bar{p}\cdot\bar{x}}\psi^{(f)}_{M,\kappa}(x_{\perp}). \tag*{(14)}

arrive at

Veff=−12Gσ2+Nc2∑f,M∫dpˉ(2π)d−1Tr⁡(ψM,+1(f)†(pˉ)ψM,+1(f)(pˉ))VT,\mathcal{V}_{\mathrm{eff}}=-\frac{1}{2G}\sigma^{2}+\frac{N_{c}}{2}\sum_{f,M}\int\frac{d\bar{p}}{(2\pi)^{d-1}}\operatorname{Tr}\left(\psi^{(f)\dagger}_{M,+1}(\bar{p})\psi^{(f)}_{M,+1}(\bar{p})\right)\mathcal{V}_{T},
VT≡E(f)+Tlog⁡(1+e−β(E(f)+μB+LM(f)))\mathcal{V}_{T}\equiv\mathcal{E}^{(f)}+T\log\left(1+e^{-\beta(\mathcal{E}^{(f)}+\mu_{B}+L^{(f)}_{M})}\right)
+Tlog⁡(1+e−β(E(f)−μB−LM(f))).(15)\qquad\qquad+T\log\left(1+e^{-\beta(\mathcal{E}^{(f)}-\mu_{B}-L^{(f)}_{M})}\right). \tag*{(15)}

In this relation pˉ\bar{p} is the non-vanishing spatial part of the center-momenta and β=1T\beta=\frac{1}{T}. The Eq. (II.14) is our general master formula for the in-medium effective action of σ\sigma field in curved space. In what follows we try to clarify our proposal with a few examples.

Veff in eB = =0

Effective action of σ\sigma field in the medium of free quarks is a very well-known issue. We would like to compare our results with the existing results. In the case of eB=Ω=0eB=\Omega=0, the eigenstates and eigenmomenta defined in the Eq. (II.9), are given by

ψM(f)(pˉ)=eipˉ⋅x14×4,pˉa=p~a(f)=(p0,p1,p2,p3),LM(f)=0,\psi^{(f)}_{M}(\bar{p})=e^{i\bar{p}\cdot x}\mathbb{1}_{4\times4},\qquad\bar{p}_{a}=\tilde{p}^{(f)}_{a}=(p_{0},p_{1},p_{2},p_{3}),\qquad L^{(f)}_{M}=0,

where 14×41_{4\times4} is the rank 4 identity matrix. Inserting these values into the Eq. (II.14), we obtain

Veff0=−σ22G+2NcNf∫d3p(2π)3VT0\mathcal{V}_{\rm eff}^{0}=-\frac{\sigma^{2}}{2G}+2N_{c}N_{f}\int\frac{d^{3}p}{(2\pi)^{3}}\mathcal{V}_{T}^{0}
VT0=E0+Tlog⁡(1+e−β(E0+μB))+Tlog⁡(1+e−β(E0−μB)),(16)\mathcal{V}_{T}^{0}=\mathcal{E}_{0}+T\log\left(1+e^{-\beta(\mathcal{E}_{0}+\mu_{B})}\right)+T\log\left(1+e^{-\beta(\mathcal{E}_{0}-\mu_{B})}\right), \tag*{(16)}

where E0=σ2+∑i=13pi2\mathcal{E}_{0}=\sqrt{\sigma^{2}+\sum_{i=1}^{3}p_{i}^{2}} and d3p=dp1 dp2 dp3d^{3}p=dp_{1}\,dp_{2}\,dp_{3}.

Veff\mathcal{V}_{\rm eff} in eB≠0eB\ne0, Ω=0\Omega=0

In a non-rotating medium (Ω=0\Omega=0) with a constant magnetic field aligned in the zz axis B⃗=Bz^\vec{B}=B\hat{z}, the eigensatets and eigenmomenta of particles with charge QfQ_{f} are given in terms of harmonic oscillator functions [37]

ψM(f)(pˉ)=eipˉ⋅xˉ(P+sffn+sf(ξx)+ΠnP−sffn−sf(ξx)),\psi_{M}^{(f)}(\bar{p})=e^{i\bar{p}\cdot\bar{x}}\left(\mathcal{P}_{+}^{sf}f_{n}^{+sf}(\xi_{x})+\Pi_{n}\mathcal{P}_{-}^{sf}f_{n}^{-sf}(\xi_{x})\right),
P±sf=1±isfγ1γ22,sf=sgn⁡(eQfB),Πn=1−δn,0,\mathcal{P}_{\pm}^{sf}=\frac{1\pm is_{f}\gamma_{1}\gamma_{2}}{2},\qquad s_{f}=\operatorname{sgn}(eQ_{f}B),\qquad\Pi_{n}=1-\delta_{n,0},
f+n(ξx)=1ℓB2nn!πe−ξx22Hn(ξx),ξx=x1−sfp2ℓB2ℓB,f_{+}^{n}(\xi_{x})=\frac{1}{\sqrt{\ell_{B}2^{n}n!\sqrt{\pi}}}e^{-\frac{\xi_{x}^{2}}{2}}H_{n}(\xi_{x}),\qquad\xi_{x}=\frac{x_{1}-s_{f}p_{2}\ell_{B}^{2}}{\ell_{B}},
ℓB2=1∣eQfB∣,fn−sf(ξx)=fn−1+sf(ξx),LM(f)=0,\ell_{B}^{2}=\frac{1}{\lvert eQ_{f}B\rvert},\qquad f_{n}^{-sf}(\xi_{x})=f_{n-1}^{+sf}(\xi_{x}),\qquad L_{M}^{(f)}=0,
pˉa(f)=(p0,0,p2,p3),pˉ~a(f)=(p0,0,−sf2n∣eQfB∣,p3).(17)\bar{p}_{a}^{(f)}=(p_{0},0,p_{2},p_{3}),\qquad\widetilde{\bar{p}}_{a}^{(f)}=\left(p_{0},0,-s_{f}\sqrt{2n\lvert eQ_{f}B\rvert},p_{3}\right). \tag*{(17)}

The number "n=j+(1−sfσ3)/2n=j+(1-s_{f}\sigma_{3})/2" labels the Landau levels and σ3=±1\sigma_{3}=\pm1 represent the eigenvalues of spin. From this definition we see that the lowest Landau level has one spin degrees of freedom, while two spin degrees of freedom occupy higher Landau levels. Plugging the eigenstates and eigenmomenta of the Eq. (II.16) into the Eq. (II.14), we obtain the following result

VeffB=−σ22G+Nc∑f=u,d∑n=0∞∫dp2 dp3(2π)2Bn(f)(x)VTB,\mathcal{V}_{\rm eff}^{B}=-\frac{\sigma^{2}}{2G}+N_{c}\sum_{f=u,d}\sum_{n=0}^{\infty}\int\frac{dp_{2}\,dp_{3}}{(2\pi)^{2}}\mathcal{B}_{n}^{(f)}(x)\mathcal{V}_{T}^{B},
Bn(f)(x)≡(fn+sf(ξx))2+Πn(fn−sf(ξx))2,\mathcal{B}_{n}^{(f)}(x)\equiv\left(f_{n}^{+sf}(\xi_{x})\right)^{2}+\Pi_{n}\left(f_{n}^{-sf}(\xi_{x})\right)^{2},
VTB=EB(f)+Tlog⁡(1+e−β(EB(f)+μB))+Tlog⁡(1+e−β(EB(f)−μB)),\mathcal{V}_{T}^{B}=\mathcal{E}_{B}^{(f)}+T\log\left(1+e^{-\beta(\mathcal{E}_{B}^{(f)}+\mu_{B})}\right)+T\log\left(1+e^{-\beta(\mathcal{E}_{B}^{(f)}-\mu_{B})}\right),
EB(f)=σ2+2n∣eQfB∣+p32.(18)\mathcal{E}_{B}^{(f)}=\sqrt{\sigma^{2}+2n\lvert eQ_{f}B\rvert+p_{3}^{2}}. \tag*{(18)}

The p2p_{2} integration is doable by a simple shift of fn±sf(ξx)f_{n}^{\pm sf}(\xi_{x}) argument

∫dp2(fn±sf(ξx))2=ℓB−1∫dξx(fn±sf(ξx))2=ℓB−2.\int dp_{2}\left(f_{n}^{\pm sf}(\xi_{x})\right)^{2}=\ell_{B}^{-1}\int d\xi_{x}\left(f_{n}^{\pm sf}(\xi_{x})\right)^{2}=\ell_{B}^{-2}.

Therefore, the final result of the effective action simplifies as

VeffB=−σ22G+Nc∑f=u,d∑n=0∞∣eQfB∣2παn∫dp32πVTB,(19)\mathcal{V}_{\rm eff}^{B}=-\frac{\sigma^{2}}{2G}+N_{c}\sum_{f=u,d}\sum_{n=0}^{\infty}\frac{\lvert eQ_{f}B\rvert}{2\pi}\alpha_{n}\int\frac{dp_{3}}{2\pi}\mathcal{V}_{T}^{B}, \tag*{(19)}

in which αn=2−δn,0\alpha_{n}=2-\delta_{n,0} counts the degeneracy of Landau levels.

Veff\mathcal{V}_{\rm eff} in eB=0eB=0, Ω≠0\Omega\ne0 with boundary condition

This problem is less known in community compared to the former cases. However, we can derive it by using the master Eq. (II.14). In a non-magnetized system (eB=0eB=0) which rotates uniformly along the z^\hat{z} axis Ω⃗=Ωz^\vec{\Omega}=\Omega\hat{z}, the radial size constrains to 0≤r≤R=1Ω0\le r\le R=\frac{1}{\Omega} because of the causality requirement. Eigenstates and eigenmomenta of particles in this system are given below

ψM(f)(pˉ)=eipˉ⋅xˉ(P+fℓ+(x)+P−fℓ−(x)),P±=1±iγ1γ22,\psi_{M}^{(f)}(\bar{p})=e^{i\bar{p}\cdot\bar{x}}\left(\mathcal{P}_{+}f_{\ell}^{+}(x)+\mathcal{P}_{-}f_{\ell}^{-}(x)\right),\qquad\mathcal{P}_{\pm}=\frac{1\pm i\gamma_{1}\gamma_{2}}{2},
fℓ±(x)=2Ωeiℓ±ϕJℓ±(qℓ,jrΩ)Jℓ(qℓ,j),ℓ±=j∓12,f_{\ell_{\pm}}(x)=\sqrt{2\Omega}e^{i\ell_{\pm}\phi}\frac{J_{\ell_{\pm}}(q_{\ell,j}r\Omega)}{J_{\ell}(q_{\ell,j})},\qquad\ell_{\pm}=j\mp\frac{1}{2},
pˉa(f)=(p0,0,0,p3),pˉ~a(f)=(p~0,0,qℓ,jΩ,p3),\bar{p}_{a}^{(f)}=(p_{0},0,0,p_{3}),\qquad\widetilde{\bar{p}}_{a}^{(f)}=(\widetilde{p}_{0},0,q_{\ell,j}\Omega,p_{3}),
LM(f)=LzΩ=Ωj,ℓ=Ωj.(20)L_{M}^{(f)}=L_{z}^{\Omega}=\Omega j,\qquad\ell=\Omega j. \tag*{(20)}

where “ℓ\ell” is the eigenvalue of orbital angular momentum Lz=−i∂ϕL_{z}=-i\partial_{\phi} and j=ℓ+12j=\ell+\frac{1}{2} is the eigenvalue of total angular momentum. Jℓ(x)J_{\ell}(x) is the first kind of Bessel functions with order ℓ\ell and qℓ,jq_{\ell,j} is the jj’th zero of ℓ\ell’th Bessel function. We use the global (spectral) boundary condition on which the total flux of fermion’s current on the cylinder surface R=1ΩR=\frac{1}{\Omega} becomes zero. Final result of the effective action in this medium is derived after plugging the eigenstates and eigenmomenta of the Eq. (II.19) into the Eq. (II.14)

VeffΩ=−σ22G+2NcNfΩ2π∑ℓ=−∞∞∑j=1∞∫dp32πJℓ(r)VTΩ,\mathcal{V}_{\rm eff}^{\Omega}=-\frac{\sigma^{2}}{2G}+\frac{2N_{c}N_{f}\Omega^{2}}{\pi}\sum_{\ell=-\infty}^{\infty}\sum_{j=1}^{\infty}\int\frac{dp_{3}}{2\pi}\mathcal{J}_{\ell}(r)\mathcal{V}_{T}^{\Omega},
Jℓ(r)≡Jℓ2(qℓ,jrΩ)+Jℓ+12(qℓ,jrΩ)Jℓ+12(qℓ,j),\mathcal{J}_{\ell}(r)\equiv\frac{J_{\ell}^{2}(q_{\ell,j}r\Omega)+J_{\ell+1}^{2}(q_{\ell,j}r\Omega)}{J_{\ell+1}^{2}(q_{\ell,j})},
VTΩ=EΩ+Tlog⁡(1+e−β(EΩ+μB+LℓΩ))+Tlog⁡(1+e−β(EΩ−μB−LℓΩ)),\mathcal{V}_{T}^{\Omega}=\mathcal{E}^{\Omega}+T\log\left(1+e^{-\beta(\mathcal{E}^{\Omega}+\mu_{B}+L_{\ell}^{\Omega})}\right)+T\log\left(1+e^{-\beta(\mathcal{E}^{\Omega}-\mu_{B}-L_{\ell}^{\Omega})}\right),
EΩ=σ2+qℓ,j2Ω2+p32.(21)\mathcal{E}^{\Omega}=\sqrt{\sigma^{2}+q_{\ell,j}^{2}\Omega^{2}+p_{3}^{2}}. \tag*{(21)}

Veff\mathcal{V}_{\rm eff} in eB≠0eB\ne0, Ω≠0\Omega\ne0 with boundary condition

To the best of our knowledge, this problem has not been solved so far and we intend to obtain the effective action of σ\sigma field in a magnetized and rotating medium from the master formula, the Eq. (II.14). Indeed, the main purpose of this section is to derive this relation. We consider a bounded system which Ω=Ωz^\boldsymbol{\Omega}=\Omega\hat{\boldsymbol{z}} and B=Bz^\boldsymbol{B}=B\hat{\boldsymbol{z}}. Eigenstates and eigenmomenta for particles with charge QfQ_f are given below [31]

Ψλ,ℓ,sf(f)(pˉ)=eipˉ⋅Xˉ(Pλ,ℓ,sf(sf)+fλ,ℓ,sf+(X)+Pλ,ℓ,sf(sf)−fλ,ℓ,sf−(X)),\Psi^{(f)}_{\lambda,\ell,s_f}(\bar{p})=e^{i\bar{p}\cdot\bar{X}}\left(\mathcal{P}^{(s_f)+}_{\lambda,\ell,s_f}f^+_{\lambda,\ell,s_f}(X)+\mathcal{P}^{(s_f)-}_{\lambda,\ell,s_f}f^-_{\lambda,\ell,s_f}(X)\right),
fλ,ℓ,sf±=Cλ,ℓ,sf±eiℓ±ϕΨλ,ℓ,sf±(X),f^\pm_{\lambda,\ell,s_f}=C^\pm_{\lambda,\ell,s_f}e^{i\ell_\pm\phi}\Psi^\pm_{\lambda,\ell,s_f}(X),
Ψλ,ℓ,sf±(X)=a±e−X2X∣ℓ±∣21F1(−Nλ,ℓ,sf±;∣ℓ±∣+1;X),\Psi^\pm_{\lambda,\ell,s_f}(X)=a_\pm e^{-\frac{X}{2}}X^{\frac{|\ell_\pm|}{2}}{}_1F_1\left(-\mathcal{N}^{\pm}_{\lambda,\ell,s_f};|\ell_\pm|+1;X\right),
a±≡1∣ℓ±∣!(∣eQfB∣2π(Nλ,ℓ,sf±+∣ℓ±∣)!Nλ,ℓ,sf±!)12,(22)a_\pm\equiv\frac{1}{|\ell_\pm|!}\left(\frac{|eQ_fB|}{2\pi}\frac{(\mathcal{N}^{\pm}_{\lambda,\ell,s_f}+|\ell_\pm|)!}{\mathcal{N}^{\pm}_{\lambda,\ell,s_f}!}\right)^{\frac{1}{2}}, \tag*{(22)}
Cλ,ℓ,sf±=(∣eQfB∣2π∫0αQdX(Ψλ,ℓ,sf±(X))2)12,C^\pm_{\lambda,\ell,s_f}=\left(\frac{|eQ_fB|}{2\pi\int_0^{\alpha_Q}dX\left(\Psi^\pm_{\lambda,\ell,s_f}(X)\right)^2}\right)^{\frac{1}{2}},
pˉa(f)=(pˉ0,0,0,p3),p~a(f)=(p~0,0,κsℓ2λ∣eQfB∣,p3),\bar{p}^{(f)}_a=(\bar{p}_0,0,0,p_3),\qquad\widetilde{p}^{(f)}_a=\left(\widetilde{p}_0,0,\kappa s_\ell\sqrt{2\lambda|eQ_fB|},p_3\right),

where

λ=(p~0(f))2−p32−σ22∣eQfB∣,αQ≡∣eQfB∣R22,\lambda=\frac{(\widetilde{p}^{(f)}_0)^2-p_3^2-\sigma^2}{2|eQ_fB|},\qquad\alpha_Q\equiv\frac{|eQ_fB|R^2}{2},
Nλ,ℓ,sf±=λ+sfℓ∓−∣ℓ±∣−12,(23)\mathcal{N}^{\pm}_{\lambda,\ell,s_f}=\lambda+\frac{s_f\ell_\mp-|\ell_\pm|-1}{2}, \tag*{(23)}
X≡∣eQfB∣r22,LM(f)=Lℓ(f)=Ωj.X\equiv\frac{|eQ_fB|r^2}{2},\qquad L_M^{(f)}=L_\ell^{(f)}=\Omega j.

The 1F1(a;b;X){}_1F_1(a;b;X) is the first kind of confluent hypergeometric functions and ℓ±=j∓12\ell_\pm=j\mp\frac{1}{2}. We elaborate on these functions in the next section. Inserting the appropriate quantities into the Eq. (II.14) would enable us to obtain the effective action of the σ\sigma field

VeffB,Ω=−σ22G+Nc∑f=u,d∑ℓ=−∞∞∫dp32πGλ,ℓ,sf(X)VTB,Ω,\mathcal{V}_{\mathrm{eff}}^{B,\Omega}=-\frac{\sigma^2}{2G}+N_c\sum_{f=u,d}\sum_{\ell=-\infty}^{\infty}\int\frac{dp_3}{2\pi}\mathcal{G}_{\lambda,\ell,s_f}(X)\mathcal{V}^{B,\Omega}_T,
Gλ,ℓ,sf(X)≡(Cλ,ℓ,sf+)2((Ψλ,ℓ,sf+(X))2+(Ψλ,ℓ,sf−(X))2),\mathcal{G}_{\lambda,\ell,s_f}(X)\equiv\left(C^+_{\lambda,\ell,s_f}\right)^2\left(\left(\Psi^+_{\lambda,\ell,s_f}(X)\right)^2+\left(\Psi^-_{\lambda,\ell,s_f}(X)\right)^2\right),
VTB,Ω≡EB,Ω+Tlog⁡(1+e−β(EB,Ω+μB+LℓΩ))+Tlog⁡(1+e−β(EB,Ω−μB−LℓΩ)),\mathcal{V}^{B,\Omega}_T\equiv\mathcal{E}^{B,\Omega}+T\log\left(1+e^{-\beta(\mathcal{E}^{B,\Omega}+\mu_B+L_\ell\Omega)}\right) +T\log\left(1+e^{-\beta(\mathcal{E}^{B,\Omega}-\mu_B-L_\ell\Omega)}\right),
EB,Ω=2λ∣eQfB∣+p32+σ2.(24)\mathcal{E}^{B,\Omega}=\sqrt{2\lambda|eQ_fB|+p_3^2+\sigma^2}. \tag*{(24)}

Throughout this paper, we take this effective action as our starting point for the numerical computations and derive the chiral phase diagram of Nf=2N_f=2 and Nc=3N_c=3 QM.

Analytical method

We put the QM system in a cylinder with radius R that rotates uniformly along the z^\hat{z} axis, Ω=Ωz^\boldsymbol{\Omega}=\Omega\hat{\boldsymbol{z}}. Hence, the line element associated with the particle trajectories in 3+13+1 dimensions is

ds2=(1−r2Ω2)dt2−dx2−dy2−dz2+2Ωdt(ydx−xdy).(25)ds^2=(1-r^2\Omega^2)dt^2-dx^2-dy^2-dz^2+2\Omega dt(ydx-xdy). \tag*{(25)}

The coordinate system is xμ=(t,x,y,z)=(t,rcos⁡ϕ,rsin⁡ϕ,z)x^\mu=(t,x,y,z)=(t,r\cos\phi,r\sin\phi,z) and −g=1\sqrt{-g}=1. The metric of Eq. (III.25) is not curved because all the Riemann tensor components are zero, Rabcd=0R^a{}_{bcd}=0, but it is curvilinear which means Γabc≠0\Gamma^a{}_{bc}\ne0. To connect the global curved space to the local flat space, we use the vierbeins. Due to the local Lorentz symmetry, there are more choices to the eμae^a_\mu. We choose the following bases[21, 31]

et0=ex1=ey2=ez3=1,e0x=yΩ,e0y=−xΩ.(26)e^0_t=e^1_x=e^2_y=e^3_z=1,\qquad e^x_0=y\Omega,\quad e^y_0=-x\Omega. \tag*{(26)}

Likewise, the orthogonality condition eμaebμ=δbae^a_\mu e^\mu_b=\delta^a_b rules

et0=ex1=ey2=ez3=1,et1=−yΩ,et2=xΩ.(27)e^0_t=e^1_x=e^2_y=e^3_z=1,\qquad e^1_t=-y\Omega,\quad e^2_t=x\Omega. \tag*{(27)}

Our model Lagrangian is given by Eq. (II.2) with turning off the Pauli term. It results in the following Dirac equation for massive charged fermion with flavor ff

(iγμDμ(f)−σ)ψf(x)=0.(28)\left(i\gamma^\mu D^{(f)}_\mu-\sigma\right)\psi_f(x)=0. \tag*{(28)}

The corresponding Dμ(f)D^{(f)}_\mu kernel contains free partial term as well as the interacting Dμ(f)D^{(f)}_\mu contribution which comes from the magnetic and rotational fields

Dμ(f)=−ieQfAμ−i4ωμabσab,(29)D^{(f)}_\mu=-ieQ_fA_\mu-\frac{i}{4}\omega_{\mu ab}\sigma^{ab}, \tag*{(29)}

where

σab≡i2[γa,γb]\sigma^{ab}\equiv\frac{i}{2}\left[\gamma^a,\gamma^b\right]

is the spin matrix,

ωμab≡gαβeaα(∂μebβ+Γμνβebν)\omega_{\mu ab}\equiv g_{\alpha\beta}e^\alpha_a\left(\partial_\mu e^\beta_b+\Gamma^\beta_{\mu\nu}e^\nu_b\right)

is the spin connection and the Christoffel symbol is Γμνβ=12gβα(∂μgαν+∂νgαμ−∂αgμν)\Gamma^\beta_{\mu\nu}=\frac{1}{2}g^{\beta\alpha}(\partial_\mu g_{\alpha\nu}+\partial_\nu g_{\alpha\mu}-\partial_\alpha g_{\mu\nu}). We set the local gauge field to be Aa=(0,A)=(0,−By2,Bx2,0)A_a=(0,\boldsymbol{A})=(0,-\frac{By}{2},\frac{Bx}{2},0) and therefore the magnetic field is Ba=Bz^\boldsymbol{B}_a=B\hat{\boldsymbol{z}}. The vierbein choices in Eqs. (III.25) and (III.26) leads to the following components for Dμ(f)D^{(f)}_\mu

Dt(f)=−ieQfBΩr22−i2σ12,Dx(f)=ieQfBy2,D^{(f)}_t=-ieQ_f\frac{B\Omega r^2}{2}-\frac{i}{2}\sigma^{12},\qquad D^{(f)}_x=ieQ_f\frac{By}{2},
Dy(f)=−ieQfBx2,Dz(f)=0.(30)D^{(f)}_y=-ieQ_f\frac{Bx}{2},\qquad D^{(f)}_z=0. \tag*{(30)}

Also, the Dirac matrices in the global curved space are

γt=γ0,γx=γ1+yΩγ0,γy=γ2−xΩγ0,γz=γ3.(31)\gamma^t=\gamma^0,\qquad\gamma^x=\gamma^1+y\Omega\gamma^0,\qquad\gamma^y=\gamma^2-x\Omega\gamma^0,\qquad\gamma^z=\gamma^3. \tag*{(31)}

Therefore, the Dirac equation (III.27) reads to

(γ⋅D(f)−σ)ψf=0,(32)\left(\gamma\cdot D^{(f)}-\sigma\right)\psi_f=0, \tag*{(32)}
γ⋅D(f)=iγ0(∂t−iΩJ^z)+iγ1(∂x+i2eQfBy)+  iγ2(∂y−i2eQfBx)+iγ3∂z,\gamma\cdot D^{(f)}=i\gamma^0\left(\partial_t-i\Omega\hat{J}_z\right)+i\gamma^1\left(\partial_x+\frac{i}{2}eQ_fBy\right) +\;i\gamma^2\left(\partial_y-\frac{i}{2}eQ_fBx\right)+i\gamma^3\partial_z,

which J^z=L^z+Σz2\hat{J}_z=\hat{L}_z+\frac{\Sigma_z}{2} is total angular momentum operator with L^z=−i(x∂y−y∂x)=−i∂ϕ\hat{L}_z=-i(x\partial_y-y\partial_x)=-i\partial_\phi is the azimuthal angular momentum and Σz=iγ1γ2\Sigma_z=i\gamma^1\gamma^2 is the spin matrix along the z^\hat{z} direction.

According to the Ritus ansatz, we choose ψλ,ℓ,κ(f)=Eλ,ℓ,κ(f)uκ(p~λ,ℓ,κ)\psi_{\lambda,\ell,\kappa}^{(f)}=\mathrm{E}_{\lambda,\ell,\kappa}^{(f)}u_\kappa(\tilde{p}_{\lambda,\ell,\kappa}) in such a way that γ⋅D(f)Eλ,ℓ,κ(f)=κEλ,ℓ,κ(f)γ⋅p~λ,ℓ,κ\gamma\cdot D^{(f)}\mathrm{E}_{\lambda,\ell,\kappa}^{(f)}=\kappa\mathrm{E}_{\lambda,\ell,\kappa}^{(f)}\gamma\cdot\tilde{p}_{\lambda,\ell,\kappa}. The main job of Ritus method is to find the eigenbasis Eλ,ℓ,κ(f)\mathrm{E}_{\lambda,\ell,\kappa}^{(f)} and the eigenmomenta p~λ,ℓ,κμ\tilde{p}_{\lambda,\ell,\kappa}^{\mu}. Due to the Hamiltonian structure of the Eq. (III.31), the eigenbasis Eλ,ℓ,κ(f)\mathrm{E}_{\lambda,\ell,\kappa}^{(f)} has well-defined quanta under the (∂t,J^z,∂z)(\partial_t,\hat{J}_z,\partial_z) operators

i∂tEλ,ℓ,κ(f)=κEλ,ℓ,κ(f)Eλ,ℓ,κ(f),i∂zEλ,ℓ,κ(f)=κpzEλ,ℓ,κ(f),i\partial_t\mathrm{E}_{\lambda,\ell,\kappa}^{(f)}=\kappa E_{\lambda,\ell,\kappa}^{(f)}\mathrm{E}_{\lambda,\ell,\kappa}^{(f)},\qquad i\partial_z\mathrm{E}_{\lambda,\ell,\kappa}^{(f)}=\kappa p_z\mathrm{E}_{\lambda,\ell,\kappa}^{(f)},
J^zEλ,ℓ,κ(f)=jEλ,ℓ,κ(f),(33)\hat{J}_z\mathrm{E}_{\lambda,\ell,\kappa}^{(f)}=j\mathrm{E}_{\lambda,\ell,\kappa}^{(f)}, \tag*{(33)}

with j≡ℓ+12j\equiv\ell+\frac{1}{2}. Regarding the latter equation, the eigenbasis can be written as

Eλ,ℓ,κ(f)=e−iκ(Eλ,ℓ,κ(f)t−pzz)∑p=±1P±(f)fλ,ℓ,κpsf(r,ϕ),(34)\mathrm{E}_{\lambda,\ell,\kappa}^{(f)}=e^{-i\kappa(E_{\lambda,\ell,\kappa}^{(f)}t-p_zz)}\sum_{p=\pm1}\mathcal{P}_{\pm}^{(f)}f_{\lambda,\ell,\kappa}^{p s_f}(r,\phi), \tag*{(34)}

which P±(f)=1±isfγ1γ22\mathcal{P}_{\pm}^{(f)}=\frac{1\pm i s_f\gamma^1\gamma^2}{2} is the spin projector operator, sf≡sgn⁡(eQfB)s_f\equiv\operatorname{sgn}(eQ_fB) and λ\lambda is the generalized Landau levels. They are integers for systems without boundaries, while in systems with boundaries they are rational numbers. If we write fλ,ℓ,κ±(r,ϕ)=eiℓ±ϕFλ,ℓ,κ±(r)f_{\lambda,\ell,\kappa}^{\pm}(r,\phi)=e^{i\ell_{\pm}\phi}F_{\lambda,\ell,\kappa}^{\pm}(r), then the governing equation for Fλ,ℓ,κ±(r)F_{\lambda,\ell,\kappa}^{\pm}(r) will be

(X∂X2+∂X+λ−ℓ±24X+sfℓ±2−X4)Fλ,ℓ(f)±(X)=0,(35)\left(X\partial_X^2+\partial_X+\lambda-\frac{\ell_{\pm}^2}{4X}+\frac{s_f\ell_{\pm}}{2}-\frac{X}{4}\right)F_{\lambda,\ell}^{(f)\pm}(X)=0, \tag*{(35)}

where X≡x∣Qf∣=∣eQfB∣r22X\equiv x|Q_f|=\frac{|eQ_fB|r^2}{2} and

λ≡(Eλ,ℓ,κ(f)+κΩj)2−pz2−σ22∣eQfB∣.(36)\lambda\equiv\frac{(E_{\lambda,\ell,\kappa}^{(f)}+\kappa\Omega j)^2-p_z^2-\sigma^2}{2|eQ_fB|}. \tag*{(36)}

The quanta λ\lambda can be seen as the eigenvalues of Eλ,ℓ,κ(f)\mathrm{E}_{\lambda,\ell,\kappa}^{(f)} in the perpendicular direction to the eBeB and Ω\Omega axis

γ⊥⋅D⊥(f)Eλ,ℓ,κ(f)=κsℓ2λ∣eQfB∣γ2Eλ,ℓ,κ(f),(37)\gamma_\perp\cdot D_\perp^{(f)}\mathrm{E}_{\lambda,\ell,\kappa}^{(f)}=\kappa s_\ell\sqrt{2\lambda|eQ_fB|}\gamma^2\mathrm{E}_{\lambda,\ell,\kappa}^{(f)}, \tag*{(37)}

with sℓ=sgn⁡(ℓ)s_\ell=\operatorname{sgn}(\ell) and γ⊥⋅D⊥(f)≡γ1Dx(f)+γ2Dy(f)\gamma_\perp\cdot D_\perp^{(f)}\equiv\gamma^1D_x^{(f)}+\gamma^2D_y^{(f)}. Solutions to the Eq. (III.34) which are confined within a boundary RR can be represented as

Fλ,ℓ(f)±(X)=Cλ,ℓ(f)±Ψλ,ℓ(f)±(X),(38)F_{\lambda,\ell}^{(f)\pm}(X)=C_{\lambda,\ell}^{(f)\pm}\Psi_{\lambda,\ell}^{(f)\pm}(X), \tag*{(38)}

where the two variables Ψλ,ℓ(f)±(X)\Psi_{\lambda,\ell}^{(f)\pm}(X) and Cλ,ℓ(f)±C_{\lambda,\ell}^{(f)\pm} are

Ψλ,ℓ(f)±(X)≡a±e−X/2X∣ℓ±∣/2 1F1(−Nλ,ℓ,sf±;∣ℓ±∣+1;X),\Psi_{\lambda,\ell}^{(f)\pm}(X)\equiv a_{\pm}e^{-X/2}X^{|\ell_{\pm}|/2}\,{}_1F_1(-\mathcal{N}_{\lambda,\ell,s_f}^{\pm};|\ell_{\pm}|+1;X),
a±≡1ℓ±!(∣eQfB∣2π(Nλ,ℓ,sf±+∣ℓ±∣)!Nλ,ℓ,sf±!)1/2,a_{\pm}\equiv\frac{1}{\ell_{\pm}!}\left(\frac{|eQ_fB|}{2\pi}\frac{(\mathcal{N}_{\lambda,\ell,s_f}^{\pm}+|\ell_{\pm}|)!}{\mathcal{N}_{\lambda,\ell,s_f}^{\pm}!}\right)^{1/2},
Cλ,ℓ(f)±=(∣eQfB∣2π∫0αBdX(Ψλ,ℓ±(X))2)1/2,(39)C_{\lambda,\ell}^{(f)\pm}=\left(\frac{|eQ_fB|}{2\pi\int_0^{\alpha_B}dX(\Psi_{\lambda,\ell}^{\pm}(X))^2}\right)^{1/2}, \tag*{(39)}

with

Nλ,ℓ,sf±≡λ+sfℓ∓−∣ℓ±∣−12.(40)\mathcal{N}_{\lambda,\ell,s_f}^{\pm}\equiv\lambda+\frac{s_f\ell_{\mp}-|\ell_{\pm}|-1}{2}. \tag*{(40)}

This number is given for different charged quarks and spin in Table I. For positive charges changing λ→λ−1\lambda\to\lambda-1 and ℓ→ℓ+1\ell\to\ell+1 turn positive spin to negative spin solutions, while for negative charges λ→λ−1\lambda\to\lambda-1 and ℓ→ℓ−1\ell\to\ell-1 turn negative spin to positive spin solutions. Likewise, the eigenmomenta are determined to be as

sf>0s_f>0Nλ,ℓ≥0,++=λ\mathcal{N}_{\lambda,\ell\geq0,+}^{+}=\lambdaNλ,ℓ≤−1,++=λ+ℓ\mathcal{N}_{\lambda,\ell\leq-1,+}^{+}=\lambda+\ell
sf>0s_f>0Nλ,ℓ≥0,+−=λ−1\mathcal{N}_{\lambda,\ell\geq0,+}^{-}=\lambda-1Nλ,ℓ≤−1,+−=λ+ℓ\mathcal{N}_{\lambda,\ell\leq-1,+}^{-}=\lambda+\ell
sf<0s_f<0Nλ,ℓ≥0,−+=λ−ℓ−1\mathcal{N}_{\lambda,\ell\geq0,-}^{+}=\lambda-\ell-1Nλ,ℓ≤−1,−+=λ−1\mathcal{N}_{\lambda,\ell\leq-1,-}^{+}=\lambda-1
sf<0s_f<0Nλ,ℓ≥0,−−=λ−ℓ−1\mathcal{N}_{\lambda,\ell\geq0,-}^{-}=\lambda-\ell-1Nλ,ℓ≤−1,−−=λ\mathcal{N}_{\lambda,\ell\leq-1,-}^{-}=\lambda

Table I. The Nλ,ℓ,sf±\mathcal{N}_{\lambda,\ell,s_f}^{\pm} numbers for different charges, spins and orbital numbers.

p~λ,ℓ,κμ=(Eλ,ℓ,κ(f),0,κsℓ2λ∣eQfB∣,pz),(41)\tilde{p}_{\lambda,\ell,\kappa}^{\mu}=\left(E_{\lambda,\ell,\kappa}^{(f)},0,\kappa s_\ell\sqrt{2\lambda|eQ_fB|},p_z\right), \tag*{(41)}

where Eλ,ℓ,κ(f)=κΩj+Eλ,ℓ,κ(f)E_{\lambda,\ell,\kappa}^{(f)}=\kappa\Omega j+E_{\lambda,\ell,\kappa}^{(f)}. Due to the p~λ,ℓ,κ2=σ2\tilde{p}_{\lambda,\ell,\kappa}^2=\sigma^2, we get Eλ(f)=pz2+σ2+2λ∣eQfB∣E_{\lambda}^{(f)}=\sqrt{p_z^2+\sigma^2+2\lambda|eQ_fB|}.

As long as we don’t impose boundary condition, the λ\lambdas are arbitrary numbers. We set a global boundary condition, namely the spectral boundary condition in which the net flux of each flavor going out of the boundary r=Rr=R is zero

∫−∞∞dz∫02πdϕ ψˉ(f)γrψ(f)∣r=R=0,(42)\int_{-\infty}^{\infty} dz\int_0^{2\pi}d\phi\,\bar{\psi}^{(f)}\gamma^r\psi^{(f)}\bigg|_{r=R}=0, \tag*{(42)}

with γr≡γ1cos⁡ϕ+γ2sin⁡ϕ\gamma^r\equiv\gamma^1\cos\phi+\gamma^2\sin\phi. Substituting the Ritus ansatz given in the Eqs. (III.33) along with Eq. (III.37) and integrating over the angle, one get the following constraints to fulfill the boundary condition of the Eq. (III.41)

for ℓ≥0,1F1(−Nλ,ℓ,sf+;∣ℓ∣+1;αQf)=0,\text{for }\ell\geq0,\qquad{}_1F_1(-\mathcal{N}_{\lambda,\ell,s_f}^{+};|\ell|+1;\alpha Q_f)=0,
for ℓ≤−1,1F1(−Nλ,ℓ,sf−;∣ℓ∣+1;αQf)=0.(43)\text{for }\ell\leq-1,\qquad{}_1F_1(-\mathcal{N}_{\lambda,\ell,s_f}^{-};|\ell|+1;\alpha Q_f)=0. \tag*{(43)}

If we are given the ℓ\ells for each spin direction, then the λ\lambdas will be derived numerically by solving these equations. Needless to say that changing the kind of boundary condition would lead to different values and therefore may alter the entire results. 9 We will show that implying the spectral boundary condition and setting the energies to be always positive, turn into a correlation between the strength of magnetic and rotation field. In next section, we will present numerical results derived from the Dirac equation along with the solutions of the gap equation at zero and finite temperature.

Numerical results

Having set up the fundamental relations for the effective action from a general view and giving some examples, and reviewing the Dirac equation, we proceed to investigate chiral symmetry breaking/restoration in a magnetized and rotating QM in presence of non-vanishing μ\mu and TT. To explore the phase diagram, we need to some prior information about the properties of QM at T=0T=0. These pieces of information contain knowledge about the values of coupling constant, roots of the Dirac equation (which are already discussed) as inputs to the gap equation and the relation between the magnetic and rotation fields. To this reason, we break this section into two parts. In the first subsection, we will study the gap equation at T=0T=0. These data will serve as inputs in the phase transition studies. In the second subsection, we are going to solve numerically the gap equation at T≠0T\ne0 and μ≠0\mu\ne0 and sketch the phase diagram as Tc(μ)T_c(\mu) and Tc(RΩ)T_c(R\Omega) plots for different cases. We are able to pinpoint exactly the position of CEP by the new technique which is solving the gap-equation and its derivative with respect to the order parameter ”σ\sigma”, simultaneously.

Results at T=0 and μ\mu=0

The λ\lambda numbers are the key ingredients. They play the role of initial data to solve the gap equation and represent the physical points consistent with the boundary condition. They are derived from the Eq. (III.42) for each flavor separately. In Fig. 1 we show the first root of each ℓ\ells for different α\alpha values (1, 5, 10, 30, 50, 100). The top plot indicates the up quark roots with Qf=2/3Q_f=2/3 and the bottom one shows the down quark roots with Qf=−1/3Q_f=-1/3.

There are some points worth to comment on here. We run our numerical codes with

Λ=500 MeV,R=6 fm,(44)\Lambda=500\,\mathrm{MeV},\qquad R=6\,\mathrm{fm}, \tag*{(44)}

where Λ\Lambda is the momentum cut-off and determines the validity region of NJL model. This cut-off puts a limit

on the energy of massless particles at rest

Λ=ϵλ(f)(pz=0,σ=0)=4R2λmax⁡α∣Qf∣,(45)\Lambda=\epsilon_{\lambda}^{(f)}(p_z=0,\sigma=0)=\sqrt{\frac{4}{R^2}\lambda_{\max}\alpha|Q_f|}, \tag*{(45)}

that in turn gives the maximum value of λ\lambda for each ℓ\ell. Also, this leads ”ℓ\ell” to be bounded from below and above ℓmin⁡≤ℓ≤ℓmax⁡\ell_{\min}\le\ell\le\ell_{\max}, and beyond these values there is no solution to λ\lambda compatible with the Eq. (IV.44). That is why the plots in Fig. 1 are shown in between ℓmin⁡≤ℓ≤ℓmax⁡\ell_{\min}\le\ell\le\ell_{\max} for each α\alpha. Values of (ℓmin⁡,ℓmax⁡)(\ell_{\min},\ell_{\max}) depend on α\alpha and quark’s charge in a way that for in small α\alpha they are almost symmetric ℓmin⁡∼−ℓmax⁡\ell_{\min}\sim-\ell_{\max}, but increasing α\alpha leads to pushing the ℓmin⁡→0\ell_{\min}\to0 and ℓmax⁡→∞\ell_{\max}\to\infty for positive quarks and ℓmax⁡→0\ell_{\max}\to0 and ℓmin⁡→−∞\ell_{\min}\to-\infty for negative quarks. This feature is due to the magnetic-orbit coupling, i.e. eL⃗⋅B⃗e\vec{L}\cdot\vec{B}. Also, values of λ\lambda decrease by increasing α\alpha. These are very obvious in the Fig. 1. It is worthwhile to mention that the strength of magnetic field is given by the definition eB=2α/R2eB=2\alpha/R^2 with RR set in the Eq. (IV.43)

∣eB∣=2αR2=0.11αmπ2.(46)|eB|=2\frac{\alpha}{R^2}=0.11\alpha m_\pi^2. \tag*{(46)}

In Figure 2, we show dimensionless energy levels for massless positively and negatively quarks at zero momentum.

Figure 2. color online. Top panel: energy derived from first root of positive quark for different α\alpha at RΩ=0R\Omega=0. Bottom panel: the same plot for down quark.

These energy levels are

REλ,ℓ,κ(f)(σ→0,pz→0)=−κjRΩ+4λα∣qf∣.(47)R E_{\lambda,\ell,\kappa}^{(f)}(\sigma\to0,p_z \to0)=-\kappa j R\Omega+\sqrt{4\lambda\alpha\lvert q_f\rvert}. \tag*{(47)}

The top panel of Figure 2 shows the energies for up quark derived from the first λ\lambda of the top panel of Figure 1 at RΩ=0R\Omega=0 and the bottom one shows the same quantity for down quark at RΩ=0R\Omega=0 derived from the first λ\lambda of the bottom plot of Figure 1. As it is evident from Figure 2, increasing the α\alpha leads to decreasing energies. In Figure 3, we show the RΩR\Omega dependence of energies for α=5\alpha=5 and α=10\alpha=10. The two top plots shows energies for α=5\alpha=5 which is derived from the first λ\lambda for (RΩ=0,0.5,1)(R\Omega=0,0.5,1). The upper (lower) part shows the energy levels for up (down) quarks. The two bottom plots show energy levels for α=10\alpha=10 with (RΩ=0,0.3,0.6)(R\Omega=0,0.3,0.6). We observe that for ℓ≥0\ell\geq0 increasing RΩR\Omega results in decreasing energies, while for ℓ≤−1\ell\leq-1 the opposite case has happened. Also, increasing the α\alpha decreases the energies because λ\lambda behaves in this kind.

Figure 1. color online. Top panel: first root of up quark for different α\alpha values (1, 5, 10, 30, 50, 100). Bottom panel: the same one for down quark.

Figure 3. color online. Two top panels: energy levels derived from first root for α=5\alpha=5 with (RΩ=0,0.5,1)(R\Omega=0,0.5,1). Upper part shows the up quark and lower part shows the down quark energy levels. Two bottom panel: the same ones for α=10\alpha=10 with (RΩ=0,0.3,0.6)(R\Omega=0,0.3,0.6).

One of the main results of our paper is to find a non-trivial correlation between the magnetic and rotation field. To have a consistent quantization scheme, we take the energies of the Eq. (IV.46) to be always non-negative for each flavor and κ=±1\kappa=\pm1. We find that by keeping α\alpha fixed and varying the RΩR\Omega from 0 to 1, it can be found a reasonable region 0≤RΩ≤RΩmax⁡0\leq R\Omega\leq R\Omega_{\max} in which

Eλ,ℓ,κ(f)≥0E_{\lambda,\ell,\kappa}^{(f)} \ge0. In the Figure 4 we sketch this region in terms of α\alpha. At small α\alpha all the physical region RΩMax=1R\Omega_{\mathrm{Max}} = 1 is accessible, but at the point αc=7\alpha_c = 7 this RΩMaxR\Omega_{\mathrm{Max}} starts to shrink. At very strong magnetic field, this accessible part reaches to zero and therefore strongly magnetized plasma can not rotate anymore. This result is a clear consequence of the spectral boundary condition.

Plot of $R\Omega_{\mathrm{Max}}$ versus $\alpha$

Figure 4. color online. RΩMaxR\Omega_{\mathrm{Max}} v.s. α\alpha. This states that strongly magnetized plasma can not rotate.

Now, we are going to investigate the gap equation at T=0T = 0. The effective action in a cold rotating and magnetized plasma is

VeffB,Ω=−σ22G+Nc∑f=u,d∑ℓ=ℓmin⁡ℓmax⁡∫dpz2πf(p,Λ,δΛ)Gλ,ℓ,sf(f)(X)ϵλ(f),(48)\mathcal{V}_{\mathrm{eff}}^{B,\Omega} = -\frac{\sigma^2}{2G} + N_c \sum_{f=u,d} \sum_{\ell=\ell_{\min}}^{\ell_{\max}} \int\frac{dp_z}{2\pi} f(p,\Lambda,\delta\Lambda) \mathcal{G}_{\lambda,\ell,s_f}^{(f)}(X)\epsilon_{\lambda}^{(f)}, \tag*{(48)}

with ϵλ(f)=pz2+σ2+4R2λα∣Qf∣\epsilon_{\lambda}^{(f)} = \sqrt{p_z^2+\sigma^2+\frac{4}{R^2}\lambda\alpha|Q_f|} and Gλ,ℓ,sf(X)\mathcal{G}_{\lambda,\ell,s_f}(X) is defined in Eq. (II.23). Note that Cλ,ℓ=Cλ,ℓ±C_{\lambda,\ell}=C_{\lambda,\ell}^{\pm} due to the boundary condition (III.42). To smooth out the pzp_z integration, we introduce the cut-off function as

f(p,Λ,δΛ)=sinh⁡(ΛδΛ)cosh⁡(pδΛ)+cosh⁡(ΛδΛ),(49)f(p,\Lambda,\delta\Lambda)=\frac{\sinh\left(\frac{\Lambda}{\delta\Lambda}\right)}{\cosh\left(\frac{p}{\delta\Lambda}\right)+\cosh\left(\frac{\Lambda}{\delta\Lambda}\right)}, \tag*{(49)}

and p≡pz2+4R2λα∣Qf∣p\equiv\sqrt{p_z^2+\frac{4}{R^2}\lambda\alpha|Q_f|}. For the limit δΛΛ→0\frac{\delta\Lambda}{\Lambda}\to0, the cut-off function tends to the step function

lim⁡δΛΛ→0f(p,Λ,δΛ)=Θ(Λ2−p2).(50)\lim_{\frac{\delta\Lambda}{\Lambda}\to0}f(p,\Lambda,\delta\Lambda)=\Theta(\Lambda^2-p^2). \tag*{(50)}

Inserting the step function into the gap equation (IV.47), makes easier the pzp_z integration and the gap equation ∂VeffB,Ω/∂σ=0\partial\mathcal{V}_{\mathrm{eff}}^{B,\Omega}/\partial\sigma=0 results in

σ(1−NcGπ∑f=u,d∑ℓ=ℓmin⁡ℓmax⁡Gλ,ℓ,sf(f)(X)tanh⁡−1(Λ2−4R2λα∣Qf∣Λ2+σ2)Θ(Λ2−4R2λα∣Qf∣))=0.(51)\sigma\left(1-\frac{N_cG}{\pi}\sum_{f=u,d}\sum_{\ell=\ell_{\min}}^{\ell_{\max}}\mathcal{G}_{\lambda,\ell,s_f}^{(f)}(X)\tanh^{-1}\left(\sqrt{\frac{\Lambda^2-\frac{4}{R^2}\lambda\alpha|Q_f|}{\Lambda^2+\sigma^2}}\right)\Theta\left(\Lambda^2-\frac{4}{R^2}\lambda\alpha|Q_f|\right)\right)=0. \tag*{(51)}

The trivial solution to this equation is σ=0\sigma= 0, but it does not lead to symmetry breaking. We find non-trivial solution and solve the latter equation numerically. In the Figure 5, the non-trivial solution to the gap equation (IV.50) is shown at α=7\alpha= 7 for different couplings. As we would expect, increasing the coupling causes to increasing the dynamical mass. In the Figure 6 it is seen the α\alpha dependence of dynamical mass at fixed coupling, GΛ2=5G\Lambda^2 = 5. Near the axis x=0x = 0, dynamical masses have no distinct α\alpha dependence and rotational magnetic inhibition is excluded [34]. This is because the vorticity vanishes at T=0T = 0 gap equations [21, 31]. The condensate falls off near the boundary region, since the condition ∂rσ≪σ2\partial_r\sigma\ll\sigma^2 violates there. In this region chiral

Dynamical mass generated at $\alpha=7$ for different couplings

Figure 5. color online. Dynamical mass generated at α=7\alpha= 7 for x∈[0,2.3]x \in[0,2.3] with different couplings (GΛ2=4,5,6G\Lambda^2 = 4,5,6) from bottom to top.

Figure 6. color online. Dynamical mass generated at coupling GΛ2=5G\Lambda^2 = 5 for different α=(3,6,9,12)\alpha= (3,6,9,12) in their corresponding interval xx.

condensate grows with α\alpha due to the mode accumulation in which the magnetic field is enhanced for larger angular momenta [33]. 10 In the Figure 7, we show another realization of α\alpha dependence. The gap equation is solved at x=0.1x = 0.1 for different couplings and α\alphas. The profile σ(α)\sigma(\alpha) looks oscillatory at these small magnetic fields around the axis and no definitive conclusion can be made about the magneto or inverse-magneto catalysis.

Dynamical mass generated versus $\alpha$ for different couplings

Figure 7. color online. Dynamical mass generated vs. α\alpha for couplings GΛ2=(4,5,6)G\Lambda^2 = (4,5,6) from bottom to top, at x=0.1x=0.1.

Critical coupling is very important quantity in the gap equation. By definition, it is the minimum coupling needed to provide a non-trivial solution σ≠0\sigma\ne0, for (Minkowski) temperature T0T_0 to a global temperature

the gap equation. In the Figure 8, this parameter is plotted against α\alpha. We derived it by searching the entire region 0≤x≤α0 \leq x \leq\alpha for each α\alpha and we call the GcΛ2G_c\Lambda^2 as the minimum value of coupling which gives a solution σ≠0\sigma\ne0 at the very first xx. Near the α→0\alpha\to0, it closes to GcΛ2∼2π2NcNfG_c\Lambda^2 \sim\frac{2\pi^2}{N_cN_f}, but increasing the magnetic field turns this coupling to smaller values and at strong magnetized plasma GcΛ2→0G_c\Lambda^2 \to0 which is a direct sign of the magneto-catalysis.

Critical coupling versus $\alpha$

Figure 8. color online. Critical coupling v.s. α\alpha. It is the minimum coupling needed to develop a non-trivial solution at the very first xx.

Results at T≠0T \ne0 and μ≠0\mu\ne0

In curved spaces, temperature has to be defined locally and thermal states are well-defined within the box that metric is flat. So, the notion of global temperature is ambiguous. Red shift equation relates the local

T(r)T(r) as follows [39, 40]

T(r)=T0g00(r),(52)T(r) = \frac{T_0}{\sqrt{g_{00}(r)}}, \tag*{(52)}

where T0T_0 is the temperature seen by a local rest frame observer when g00(r)→1\sqrt{g_{00}(r)} \to1, while T(r)T(r) is the temperature seen by an accelerated observer. We infer from Eq. (IV.51) that at very high gravity points, i.e. g00(r)→0g_{00}(r) \to0 accelerated observer sees a high temperature medium T(r)→∞T(r) \to\infty, while in low gravity points when g00(r)→∞g_{00}(r) \to\infty the medium seems to be cold, T(r)→0T(r) \to0. For rotating plasma according to the Eq. (III.24), T(r)→T0/1−r2Ω2T(r) \to T_0/\sqrt{1-r^2\Omega^2} and therefore, if rΩ→0r\Omega\to0 then T(r)→T0T(r) \to T_0. To this reason, we solve the gap equation at points close to the axis which is x=0.1x=0.1.

Effective action at T≠0T \ne0 for rotating and magnetized plasma with NfN_f flavors and NcN_c colors is given by the Eq. (II.23). Hence, the non-trivial solution to the gap equation comes from the non-solutions of the following relation

1+NcG∑f=u,d∑ℓ=ℓmin⁡ℓmax⁡∑λ∫dpz2π1Eλ(f)σλ,ℓ(f)(x)(fFD(Eλ,ℓ,−(f)+μ)+fFD(Eλ,ℓ,+(f)−μ)−1)=0,(53)1+N_cG\sum_{f=u,d}\sum_{\ell=\ell_{\min}}^{\ell_{\max}}\sum_{\lambda}\int\frac{dp_z}{2\pi}\frac{1}{E_{\lambda}^{(f)}}\sigma_{\lambda,\ell}^{(f)}(x)\left(f_{FD}\left(E_{\lambda,\ell,-}^{(f)}+\mu\right)+f_{FD}\left(E_{\lambda,\ell,+}^{(f)}-\mu\right)-1\right)=0, \tag*{(53)}

where the Fermi-Dirac distribution is

fFD(ε)=11+eβε.(54)f_{FD}(\varepsilon)=\frac{1}{1+e^{\beta\varepsilon}}. \tag*{(54)}

We solve the gap equation (IV.52) for the chiral field σ\sigma by keeping fixed (α,RΩ,T,μ,G,x)(\alpha,R\Omega,T,\mu,G,x) and results are shown in Figure 9. All the plots of this subsection are sketched at GΛ2=5G\Lambda^2=5 and x=0.1x=0.1. The two top plots display non-trivial solution for μ=0\mu=0 and μ=200MeV\mu=200\mathrm{MeV} at RΩ=0R\Omega=0 from up to down and the two bottom plots show similar solution for μ=0\mu=0 and μ=200MeV\mu=200\mathrm{MeV} at RΩ=0.4R\Omega=0.4. In each plot different colors correspond to different (α=1,3,5,7,10)(\alpha=1,3,5,7,10). Increasing temperature leads to decreasing the mass and at a critical point σ(Tc)→0\sigma(T_c)\to0. Value of this critical point depends on (μ,α,RΩ)(\mu,\alpha,R\Omega) and we will discuss it later. A similar pattern of the σ(α)\sigma(\alpha) exist for the rest diagrams of the Figure 9. It seems that σ(α)\sigma(\alpha) is a decreasing function for 1≤α≤71\leq\alpha\leq7 and elsewhere it increases.

Figure 9. color online. Dynamical mass σ(T)\sigma(T) for different α\alpha at different situations. Two top plots indicate non-zero mass solution for μ=0\mu= 0 and μ=200\mu= 200MeV at RΩ=0R\Omega= 0. Right plots show the similar plot for μ=200\mu= 200 at (RΩ=0,0.2,0.4)(R\Omega= 0, 0.2, 0.4) from top to bottom. In each plot different colors stand for corresponding α\alpha. The interval 0≤T≤500 \leq T \leq50 is magnified in the middle plots.

In the Figure 10 we show σ(T)\sigma(T) for α=7\alpha=7 at μ=0\mu=0 and μ=200MeV\mu=200\mathrm{MeV}. Different colors stand for different vorticities (RΩ=0,0.2,0.4,0.6,0.8,1)(R\Omega=0,0.2,0.4,0.6,0.8,1). The inside plots show magnified the interval 0≤T≤500\leq T\leq50. Looking very carefully, we observe that increasing the rotations causes to decreasing the mass and it is a direct sign of inverse-rotational catalysis. We see this observation to be valid for 1≤α≤101\leq\alpha\leq10. In the Figure 11, we plot σ(T)\sigma(T) for α=7\alpha=7 and α=10\alpha=10 which magenta(red) color shows the μ=0(μ=200)\mu=0(\mu=200) solutions. Top plots indicate solution for α=7\alpha=7 and bottom plots display the dynamical mass for α=10\alpha=10. Likewise, the left part belongs to RΩ=0.4R\Omega=0.4 and the right part belongs to RΩ=0.6R\Omega=0.6. As we would expect, increasing the chemical potential decreases the dynamical mass that is a common feature in all NJL model calculations of phase diagrams.

Figure 10. color online. Dynamical mass σ(T)\sigma(T) T for different vorticities (RΩ=0,0.2,0.4,0.6,0.8,1)(R\Omega= 0, 0.2, 0.4, 0.6, 0.8, 1) shown with their corresponding colors. The inside plots show magnified the interval 0≤T≤500 \leq T \leq50.

Figure 11. The dynamical mass σ(T)\sigma(T) for different RΩR\Omega and μ\mu. Two top plots indicate the α=7\alpha=7 at (RΩ=0.4,0.6)(R\Omega=0.4,0.6) at μ=0\mu=0 and μ=200MeV\mu=200\mathrm{MeV}. The two bottom ones show similar plots for α=10\alpha=10 at (RΩ=0.4,0.6)(R\Omega=0.4,0.6) at μ=0\mu=0 and μ=200MeV\mu=200\mathrm{MeV}.

By definition, the critical point is a point of which σ(Tc)→0\sigma(T_c)\to0, i.e. the chiral symmetry is restored. However, the kind of phase transition is important. There are many ways to describe the order of phase transition. The regular way is to look for the derivatives of effective action with respect to the order parameter. If nnth derivative of effective action would have singularity, it is the so called nnth-order phase transition. Another way is to explore the behavior of dynamical mass. If phase transition is of second order, the trivial solution is the only one to the gap equation (IV.52) at T≥TcT\geq T_c and chiral symmetry is restored. If the kind of phase transition is of first order, at critical point there is at least two solutions to the gap equation (IV.52) besides the trivial one. In this kind, all the solutions have identical effective action and the system lives in a mixed state [41]. Also, we can identify the first-order transition with a jump in the dynamical mass profile [42].

In the NJL and dense effective models at low densities we get second order phase transition, while at high densities we have first-order ones. The joining point of these two section is CEP and it defines the changing of phase transition pattern. In the current paper, we are able to derive the location of CEP analytically and its derivation works as follows. As we already discussed, at critical point of first-order there are many solutions to the gap equation and they are equivalent in the effective action. You may assume that σa=(σ1,σ2,…,σN)\sigma_a=(\sigma_1,\sigma_2,\ldots,\sigma_N) is the set of ground state and its members are the solutions of the gap equation ∂Veff/∂σi=0\partial V_{\mathrm{eff}}/\partial\sigma_i=0. Expansion of the effective action around each member is

Veff(σi+δσi)=Veff(σi)+δσi22∂2Veff∂σ2∣σ=σi+⋯V_{\mathrm{eff}}(\sigma_i+\delta\sigma_i)=V_{\mathrm{eff}}(\sigma_i)+\frac{\delta\sigma_i^2}{2}\left.\frac{\partial^2V_{\mathrm{eff}}}{\partial\sigma^2}\right|_{\sigma=\sigma_i}+\cdots

Since all the σi\sigma_i have similar effective action, solving ∂2Veff/∂σ2=0\partial^2V_{\mathrm{eff}}/\partial\sigma^2=0 in addition to the ∂Veff/∂σi=0\partial V_{\mathrm{eff}}/\partial\sigma_i=0 can locate the position of CEP. Basically, the CEP is the first point in phase diagram which meet this cut. Here, we give an example to show how this mechanism works. In the normal NJL model with the effective action given in the Eq. (II.15) these two equations are given by

∂Veff∂σ=σ+NcNfGπ2∫0Λp2dpσϵ0(fFD(ϵ0+μ)+fFD(ϵ0−μ)−1)=0,\frac{\partial V_{\mathrm{eff}}}{\partial\sigma}=\sigma+\frac{N_cN_fG}{\pi^2}\int_0^\Lambda p^2dp\frac{\sigma}{\epsilon_0}\left(f_{FD}(\epsilon_0+\mu)+f_{FD}(\epsilon_0-\mu)-1\right)=0,
∂2Veff∂σ2=1+NcNfGπ2∫0Λp2dp∂∂σ(σϵ0(fFD(ϵ0+μ)+fFD(ϵ0−μ)−1))=0.(55)\frac{\partial^2V_{\mathrm{eff}}}{\partial\sigma^2}=1+\frac{N_cN_fG}{\pi^2}\int_0^\Lambda p^2dp\frac{\partial}{\partial\sigma}\left(\frac{\sigma}{\epsilon_0}\left(f_{FD}(\epsilon_0+\mu)+f_{FD}(\epsilon_0-\mu)-1\right)\right)=0. \tag*{(55)}

The latter equations are identical to the previous methods to find the CEP [43]. We are able to reproduce the normal NJL phase diagram by solving this method. Indeed, the second derivative give the susceptibility of the QM and its vanishing states the divergence of susceptibility around the first-order points.

In Figure 12 we sketch Tc(μ)T_c(\mu) for different α\alpha and RΩR\Omega. Solid(dashed) lines define the first(second) order phase transitions. Their joining, which is the CEP, is marked with a star symbol. Top-left plot represents α=1\alpha= 1 at different (RΩ=0,0.4,0.6,0.8,1)(R\Omega= 0, 0.4, 0.6, 0.8, 1), top-right plot shows α=5\alpha= 5 at different (RΩ=0,0.4,0.6,0.8,1)(R\Omega= 0, 0.4, 0.6, 0.8, 1), bottom-left plot indicates α=7\alpha= 7 at different (RΩ=0,0.4,0.6,0.8,1)(R\Omega= 0, 0.4, 0.6, 0.8, 1) and bottom-right plot shows α=10\alpha= 10 at different (RΩ=0,0.4,0.6)(R\Omega= 0, 0.4, 0.6). We should study the phase diagrams by considering the correlation between magnetic and rotation fields shown in the Figure 4. That is why we show α=10\alpha= 10 up to RΩ=0.6R\Omega= 0.6, because in this magnetic field strength RΩmax⁡=0.68R\Omega_{\max} = 0.68. To illuminate our results, we fit precisely the transition points of the Figure 12 to the following function

Figure 12. color online. Phase diagram Tc(μ)T_c(\mu) in different cases. Solid(dashed) lines represent the first(second) order phase transitions. The very top is for α=1\alpha= 1, the next top is for α=5\alpha= 5, the third- top plot is for α=7\alpha= 7 and the very bottom is for α=10\alpha= 10. In each plot, the RΩR\Omega dependence is shown by different colors. For α=10\alpha= 10, plots are sketched until RΩ=0.6R\Omega= 0.6 because RΩMax=0.68R\Omega_{\mathrm{Max}} = 0.68.

Tc(μ)=Tc0(1−κ2(μTc0)2−κ4(μTc0)4).(56)T_c(\mu) = T_c^0 \left(1 - \kappa_2 \left(\frac{\mu}{T_c^0}\right)^2 - \kappa_4 \left(\frac{\mu}{T_c^0}\right)^4\right). \tag*{(56)}

The parameters (Tc0,κ2,κ4)(T_c^0,\kappa_2,\kappa_4) are given in the Appendix. A for each α\alpha and RΩR\Omega. Although, our model is very different from the Lattice models, they are comparable and could hint us to see how the QM behaves under the rotation and magnetic field and what would we expect from other numerical methods [35]. Compared to the normal Lattice models (no rotation or magnetic fields), we find that the curvature of the phase diagram κ2\kappa_2 is enhanced. The κ2\kappa_2 decreases by increasing the RΩR\Omega and α\alpha.

We infer the inverse-rotational catalysis of these plots because increasing the RΩR\Omega causes decreasing Tc(μ)T_c(\mu). This property is very obvious in Figure 13. In this figure, we demonstrate TCEP(RΩ)T_{\mathrm{CEP}}(R\Omega) for different α\alpha. As it is shown, the magnetic field has a complex impact on the phase diagram. We see inverse-magneto catalysis until a point RΩ∼0.39R\Omega\sim0.39, which means that increasing α\alpha leads to decreasing TcT_c, while for RΩ>0.39R\Omega>0.39 we see the magneto-catalysis. This pattern is valid up to α=7\alpha=7 and for larger values of magnetic field it seems magneto-catalysis to happen every where. Generally, magneto-catalysis is the dominant picture at large RΩR\Omega. We can also express these results from the numbers given in the table II to table V of the App. A. In the Figure 14, we demonstrate Tc(RΩ)T_c(R\Omega) at α=7\alpha=7 for different chemical potentials. Types of the lines and stars are similar to the Figure 12. For μ<μ1=244.932\mu<\mu_1=244.932 phase transition is second order every where in 0≤RΩ≤10\leq R\Omega\leq1, but at the point μ1\mu_1 phase diagram starts to get some first-order transition points. Increasing the chemical potential covers more points in the Tc(RΩ)T_c(R\Omega) plot to become first order and finally in the point μ0=248.012\mu_0=248.012 all the points in 0≤RΩ≤10\leq R\Omega\leq1 possess first-order transition. For μ≥μ0\mu\geq\mu_0 the phase transition is always first order in the RΩR\Omega direction. This kind of interplay between the chemical potential and RΩR\Omega in the Tc(RΩ)T_c(R\Omega) plots, is observed everywhere in 0≤α≤100\leq\alpha\leq10. Likewise, the point that faster plasma changes sooner to become first-order, is definable within the inverse-rotational catalysis scenario. The Figure 15 shows α\alpha dependence in Tc(RΩ)T_c(R\Omega) plots. Top-right, top-left and bottom plots indicate (RΩ=0,0.4,0.6)(R\Omega=0,0.4,0.6), respectively and different colors in each plot belong to (α=1,5,7,10)(\alpha=1,5,7,10) as it is shown. The inside plots show magnified points of plots in 0≤μ≤500\leq\mu\leq50. We observe that for 0≤RΩ≤0.390\leq R\Omega\leq0.39 the dominant picture is inverse-magneto catalysis, while in faster plasma RΩ≥0.4R\Omega\geq0.4 the magneto-catalysis scenario is seen. This picture is valid up to αc=7\alpha_c=7. For larger magnetic field we see the magneto-catalysis every where. Results of this figure are in complete agreement with the Figure 13.

Figure 13. color online. Temperature of CEP as a function of RΩR\Omega for different α\alpha. Dashed line shows the RΩMaxR\Omega_{\mathrm{Max}} of α=10\alpha= 10.

Figure 14. color online. Phase diagram Tc(RΩ)T_c(R\Omega) for α=7\alpha= 7 at different μ\mu. Solid(dashed) lines illustrate first(second) order phase transitions. Star marks indicate the location of CEP at each μ\mu.

Figure 15. color online. Phase diagram Tc(μ)T_c(\mu) for different RΩR\Omega. Top, middle and bottom plots show the phase diagram for (RΩ=0,0.4,0.6R\Omega= 0, 0.4, 0.6), respectively. In each plot different colors belong to (α=1,5,7,10\alpha= 1, 5, 7, 10) shown by different colors. The inside plots magnify the points in interval 0≤μ≤500 \leq\mu\leq50.

Conclusion

Studying the phase diagrams of QM in presence of the rotation and magnetic field is of great importance in today’s high-energy research. Either of the magnetic or rotation fields has its special non-trivial effects on the evolution of QGP and they can reveal many facts about the behavior of QGP under these extreme circumstances. In this work, we examine the chiral symmetry restoration/breaking of a QM with Nf=2N_f = 2 and Nc=3N_c = 3 within the effective NJL model accompanied by constant magnetic and rotation fields. Utilizing the Ritus method has enabled us to present a master formula for the in-medium effective action of “σ\sigma” field in curved spaces. In this regard, we gave some examples to clarify our framework and showed how it works in different situations. Likewise, we used the Ritus method to solve the Dirac equation in the constantly rotating and magnetized plasma. We identified the energy eigenvalues along with the generalized Landau levels and to have a good quantization pattern, we take the positive energies. We imposed the spectral boundary condition and the profile of Landau levels as well as energy levels are sketched in terms of orbital number ℓ\ell for different α\alpha. The very important result is the correlation between the magnetic and rotation fields and we concluded that strongly magnetized plasma is not able to rotate. The critical coupling needed to provide a non-trivial solution to the gap equation decreases by increasing α\alpha and it closes to the normal NJL value GcΛ2=2π2NcNfG_c\Lambda^2 = \frac{2\pi^2}{N_cN_f} at small α\alpha.

Solutions of the gap equation at zero and finite temperature express many facts. At zero temperature, some features are in common with the normal NJL calculations. Since vorticity disappears at T=0T = 0 near the axis we didn’t see rotational-magnetic inhibition. However, near the boundary the chiral condensate grows with α\alpha because of the mode accumulation and this confirms the surface-magnetic catalysis. At finite temperature, the interplay between the chemical potential, magnetic and rotation fields makes the phase diagrams to be more complicated. TcT_c always decreases by increasing the RΩR\Omega which is nothing but the inverse-rotational catalysis. But Tc(α)T_c(\alpha) is more complex in which for less rotating plasma (0≤RΩ≤0.380 \leq R\Omega\leq0.38) TcT_c decreases by increasing α\alpha

-the inverse-magneto catalysis-, while in fast rotating plasma (0.38<RΩ≤10.38 < R\Omega\leq1) TcT_c increases by increasing the α\alpha -the magneto catalysis-. Points of the phase diagrams are fitted to the polynomial function and we discover that the magnetic and rotation fields try to increase the curvature of the phase diagrams. This curvature decreases by increasing α\alpha and RΩR\Omega. The position of CEP is found exactly by solving the gap equation and its derivatives, simultaneously. The TCEPT_{\mathrm{CEP}} always decrease by increasing the RΩR\Omega and at each fixed μ\mu, the dTCEPdRΩ\frac{dT_{\mathrm{CEP}}}{dR\Omega} is larger for smaller α\alpha.

Following the current paper, we can extend it to more cases. We study the chiral symmetry dynamics, but an interesting topic is confinement/deconfinement phase transition in presence of constant magnetic and rotation fields. Also, the results of this work can be more investigated within the other effective models such as quark-hadron models to see whether they change. Including the strange quarks into the effective models is also very interesting. Spectrum of mesons in the constant rotation and magnetic field deserves an independent study.

Acknowledgement

We would like to thank M. Asadi for reading carefully the manuscript. We kindly appreciate M. N. Chernodub for editing the primary draft, giving valuable comments and having fruitful discussions. We also thank Xu-Guang Huang and Igor Shovkovy for their precious discussions and clarification of some issues.

A Parameters of fit

The parameters of the fit equation (IV.55) are shown below for each α\alpha and RΩR\Omega

RΩR\OmegaTc0T_c^0κ2\kappa_2κ4\kappa_4
0197.9850.1740.074
0.4196.3650.1690.08
0.6194.3980.1660.083
0.8191.590.1610.089
1187.9350.1550.096

Table II. Parameters of fit for α=1\alpha= 1.

RΩR\OmegaTc0T_c^0κ2\kappa_2κ4\kappa_4
0194.4450.1060.122
0.4194.0550.1050.124
0.6193.5240.1010.127
0.8192.9290.10.128
1192.1940.10.13

Table III. Parameters of fit for α=5\alpha= 5.

RΩR\OmegaTc0T_c^0κ2\kappa_2κ4\kappa_4
0193.9170.1060.122
0.4193.5980.1040.124
0.6193.1890.10.127
0.8192.8170.1020.126
1192.3960.1070.125

Table IV. Parameters of fit for α=7\alpha= 7.

RΩR\OmegaTc0T_c^0κ2\kappa_2κ4\kappa_4
0194.1990.1060.123
0.4193.9310.1040.125
0.6193.6320.1020.127

Table V. Parameters of fit for α=10\alpha= 10.

References

References

  1. [1]E. V. Shuryak, “What RHIC experiments and theory tell us about properties of quark-gluon plasma?,” Nucl. Phys. A 750 (2005) 64–83, hep-ph/0405066.DOI
  2. [2]ALICE Collaboration, B. Abelev et al., “Long-range angular correlations on the near and away side in p–Pb collisions at √sNN = 5.02 TeV,” Phys. Lett. B 719 (2013) 29–41, 1212.2001.DOI
  3. [3]STAR Collaboration, J. Adams et al., “Experimental and theoretical challenges in the search for the quark gluon plasma: The STAR Collaboration’s critical assessment of the evidence from RHIC collisions,” Nucl. Phys. A 757 (2005) 102–183, nucl-ex/0501009.
  4. [4]M. G. Alford, K. Rajagopal, and F. Wilczek, “QCD at finite baryon density: Nucleon droplets and color superconductivity,” Phys. Lett. B 422 (1998) 247–256, hep-ph/9711395.
  5. [5]K. Fukushima and T. Hatsuda, “The phase diagram of dense QCD,” Rept. Prog. Phys. 74 (2011) 014001, 1005.4814.DOI
  6. [6]V. Skokov, A. Y. Illarionov, and V. Toneev, “Estimate of the magnetic field strength in heavy-ion collisions,” Int. J. Mod. Phys. A 24 (2009) 5925–5932, 0907.1396.DOI
  7. [7]STAR Collaboration, L. Adamczyk et al., “Global Λ hyperon polarization in nuclear collisions: evidence for the most vortical fluid,” Nature 548 (2017) 62–65, 1701.06657.DOI
  8. [8]V. A. Miransky and I. A. Shovkovy, “Quantum field theory in a magnetic field: From quantum chromodynamics to graphene and Dirac semimetals,” Phys. Rept. 576 (2015) 1–209, 1503.00732.DOI
  9. [9]A. J. Mizher, M. N. Chernodub, and E. S. Fraga, “Phase diagram of hot QCD in an external magnetic field: possible splitting of deconfinement and chiral transitions,” Phys. Rev. D 82 (2010) 105016, 1004.2712.DOI
  10. [10]K. Fukushima and J. M. Pawlowski, “Magnetic catalysis in hot and dense quark matter and quantum fluctuations,” Phys. Rev. D 86 (2012) 076013, 1203.4330.DOI
  11. [11]C. V. Johnson and A. Kundu, “External Fields and Chiral Symmetry Breaking in the Sakai-Sugimoto Model,” JHEP 12 (2008) 053, 0803.0038.DOI
  12. [12]N. Evans, T. Kalaydzhyan, K.-y. Kim, and I. Kirsch, “Non-equilibrium physics at a holographic chiral phase transition,” JHEP 01 (2011) 050, 1011.2519.DOI
  13. [13]E. S. Fraga and A. J. Mizher, “Can a strong magnetic background modify the nature of the chiral transition in QCD?,” Nucl. Phys. A 820 (2009) 103C–106C, 0810.3693.DOI
  14. [14]G. S. Bali, F. Bruckmann, G. Endrodi, Z. Fodor, S. D. Katz, S. Krieg, A. Schaefer, and K. K. Szabo, “The QCD phase diagram for external magnetic fields,” JHEP 02 (2012) 044, 1111.4956.DOI
  15. [15]G. Endrodi, “Critical point in the QCD phase diagram for extremely strong background magnetic fields,” JHEP 07 (2015) 173, 1504.08280.arxiv.org/abs/1504.08280
  16. [16]F. Bruckmann, G. Endrodi, and T. G. Kovacs, “Inverse magnetic catalysis and the Polyakov loop,” JHEP 04 (2013) 112, 1303.3972.DOI
  17. [17]F. Preis, A. Rebhan, and A. Schmitt, “Inverse magnetic catalysis in dense holographic matter,” JHEP 03 (2011) 033, 1012.4785.DOI
  18. [18]N. O. Agasian and S. M. Fedorov, “Quark-hadron phase transition in a magnetic field,” Phys. Lett. B 663 (2008) 445–449, 0803.3156.DOI
  19. [19]K. Fukushima, D. E. Kharzeev, and H. J. Warringa, “The Chiral Magnetic Effect,” Phys. Rev. D 78 (2008) 074033, 0808.3382.DOI
  20. [20]D. E. Kharzeev, L. D. McLerran, and H. J. Warringa, “The Effects of topological charge change in heavy ion collisions: “Event by event P and CP violation”,” Nucl. Phys. A 803 (2008) 227–253, 0711.0950.DOI
  21. [21]M. N. Chernodub and S. Gongyo, “Interacting fermions in rotation: chiral symmetry restoration, moment of inertia and thermodynamics,” JHEP 01 (2017) 136, 1611.02598.DOI
  22. [22]M. N. Chernodub and S. Gongyo, “Effects of rotation and boundaries on chiral symmetry breaking of relativistic fermions,” Phys. Rev. D 95 (2017) no. 9, 096006, 1702.08266.DOI
  23. [23]X. Wang, M. Wei, Z. Li, and M. Huang, “Quark matter under rotation in the NJL model with vector interaction,” Phys. Rev. D 99 (2019), no. 1, 016018, 1808.01931.DOI
  24. [24]Z. Zhang, C. Shi, X. Luo, and H.-S. Zong, “Chiral phase transition in a rotating sphere,” Phys. Rev. D 101 (2020), no. 7, 074036, 2003.03765.DOI
  25. [25]V. V. Braguta, A. Y. Kotov, D. D. Kuzne­d­elev, and A. A. Roenko, “Influence of relativistic rotation on the confinement-deconfinement transition in gluodynamics,” Phys. Rev. D 103 (2021), no. 9, 094515, 2102.05084.DOI
  26. [26]M. N. Chernodub and V. E. Ambrus, “Phase diagram of helically imbalanced QCD matter,” Phys. Rev. D 103 (2021), no. 9, 094015, 2005.03575.DOI
  27. [27]Y. Liu and I. Zahed, “Pion Condensation by Rotation in a Magnetic field,” Phys. Rev. Lett. 120 (2018), no. 3, 032001, 1711.08354.DOI
  28. [28]Y. Liu and I. Zahed, “Rotating Dirac fermions in a magnetic field in 1+2 and 1+3 dimensions,” Phys. Rev. D 98 (2018), no. 1, 014017, 1710.02895.DOI
  29. [29]H.-L. Chen, X.-G. Huang, and K. Mameda, “Do charged pions condense in a magnetic field with rotation?,” 1910.02700.DOI
  30. [30]H.-L. Chen, X.-G. Huang, and J. Liao, “QCD phase structure under rotation,” Lect. Notes Phys. 987 (2021) 349–379, 2108.00586.DOI
  31. [31]N. Sadooghi, S. M. A. Tabatabaee Mehr, and F. Taghinavaz, “Inverse magnetorotational catalysis and the phase diagram of a rotating hot and magnetized quark matter,” Phys. Rev. D 104 (2021), no. 11, 116022, 2108.12760.DOI
  32. [32]V. I. Ritus, “Radiative corrections in quantum electrodynamics with intense field and their analytical properties,” Annals Phys. 69 (1972) 555–582.DOI
  33. [33]H.-L. Chen, K. Fukushima, X.-G. Huang, and K. Mameda, “Surface Magnetic Catalysis,” Phys. Rev. D 96 (2017), no. 5, 054032, 1707.09130.DOI
  34. [34]H.-L. Chen, K. Fukushima, X.-G. Huang, and K. Mameda, “Analogy between rotation and density for Dirac fermions in a magnetic field,” Phys. Rev. D 93 (2016), no. 10, 104052, 1512.08974.DOI
  35. [35]H.-T. Ding, F. Karsch, and S. Mukherjee, “Thermodynamics of strong-interaction matter from Lattice QCD,” Int. J. Mod. Phys. E 24 (2015), no. 10, 1530007, 1504.05274.
  36. [36]S. M. A. Tabatabaee and N. Sadooghi, “Wigner function formalism and the evolution of thermodynamic quantities in an expanding magnetized plasma,” Phys. Rev. D 101 (2020), no. 7, 076022, 2003.01686.DOI
  37. [37]N. Sadooghi and F. Taghinavaz, “Dilepton production rate in a hot and magnetized quark-gluon plasma,” Annals Phys. 376 (2017) 218–253, 1601.04887.DOI
  38. [38]J. I. Kapusta and C. Gale, Finite-temperature field theory: Principles and applications. Cambridge Monographs on Mathematical Physics. Cambridge University Press, 2011.
  39. [39]J. Santiago and M. Visser, “Tolman temperature gradient in a gravitational field,” Eur. J. Phys. 40 (2019), no. 2, 025604, 1803.04106.DOI
  40. [40]R. Tolman and P. Ehrenfest, “Temperature Equilibrium in a Static Gravitational Field,” Phys. Rev. 36 (1930), no. 12, 1791–1798.DOI
  41. [41]L. D. McLerran, “Lecture on Quarkyonic Effective Field Theory,” Acta Phys. Polon. B 52 (2021), no. 3, 229–241.DOI
  42. [42]M. Asakawa and K. Yazaki, “Chiral Restoration at Finite Density and Temperature,” Nucl. Phys. A 504 (1989) 668–684.DOI
  43. [43]J. R. Morones-Ibarra, A. Enriquez-Perez-Gavilan, A. I. H. Rodriguez, F. V. Flores-Baez, N. B. Mata-Carrizalez, and E. V. Ordoñez, “Chiral Symmetry restoration and the critical end point in QCD,” Open Phys. 15 (2017), no. 1, 1039–1044.DOI

Paper details

vlnc:W_gz2yzsp9
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 10

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents