Isospin Effects on the Mixed Hadron-Quark Phase at High Baryon Density M. Di Toroa,b B. Liuc,d V. Grecoa,b V. Barane M. Colonnaa S. Plumarib,f 0 1 0 aLaboratori Nazionali del Sud INFN, I-95123 Catania, Italy 2 bPhysics and Astronomy Dept., University of Catania r p cIHEP, Chinese Academy of Sciences, Beijing, China A dTheoretical Physics Center for Scientific Facilities, Chinese Academy of Sciences, 9 100049 Beijing, China 2 ePhysics Faculty, Univ. of Bucharest and NIPNE-HH, Romania ] h fINFN, Sezione di Catania, Italy t - e-mail: [email protected] l c u n [ Abstract 2 v 7 The phase transition of hadronic to quark matter at high baryon and isospin 4 density is analyzed. Nonlinear relativistic mean field models are used to describe 2 hadronic matter, and the MIT bag model is adopted for quark matter. The bound- 3 . aries of the mixed phaseand therelated critical points for symmetricand asymmet- 9 0 ric matter are obtained. Isospin effects appear to be rather significant. The binodal 9 transitionlineofthe(T,ρ )diagram islowered toaregionaccessible throughheavy B 0 ioncollisionsintheenergyrangeofthenewplannedfacilities,e.g.theFAIR/NICA : v projects. Some observable effects are suggested, in particular an Isospin Distillation i X mechanism. The dependence of the results on a suitable treatment of isospin con- r tributions in effective QCD Lagrangian approaches, at the level of explicit isovector a parts and/or quark condensates, is critically discussed. Key words: Nuclear Matter at High Baryon Density; Symmetry Energy; Deconfinement Transition; Critical End Point; Effective QCD Lagrangians PACS numbers: 05.70.Fh, 05.70.Ce, 21.65.+f, 12.38Mh 1 Introduction Several suggestions are already present about the possibility of interesting isospin effects on the transition to a mixed hadron-quark phase at high baryon Preprint submitted to Elsevier 30 April 2010 600 500 a H=1.0 a =0.0 NLr : a =0.0, 1.0 H H V) 400 MIT : a Q=0.0, 1.0 e M a =1.0 300 Q a =0.0 A ( X Q / E 200 O 100 0 0 1 2 3 4 5 6 7 8 9 10 r /r 0 Fig. 1. Zero temperature EoS of Symmetric/Neutron Matter: Hadron (NLρ), solid lines, vs. Quark (MIT-Bag), dashed lines. α represent the isospin asymmetry H,Q parameters respectively of the hadron,quark matter: α = 0, Symmetric Matter; H,Q α = 1, Neutron Matter. H,Q density [1,2,3]. This seems to be a very appealing physics program for the new facilities, FAIR at GSI-Darmstadt [4] and NICA at JINR-Dubna [5], where heavy ion beams (even unstable, with large isospin asymmetry) will be available with good intensities in the 1-30 AGeV energy region. The weak point of those predictions is the lack of a reliable equation of state that can describe with the same confidence the two phases, hadronic and deconfined. On the other hand this also represents a strong theory motivation to work on more refined effective theories for a strong interacting matter. The main qualitative argument in favor of noticeable isospin effects on the hadron-quark transition at high density can be derived from the Fig.1, where we compare typical Equations of State (EoS) for Hadron (Nucleon) andQuark Matter, at zero temperature, for symmetric (α ≡ (ρ −ρ )/ρ ≡ −ρ /ρ = n p B 3 B 0.0)andneutronmatter(α = 1.0),whereρ aretheneutron/protondensities n,p and ρ = ρ +ρ the total baryon density. B n p For the Hadron part we use a Relativistic Mean Field (RMF) EoS ([6,7,8]) with non-linear terms and effective ρ−meson coupling for the isovector part, largely used to study isospin effects in relativistic heavy ion collisions [3,8]. In order to keep a smooth flow of the physics points in the discussion, details about the adopted effective nucleon-meson Lagrangians are presented in the Appendix A. 2 The energy density and the pressure for the quark phase are given by the MIT Bag model [9] (two-flavor case) and read, respectively: d3k ǫ = 3×2 k2 +m2(f +f¯)+B , (1) qX=u,dZ (2π)3q q q q 3×2 d3k k2 ¯ P = (f +f )−B , (2) 3 qX=u,dZ (2π)3 k2 +m2q q q q where B denotes the bag constant (the bag pressure), taken as a rather stan- dard value from the hadron spectra (B = 85.7 MeV fm−3, no density de- pendence), m are the quark masses (m = m = 5.5 MeV choice), and f , q u d q ¯ f represent the Fermi distribution functions for quarks and anti-quarks. The q quark number density is given by d3k + ¯ ρ =< q q >= 3×2 (f −f ) , i = u,d. (3) i i i Z (2π)3 i i The transition to the more repulsive quark matter will appear around the crossing pointsofthetwo EoS. Weseethatsuch crossing forsymmetric matter (α = α = 0.0) is located at rather high density, ρ ≃ 7ρ , while for pure H Q B 0 neutron matter (α = α = 1.0) it is moving down to about three times ρ . H Q 0 Of course the Fig.1 represents just a simple energetic argument to support the hadron-quark transition to occur at lower baryon densities for more isospin asymmetric matter. In the rest of the paper we will rigourously consider the case of a first order phase transition in the Gibbs frame for a system with two conserved charges (baryon and isospin), in order to derive more detailed results. Since the first order phase transition presents a jump in the energy, we can expect the mixed phase to start at densities even before the crossing pointsoftheFig.1.The lower boundarythencanbepredicted at relatively low baryon densities for asymmetric matter, likely reached in relativistic heavy ion collisions. Moreover this point is certainly of interest for the structure of the crust and the inner core of neutron stars, e.g. see refs. [10,11] and the review [12]. The ref.[10] is particularly interesting since similar results are obtained with rather different hadronic approaches, the RMF and the non-relativistic Brueckner-Hartree-Fock (BHF) theory. We finally note that the above conclusions are rather independent on the isoscalar part of the used Hadron EoS at high density, that is chosen to be rather soft in agreement with collective flow and kaon production data [13,14]. 3 In the used Bag Model no gluon interactions, the α -strong coupling parame- s ter, are included. We remark that this in fact would enhance the above effect, since it represents an attractive correction for a fixed B-constant, see [15]. A reduction of the Bag-constant with increasing baryon density, as suggested by various models, see ref.[10], will also go in the direction of an “earlier” (lower density) transition, as already seen in ref.[2]. At variance, the presence of ex- plicit isovector contributions in the quark phase could play an important role, as shown in the following also for other isospin properties inside the mixed phase. 2 Isospin effects on the Mixed Phase Wecanstudyindetailtheisospindependenceofthetransitiondensities[1,2,3]. ThestructureofthemixedphaseisobtainedbyimposingtheGibbsconditions [16] for chemical potentials and pressure and by requiring the conservation of the total baryon and isospin densities: µH(ρH,ρH,T) = µQ(ρQ,ρQ,T) , B B 3 B B 3 µH(ρH,ρH,T) = µQ(ρQ,ρQ,T) , 3 B 3 3 B 3 PH(T)(ρH,ρH,T) = PQ(T)(ρQ,ρQ,T) , B 3 B 3 ρ = (1−χ)ρH +χρQ , B B B ρ = (1−χ)ρH +χρQ , (4) 3 3 3 where χ is the fraction of quark matter in the mixed phase and T is the temperature. The consistent definitions for the densities and chemical potentials in the two phases are given by : ρH = ρ +ρ , ρH = ρ −ρ , B p n 3 p n µ +µ µ −µ µH = p n, µH = p n , (5) B 2 3 2 for the Hadron Phase and 1 ρQ = (ρ +ρ ) , ρQ = ρ −ρ , B 3 u d 3 u d 3 µ −µ µQ = (µ +µ ), µQ = u d , (6) B 2 u d 3 2 4 for the Quark Phase. The related asymmetry parameters are: ρH ρ −ρ ρQ ρ −ρ αH ≡ − 3 = n p , αQ ≡ − 3 = 3 d u . (7) ρH ρ +ρ ρQ ρ +ρ B n p B d u Nucleon and quark chemical potentials, as well as the pressures in the two phases, are directly derived from the respective EoS. Inthis way we get the binodal surface which gives thephase coexistence region in the (T,ρ ,ρ ) space. For a fixed value of the total asymmetry α = −ρ /ρ B 3 T 3 B we will study the boundaries of the mixed phase region in the (T,ρ ) plane. B Since in general the charge chemical potential is related to the symmetry term of the EoS, [8], µ = 2E (ρ )ρ3, we expect critical and transition densities 3 sym B ρB rather sensitive to the isovector channel in the two phases. In the hadron sector we will use the Non-Linear Relativistic Mean Field mod- els, [7,8,3], with different structure of the isovector part, already tested to describe the isospin dependence of collective flows and meson production for heavy ion collisions at intermediate energies, [17,18,19]. We will refer to these different Iso-Lagrangians as: i) NL, where no isovector meson is included and the symmetry term is only given by the kinetic Fermi contribution, ii) NLρ when the interaction contribution of an isovector-vector meson is considered and finally iii) NLρδ where also the contribution of an isovector-scalar meson is accounted for. See details in Appendix A and refs.[7,8,3]. We will look at the effect on the hadron-quark transition of the different stiff- nessofthesymmetrytermathighbaryondensitiesinthedifferentparametriza- tions. As clearly shown in Appendix A, where a rather transparent form for the density dependence of the symmetry energy in RMF approaches is dis- cussed, the potential part of the symmetry term will be proportional to the baryon density in the NLρ choice and even stiffer in the NLρδ case. As already mentioned, in the quark phase we use the MIT-Bag Model, where the symmetry term is only given by the Fermi contribution. The Bag param- eter B is fixed for each baryon density to a constant, rather standard, value B1/4 = 160MeV, corresponding to a Bag Pressure of 85.7 MeV fm−3. In general for each effective interactive Lagrangian we can simulate the solu- tion of the highly non-linear system of Eqs.(4), via an iterative minimization procedure, in order to determine the binodal boundaries and the Critical End Point (CEP) (T , ρB) of the mixed phase. c c A relatively simple calculation can be performed at zero temperature. The 5 10 T=0.0 MeV 9 r Q/r B 0 8 c =1.0 7 r H/r B 0 6 0 r/B 5 r 4 3 NL NLr c =0.0 2 NLdr 1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 a Fig. 2. Dependence on the Hadron Symmetry Energy of the Lower (χ = 0.0) and Upper (χ = 1.0) Boundaries of the Mixed Phase, at zero temperature, vs. the asymmetry parameter. Quark EoS: MIT bag model with B1/4=160 MeV. isospin effect (asymmetry dependence) on the Lower (χ = 0.0) and Upper (χ = 1.0) transition densities of the Mixed Phase are shown in Fig.2 for various choices of the Hadron EoS. The effect of a larger repulsion of the symmetry energy in the hadron sector, from NL to NLρ and to NLρδ, is clearly evident on the lower boundary with a sharp decrease of the transition density even at relatively low asymmetries. Typical results for isospin effects on the whole binodal “surface” are presented in Fig.3 for symmetric and asymmetric matter. For the hadron part we have started from a NLρ effective Lagrangian very close to other widely used rela- tivistic effective models, e.g. see the GM3 of ref.[20] and the NL3 interaction of P.Ring and collaborators [21], which has also given good nuclear structure results, even for exotic nuclei. In the symmetric matter case the mixed phase is evaluated from the simpler Maxwell conditions. The results are shown in Fig.4 for the same hadron and quark EoS’s as in Fig.3 at temperatures T=0, 50 and 80 MeV. The equal chemical potential densities (intersection of the dotted line in the lower panel) must correspond to the equal pressure densities of the upper panels. We nicely see: i) at T=0 MeV the mixed phase is centered around ρ/ρ ≃ 7.0 , exactly 0 the α = 0 crossing point of Fig.1, confirming our energetic argument about the transition location; ii) the size of the mixed phase is shrinking with tem- 6 60 a=0.2 HIC a=0.0 50 ]40 V e M 30 HM MIXED QM [ T NLr B=(160 MeV)4 20 10 0 3 4 5 6 7 8 9 r/r 0 Fig. 3. Binodal surface for symmetric (α = 0.0) and asymmetric (α = 0.2) matter. Hadron EoS from NLρ interaction. Quark EoS: MIT bag model with B1/4=160 MeV.ThegreyregioncorrespondstotheconditionsreachedinHeavyIonCollisions simulations at few AGeV beam energies, see Section 3. T=0 MeV T=50 MeV T=80 MeV 1000 1000 1000 800 800 800 -3m) 600 HQ MM 600 600 V/f e 400 400 400 M P ( 200 200 200 0 0 0 0 2 4 6 8 0 2 4 6 8 0 2 4 6 8 r /r r /r r /r 2000 0 2000 0 2000 0 r /r =6.19 r /r =4.8 1800 H 0 1800 H 0 1800 r Q/r 0=8.29 r /r =5.4 Q 0 V) 1600 m =1597 MeV 1600 m =1332 MeV 1600 Me B B (B1400 1400 1400 m 1200 1200 1200 1000 1000 1000 0 2 4 6 8 0 2 4 6 8 0 2 4 6 8 r /r 0 r /r 0 r /r 0 Fig. 4. Maxwell construction forsymmetric(α = 0.0) matter at temperatures T=0, 50 and 80 MeV. For Pressure and Chemical Potentials we use: Hadron EoS (black curves) from NLρ interaction; Quark EoS (grey curves) from MIT bag modelwith B1/4=160 MeV. 7 perature, it is very narrow at T=50 MeV and finally at T=80 MeV we cannot have anymore a first order transition. In fact the Critical End Point is found at T ≃ 58MeV, ρ /ρ ≃ 3.8, P ≃ 120MeV/fm3 and µ ≃ 1130MeV, see c c 0 c c also Fig.3. The result is dependent on the choice of the Bag constant, with increasing of the critical temperature with the Bag value due to the reduction of the pressure in the quark phase, while the chemical potentials are not af- fected. This point will not be further discussed since here the main focus is on the isospin dependence of the mixed phase at lower temperatures, that can be probed in heavy ion collisions at intermediate energies. As expected, the lower boundary of the mixed phase is mostly affected by isospin effects. In spite of the relatively small total asymmetry, α = 0.2, we clearly observe in Fig.3 a shift to the left of the first transition boundary, in particular at low temperature, and an indication of a relatively “earlier” Critical End Point, around (T = 50−55MeV,ρ /ρ = 2.5−2.8). c c 0 Actually we see from Fig.3 that for asymmetric matter, we are not able to ex- plicitly reach the CEP of the binodal surface. Since this feature appears also in the following figures we must add a general comment. From the present results we cannot exclude a CEP at higher temperature and smaller baryon density. However we have also to mention some numerical problems. The solu- tion of the Gibbs conditions, the highly non-linear system of Eqs.(4), is found through an iterative multi-parameter minimization procedure (the Newton- Raphson method). When we are close to the Critical End Point, for each χ-concentration the baryon and isospin densities of the hadron and quark phases become very similar and we start to have problems in finding a definite minimum. We are working on improving the numerical accuracy but in any case thetrends towards the CEP arealways quite clearly observed, asasevere reduction of the mixed phase. In the following we will concentrate on properties of the mixed phase mostly located at high density and relatively low temperature, well described in the calculationandwithinthereach ofheavy ioncollisions inthefew AGeVrange, see Section 3. 2.1 Inside the Mixed Phase of Asymmetric Matter For α = 0.2 asymmetric matter, in the Figs.5, 6 we show also the (T,ρ ) B curves inside the Mixed Phase corresponding to a 20% and 50% presence of the quark component (χ = 0.2,0.5), evaluated respectively with the two choices, NLρ and NLρδ, of the symmetry interaction in the hadron sector. We note, as also expected from Fig.2, that in the more repulsive NLρδ case the lower boundary is much shifted to the left. However this effect is not so 8 60 c =0.0 and 1.0 c =0.2 with a =0.2 50 c =0.5 with a =0.2 V) 40 NLr e M 30 ( T HM QM 20 a H=0.2 a Q=0.2 10 0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0 r /r 0 Fig. 5. Asymmetric α = 0.2 matter. Binodal surface and (T,ρ ) curves for various B quark concentrations (χ = 0.2,0.5) in the mixed phase. Quark EoS: MIT Bag model with B1/4=160 MeV. Hadron EoS: NLρ Effective Interaction. 60 c =0.0 and 1.0 c =0.2 with a =0.2 50 c =0.5 with a =0.2 V) 40 NLdr e M 30 ( T HM QM 20 aH=0.2 a Q=0.2 10 0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0 r /r 0 Fig. 6. As in Fig.5, for the NLρδ Effective Interaction in the Hadron sector. evident for the curve corresponding to a 20% quark concentration, and almost absent forthe50%case. Theconclusionseems tobethatforastiffersymmetry term in a heavy-ion collision at intermediate energies during the compression stage we can have more chance to probe the mixed phase, although in a region with small weight of the quark component. 9 60 c =0.0 and 1.0 c =0.2 with a =0.2 50 c =0.5 with a =0.2 ) V 40 NLr e M ( 30 T HM QM 20 a H=0.2 a Q=0.2 10 0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0 r /r 0 Fig. 7. Asymmetric α = 0.2 matter. Binodal surface and (T,ρH,ρQ) curves for B B various quark concentrations in the mixed phase. Quark EoS: MIT Bag model with B1/4=160 MeV. Hadron EoS: NLρ Effective Interaction. 60 c =0.0 and 1.0 c =0.2 with a =0.2 50 c =0.5 with a =0.2 V) 40 NLrd e M 30 ( T HM QM 20 a H=0.2 a Q=0.2 10 0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0 r /r 0 Fig. 8. As in Fig.7, for the NLρδ Effective Interaction in the Hadron sector. In fact from the solution of the system Eq.(4) we get the baryon densities ρH,ρQ in the two phases for any χ value. In the Figs. 7, 8 we present the B B results for the same weights 20%,50% of the quark phase of the previous figures. The quark phase appears always with larger baryon density, even for the lowest value of the concentration. Can we expect some signatures related to the subsequent hadronization in the following expansion? Aninteresting possibility iscoming fromthestudy oftheasymmetry αQ inthe quark phase. In fact since the symmetry energy is rather different in the two 10