Low-momentum interactions with Brown-Rho-Ericson scalings and the density dependence of the nuclear symmetry energy Huan Dong and T. T. S. Kuo Department of Physics and Astronomy, Stony Brook University, New York 11794-3800, USA R. Machleidt Department of Physics, University of Idaho, Moscow, Idaho 83844, USA We have calculated the nuclear symmetry energy Esym(ρ) up to densities of 4 ∼ 5ρ0 with the effects from the Brown-Rho (BR) and Ericson scalings for the in-medium mesons included. Using theVlow−k low-momentuminteractionwithandwithoutsuchscalings,theequationsofstate(EOS) of symmetric and asymmetric nuclear matter havebeen calculated using a ring-diagarm formalism 1 wheretheparticle-particle-hole-holeringdiagramsareincludedtoallorders. TheEOSforsymmetric 1 nuclear matter and neutron matter obtained with linear BR scaling are both overly stiff compared 0 withtheempiricalconstraintsofDanielewiczetal. [9]. Incontrast,satisfactoryresultsareobtained 2 byeitherusingthenonlinearEricsonscalingorbyaddingaSkyrme-typethree-nucleonforce(TNF) n to the unscaled Vlow−k interaction. Our results for Esym(ρ) obtained with the nonlinear Ericson scaling are in good agreement with theempirical valuesof Tsang et al. [7] and Li et al. [10], while a those with TNF are slightly below these values. For densities below the nuclear saturation density J ρ0, the results of the above calculations are nearly equivalent to each other and all in satisfactory 0 agreement with theempirical values. 1 PACSnumbers: 21.65.Jk,21.65.Mn,13.75.Cs ] h t - I. INTRODUCTION plied to the study of E . As it is well known, most l sym c realistic V contain hard cores, or strong short-range u NN repulsions. This feature makes these interactions not n The nuclear matter symmetry energy is an important [ as well as very interesting subject in nuclear and astro- suitableforbeingdirectlyusedinnuclearmany-bodycal- culations;theyneedtobe‘tamed’beforehand. Formany nuclear physics. As reviewed extensively in the litera- 1 ture [1–8], it plays a crucial role in determining many years, this taming is enacted by way of the BHF theory v whereV is convertedinto the BruecknerG-matrix. A 0 important nuclear properties, such as the neutron skin NN 1 of nuclear systems,structure ofnuclei near the drip line, complication of the G-matrix is its energy dependence (see e.g. [26]), making it rather inconvenientfor calcula- 9 and neutron stars’ masses and radii. It is especially of 1 importance that constraintson the nuclear matter equa- tions. In the Vlow−k approach, a different ‘taming’ pro- 1. tionofstate(EOS)[9]andthedensity(ρ)dependenceof cedure is employed; it is performed by ‘integrating out’ the high-momentum components of V beyond a deci- 0 the symmetry energy Esym(ρ) [7,10] up to ρ≃4ρ0 have NN 1 been experimentally extracted from heavy-ioncollisions, mation scale Λ. In this way, the resulting Vlow−k is en- 1 ρ0 being the saturation density of symmetric nuclear ergy independent. Furthermore Vlow−k is nearly unique, v: matter. There have been a large number of theoretical namely the Vlow−ks deduced from various realistic VNN potentials (such as [27–30]) are nearly identical to each i derivationsofE (ρ)using,forexample,theBrueckner X Hartree-Fock(BsHymF)[11–13],DiracBHF[6,14–16],varia- other for decimation scale Λ≃2fm−1 [23, 24]. ar tional[17],relativisticmeanfield(RMF)[18]andSkyrme Using this Vlow−k interaction, we shall first calculate HF [19] many-body methods. The results of these theo- the equations of state (EOS) E(ρ,α) for asymmetric nu- reticalinvestigationshaveexhibited, however,largevari- clearmatter,fromwhichE (ρ)canbe obtained. Here sym ations for E (ρ). Depending on the interactions and E is the ground-state energy per nucleon and ρ is the sym many-body methods used, they can give either a ‘hard’ totalbaryondensity. α is the isospinasymmetry param- E (ρ),inthesensethatitincreasesmonotonicallywith eter defined as α = (ρ −ρ )/ρ, where ρ and ρ de- sym n p n p ρ up to ∼ 5ρ , or a ‘soft’ one where E (ρ) arises to a note, repspectively, the neutron and proton density and 0 sym maximum value at ρ ≃ 1.5ρ and then descends to zero ρ = ρ +ρ . Our EOS will be calculated using a ring- 0 n p at ∼ 3ρ [4, 8]. It appears that the predicted behav- diagrammany-bodymethod[31–33]. Asweshalldiscuss 0 ior of E (ρ) may depend importantly on the nucleon- later,thismethodincludestheparticle-particlehole-hole sym nucleon (NN) interactions and the many-body methods (pphh) ring diagrams to all orders. In comparison, only employed. thediagramswithtwoholelinesareincludedinthefamil- In the present work, we shall calculate the nuclear iar HF, BHF andDBHF calculations. In other words,in symmetry energy using the low-momentum interaction theseHFmethodsaclosedFermiseaisemployedwhilein Vlow−k derivedfromrealistic NNinteractionsVNN using the ring-diagramframeworkthe effects fromthe fluctua- arenormalizationgroupapproach[21–25]. Toourknowl- tionsoftheFermiseaaretakenintoaccountbyincluding edge, this renormalized interaction has not yet been ap- the pphh ring diagrams to all orders. 2 ThenuclearsymmetryenergyE (ρ)isrelatedtothe matrix equivalence equations: sym asymmetric nuclear matter EOS by T(k′,k,k2)=V (k′,k) E(ρ,α)=E(ρ,α=0)+E (ρ)α2+O(α4). (1) NN sym 2 ∞ V (k′,q)T(q,k,k2) + P NN q2dq, (3) The contributions from terms of orders higher than α2 π Z0 k2−q2 are usually negligibly small, as illustrated by our results insectionIII.Withsuchcontributionsneglected,wehave Tlow−k(k′,k,k2)=Vlow−k(k′,k) E (ρ)=E(ρ,1)−E(ρ,0). (2) sym + 2P Λ Vlow−k(k′,q)Tlow−k(q,k,k2)q2dq, (4) Thenthesymmetryenergyisjustgivenbytheenergydif- π Z0 k2−q2 ference between neutron and symmetric nuclear matter. In calculating E (ρ), the above EOS clearly play an sym important role. In our calculation, we shall require that T(k′,k,k2)=Tlow−k(k′,k,k2);(k′,k)≤Λ. (5) the NN interaction and many-body methods employed should give satisfactory results for E(ρ,1) and E(ρ,0) In the above V represents a realistic NN interaction of, respectively, neutron and symmetric nuclear matter. NN suchas the CDBonn potential [27]. P denotes principal- TheuseofVlow−k alone,however,hasnotbeenabletore- valueintegrationandtheintermediatestatemomentumq produce the empirical nuclear saturation properties, the isintegratedfrom0to∞forthewhole-spaceT andfrom predictedsaturationdensityandbindingenergyperpar- ticle being both too large compared with the empirical 0 to Λ for Tlow−k. The above Vlow−k preserves the low- valuesofρ ≃0.16fm−3andE ≃−16MeVforsymmetric energy phase shifts (up to energy Λ2) and the deuteron nuclearma0tter[31,32]. Toimprovethe situation,itmay bindingenergyofVNN. SinceVlow−k isobtainedbyinte- grating out the high-momentum components of V , it benecessarytoincludetheeffectsfromBrown-Rho(BR) NN is a smooth ‘tamed’ potential which is suitable for being scaling [34–36] for the in-medium mesons, or a three- used directly in many-body calculations. nucleon force (TNF) [40]. BR scaling is suitable only for thelowdensityregion;itsuggeststhatthemassesoflight vector mesons in medium are reduced ‘linearly’ with the density. We consider here the EOS up to about ∼ 5ρ 0 and at such high density the linear BR scaling is clearly not applicable. In the present work we shall adopt the nonlinear Ericson scaling [41] for the in-medium mesons and apply it to our E (ρ) calculations. The effects sym fromthe linearBRandnonlinearEricsonscalingsonthe nuclear EOS and symmetry energy will be studied. The organization of this paper is as follows. In sec- tionIIweshallbrieflydescribeourderivationofthelow- momentum interaction Vlow−k using a T-matrix equiv- alence approach. Some details about the calculation of the EOS for asymmetric nuclear matter from this inter- actionwiththepphhringdiagramssummedtoallorders FIG.1: Sampleringdiagramincludedintheequationofstate willalsobepresented. The Ericsonscalingis anonlinear E(ρ,α). Each wave line represents a Vlow−k vertex. The HF one-bubbleinsertions totheFermion lines areincludedtoall extension of linear BR scaling. The difference between orders. themwillbeaddressedinthissection. Ourresultswillbe presented and discussed in section III. A summary and conclsion is contained in section IV. Weusearing-diagrammethod[31–33]tocalculatethe nuclear matter EOS. In this method, the ground-state energy is expressed as E(ρ,α) =Efree(ρ,α)+∆E(ρ,α) where Efree denotes the free (non-interacting) EOS and II. FORMALISM ∆E is the energy shift due to the NN interaction. In our ring-diagram approach, ∆E is is given by the all- We shall calculate E (ρ) using a low-momentum ordersumofthepphh ringdiagramsasillustratedinFig. sym ring-diagramapproach[31–33], where the pphh ring dia- 1. Note that we include three types of ring diagrams, grams are summed to all orders within a model space of the proton-proton, neutron-neutron and proton-neutron decimationscaleΛ. Inthisapproach,weemploythelow- ones. The proton and neutron Fermi momenta are, re- momentum interaction Vlow−k [21–25]. Briefly speaking, spectively, kFp = (3π2ρp)1/3 and kFn = (3π2ρn)1/3. this interaction is obtained by solving the following T- With such ring diagrams summed to all orders, we have 3 [32, 33] 10−15%atnormalnuclearmatterdensityandzerotem- perature. (Pions are not scaled because they are pro- 1 tected by chiral symmetry.) Density-dependent nuclear ∆E(ρ,α)= dλ Ym(ij,λ) interactions obtained by applying the above scaling to Z0 m ijkl<Λ the light mesons (ω, ρ and σ) which mediate the NN X X ×Ym∗(kl,λ)hij|Vlow−k|kli, (6) potential have been employed in studying the properties of nuclear matter [31, 32, 36, 37] and the 14C →14 N where the transition amplitudes Y are obtaind from a β-decay [38]. pphh RPA equation [31–33]. Note that λ is a strength WeareinterestedintheEOSandE uptodensities sym parameter, integrated from 0 to 1. The above ring- as high as ρ≃5ρ , and at such high densities the above 0 diagram method reduces to the usual HF method if linear scaling is clearly not suitable. How to scale the only the first-order ring diagram is included. In this mesons in such high density region is still by and large case, the above energy shift becomes ∆E(ρ,α) = uncertain. We shall adopt here the Ericson scaling [41] HF 21 ninjhij|Vlow−k|iji where nk=(1,0) if k(≤,>)kFp for which is an extension of the BR scaling. In this scaling, proton and n =(1,0) if k(≤,>)k for neutron. a new relation for the quark condensate hq¯qi based on k Fn P It is well known that the use of the free-space V chiral symmetry breaking is employed, namely NN aloneisnotadequatefordescribingnuclearpropertiesat highdensities. Tosatisfactorilydescribesuchproperties, hq¯q(ρ)i 1 = . (10) one may need to include the three-nucleon force [40] or hq¯q(0)i 1+ ρΣπN the in-medium modifications to the nuclear interaction. fπ2m2π In the present work, we shall employ in our EOS calcu- Note that this relation agrees with the linear scaling re- lationsnuclearinteractionswhichcontainthein-medium lation of Eq.(8) for small ρ. The above scaling suggests modifications suggestedby the Brown-Rho(BR) [34,35] a non-linear scaling for meson mass and Ericson [41] scalings. These scalings are based on the relation [34, 35, 39] that hadron masses scale with 1/3 ∗ m 1 the quark condensate hq¯qi in medium as = (11) m 1+D ρ ρ0! ∗ 1/3 m hq¯q(ρ)i m =(cid:18)hq¯q(0)i(cid:19) (7) with D = ρf0π2Σmπ2πN, and we shall refer to this scaling as the nonlinear Brown-Rho-Ericson (BRE) scaling. Using ∗ where m is the hadron mass in a medium of density ρ, theempiricalvaluesfor(Σ , ρ , f , m ),wehaveD= πN 0 π π and m is that in free space. The quark condenstate hq¯qi 0.35±0.06. Inthepresentwork,weshallemploytheone- measures the chiral symmetry breaking, and its density boson exchange BonnA potential [30] with its (ρ, ω, σ) dependence in the low-density limit is related [42, 43] to mesons scaledusing both the linear (Eq.(9)) andnonlin- the free πN sigma term ΣπN by ear (Eq.(11)) scalings. This potential is chosen because it has a relatively simple structure which is convenient hq¯q(ρ)i ρΣ =1− πN (8) for scaling its meson parameters. hq¯q(0)i f2m2 π π wheref =93MeVisthepiondecayconstantandΣ = π πN III. RESULTS AND DISCUSSIONS 45±7MeV [44]. Applying the above scaling to mesons inlow-densitynuclearmedium,onehasthelinearscaling Using boththe unscaledandscaledBonnApotentials, [35] we first calculate the ring-diagram EOS for symmetric m∗ ρ nuclear matter to investigate if they can give saturation =1−C (9) m ρ properties in good agreement with the empirical values. 0 We employ the low-momentum interactions Vlow−k from where m∗ and m are, respectively, the in-medium and these potentials using a decimationΛ=3.0fm−1, which free meson mass, and C is a constant of value ∼ 0.15. is chosen because we are to study the EOS up to high TheabovescalingwillbereferredtoasthelinearBRscal- densities of ∼ 5ρ . As shown in Fig. 2, the EOS (la- 0 ing. Nucleon-nucleoninteractionsaremediatedbymeson belled ’Vlow−k alone’) calculated with the unscaled po- exchanges, and clearly the in-medium modifications of tential saturates at k ≃ 1.8fm−1, which is too large F meson masses can significantly alter the NN interaction. comparedwiththe empiricalvalue,anditalsooverbinds These modifications could arise from the partial restora- nuclear matter. We then repeat the calculation includ- tion of chiral symmetry at finite density/temperature ing the medium modifications from the BR scalings. For or from traditional many-body effects. Particularly im- the linear BR scaling (Eq.(9)), we have used C =0.128, ω portant are the vector mesons, for which there is now C =0.113 and C =0.102. These parameters are chosen ρ σ evidence from both theory [35, 45, 46] and experiment so as to have satisfactory saturation properties, namely [47, 48] that the masses may decrease by approximately theygiveE /A≃-15.5MeVandρ ≃0.17fm−3. InFig. 0 0 4 2wealsopresentourresultsobtainedwiththe nonlinear 120 BRE scaling (Eq.(11)) using parameters D =D =0.40 and D =0.30. They were chosen to providωe satisfρactory 100 Vlow-k alone σ linear BR results for E0/A and ρ0. It is of interest that for den- 80 nonlinear BRE sities . ρ0 the EOS given by the linear BR and nonlin- V] 60 Vlow-k with TNF e ear BRE scalings are practically eqivalent to each other. M 40 From Eqs.(7-11), we see that the parameters D and C A [ obtained from the density dependence of quark conden- /0 20 sates is 0.29.D .0.41 and C ≃D/3. It is noteworthy E 0 that the C and D parameters we have employed in the -20 EOS calculations agree well with the above theoretical values. -40 As also seen from Fig. 2, the above equivalence be- 0.8 1 1.2 1.4 1.6 1.8 2 2.2 -1 gins to disappear for densities larger than ρ . There the k [fm ] 0 F EOSgiven by the linear scaling is much stiffer than that given by the nonlinear one; the difference between them becomes larger and larger as density increases. In ad- FIG.2: Ring-diagramEOSscalculatedwiththeVlow−k inter- action alone, with the linear BR scaling of Eq.(9), with the dition to the above two EOS, we have also calculated nonlinearBRE scaling of Eq.(11), and with theaddition of a an EOS using the interaction given by the sum of the Skyrme-typethree-nucleon force (TNF). unscaled Vlow−k and an empiriral Skyrme three-nucleon force (TNF). The well-known emipirical Skyrme force [49] is of the form we have considered the nonlinear scaling. As displayed in Fig. 3, the EOS with the nonlinear BRE scaling is in V = V(i,j)+ V(i,j,k), (12) much better agreement with the constraint than the lin- Skyrme ear BR one. It satisfies the constraint well except being i<j i<j<k X X slightly above the constraint at densiies near ∼4.5ρ . It 0 where V(i,j) is a two-nucleon momentum dependent in- isofinterestthattheEOSusingVlow−k withtheSkyrme- teraction,andV(i,j,k)isazero-rangethree-nucleoninter- type TNF exhibits even better agreement with the con- action which has playedan indispensible role for nuclear straint. saturation. For nucleons in a nuclear medium of density ρ, this three-nucleon force becomes a density-dependent two-nucleon force commonly written as t 100 3 V (i,j)= ρδ(~r −~r ). (13) ρ 6 i j 3] m InFig. 2theEOSlabelled’Vlow−k withTNF’isobtained V/f e using the combinedinteraction of Vlow−k (unscaled) and M 10 Vρ. The parameter t3 is adjusted so that the resulting P[ EOSgivessatisfactorysaturationpropertiesforsymmet- linear BR nonlinear BRE ric nuclear matter. The EOS shown has t =2000 MeV- 3 V with TNF fm6. low-k 1 It is of interest that the above three EOS (linear BR 1 1.5 2 2.5 3 3.5 4 4.5 5 and nonlinear BRE, TNF) are nearly identical for densi- ρ/ρ 0 ties .ρ , butthey deviate fromeachotherwith increas- 0 ingdensities. Withoutexperimentalguidelinesaboutthe nuclearmatterEOSaboveρ ,itwouldbe difficulttode- FIG. 3: Comparison of the calculated equations of state for 0 termine which of these three EOS has the correct high symmetricnuclearmatterwiththeconstraint(solid-linebox) of Danielewicz [9]. See text for more explanations. density behavior. Fortunately, heavy-ioncollision exper- iments conducted during the last severalyears have pro- vided us with constraints of the EOS at high densities. So far we have studied the effects of the BR scalings Danielewicz et al. [9] have obtained a constraint on the andtheTNFthree-nucleonforceontheEOSforsymmet- EOS for symmetric nuclear matter of densities between ric nuclear matter. The neutron matter EOS is also an 2ρ and 4.5ρ , as shown by the red solid-line box in Fig. interestingandimportanttopic[50,51]. Itplaysacrucial 0 0 3. Comparing our three EOSs with their constraint, the roleindeterminingthenuclearsymmetryenergiesaswell linear BR EOS is clearly not consistent with the con- asthepropertiesofneutronstars. Itshouldbeofinterest straint and should be ruled out. This linear scaling is to study also the effects of the above BR/BRE scalings suitable for low densities, but definitely needs modifica- and the TNF force on the EOS of neutron matter. Us- tionathighdensities. Itisprimarilyforthispurposethat ing the same Vlow−k ring-diagram framework employed 5 for symmetric nuclear matter and the same C, D and t3 80 parameters, we have caculated the neutron matter EOS ρ=0.090 fm-3 ρ=0.090 fm-3 upto4.5ρ0. Ourcalculatedneutron-matterEOSaredis- 70 ρρ==00..232382 ffmm--33 ρρ==00..232382 ffmm--33 played in Fig. 4. Danielewicz et al. [9] have given two V] 60 e different constraints for the neutron matter EOS: a stiff M one(upperblacksolid-linebox)andasoftone(lowerred )[ 50 0 solid-linebox)whicharebothdisplayedinFig. 4. As we ρ, 40 ( cansee, the linear BREOS is againproducing too much E pressure. The nonlinear BRE EOS agrees well with the α)- 30 stiff constraint (upper box) while the TNF EOS is fully ρ(, 20 E within the soft constraint box. To further test these two 10 EOS(nonlinearBREandTNF),itwouldbeveryhelpful tohavenarrowerexperimentalconstraintsontheneutron 0 matter EOS. 0 0.2 0.4 0.6 0.8 α2 0.2 0.4 0.6 0.8 1 FIG. 5: Ring-diagarm equations of state of asymmetric nu- clear matter. See text for more explanations. 100 3m] obtained with the ’Vlow−k with TNF’ interaction while V/f for the right panel the ’nonlinear BRE’ interaction is e used. As seen, E(ρ,α) varies with α2 almost perfectly M 10 linearly, for a wide range of ρ. (Note that in Fig. 5 P[ we plot the energy difference E (ρ,α)−E (ρ,0).) linear BR sym sym nonlinear BRE Thisisadesirableandremarkableresult,indicatingthat Vlow-k with TNF ourring-diagramsymmetryenergycanbeaccuratelyob- 1 tained from the simple relation given by Eq.(2), namely 1 1.5 2 2.5 3 3.5 4 4.5 5 theenergydifferencebetweenneutronandsymmetricnu- ρ/ρ 0 clear matter. In Fig. 6, the ’shaded area’ represents the empirical FIG. 4: Comparison of the calculated equations of state for constraint, Eq.(14), of Li et al. [10]. As seen, there are neutron-matter with the constraints of Danielewicz [9]. See large uncertainties in the high-density region. The em- text for more explanations. pirical relation Eq.(15) of Tsang et al. [7] is given by the ’second curve from bottom’ in the figure. As seen, the density dependence of this relation is slightly below The symmetry energyE is a topic ofmuchcurrent sym the softest limit (lower boundary of the shaded area) of interest, and extensive studies have been carried out to Eq.(14). Our ’nonlinear BRE’ results are in the middle extract its density dependence from heavy-ion collision oftheshadedarea,ingoodagreementwiththeempirical experiments[7,10]. Basedonsuchexperiments,Li et al. constraint of [10]. Our results with the TNF force are [10] suggested an empirical relation below the empirical ones of both [10] and [7], giving a E (ρ)≈31.6(ρ/ρ )γ; γ =0.69−1.1, (14) softer density dependence than both. It may be noticed sym 0 that for densities (ρ . ρ ), the calculated and empirical 0 for constraining the density dependence of the symme- results are all in good agreement with each other. The try energy. Also based on such experiments, Tsang et symmetry energies given by them at ρ0 are all close to al. [7] recently proposeda new empiricalrelation for the ∼ 30MeV, which is also the only well determined em- symmetry energy, namely pirical value. Furthermore, our calculated symmetry en- ergies all increase monotonically with density. We have C ρ 2/3 C ρ γi required our nuclear matter EOS to satisfy certain em- s,k s,p Esym(ρ)= + (15) pirical constraints, and with such requirements it may 2 ρ 2 ρ (cid:18) 0(cid:19) (cid:18) 0(cid:19) be difficult for our present calculations to have a soft where Cs,k = 25MeV, Cs,p = 35.2MeV and γi ≈ 0.7. It Esym(ρ) as soft as the supersoft one of [8] which satu- shouldbeusefulandofinteresttocheckifourcalculated rates at density near ∼1.5ρ0. E (ρ) is consistent with the above relations. Wehavefoundthatoursymmetryenergiescanbewell sym Using the ring-diagram framework described earlier, fitted by expressions of the same forms as Eqs.(14) and we have calculated the ground-state energy E(ρ,α) for (15),withthe exponentsγ andγ treatedasparameters. i asymmetric nuclear matter. (Recall that the asymmetry In Table I, we compare the exponents determined from parameter is α = (ρ −ρ )/ρ.) Some representative re- ourresults withthe empiricalonesof[10] and[7]. The γ n p sultsareshowninFig. 5: theresultsintheleftpanelare exponent given by the nonlinear BRE scaling is in good 6 agreement with the empirical values of [10]. The empir- saturationpropertiescansatisfactorilybereproducedby ical γ of [7] is, however, about half-way between the γ including the medium corrections from the well-known i i obtained with ’nonlinear BRE’ and that with ’TNF’. linearBrown-Rhoscalingforthein-mediummesons,this scaling produces an EOS which is too stiff compared 200 with the Danielewicz constraints. We have found that theEOSobtainedwiththenonlinearBrown-Rho-Ericson Tsang et al. 160 nonlinear BRE scaling are in good agreementwith the Danielewicz con- Vlow-k with TNF straints. We have considered another method to ren- V] der the effective interaction density dependent, namely e 120 M adding a Skyrme-type three-nucleon force (TNF) to the [m unscaled Vlow−k interaction. The EOS so obtained are 80 y alsoingoodagreementwiththeDanielewiczconstraints, s E but the resulting neutron matter EOS is significantly 40 softer than that with the nonlinear scaling. The three methods (linear and nonlinear scalings, and TNF) all 0 havereproducedwellthe empiricalsaturationproperties 0 1 2 3 4 5 ofnuclearmatter(ρ ≈0.17fm−3andE /A≈−15MeV), 0 0 ρ/ρ0 but their results at high densities are different. We have determinedthe scaling parametersC (linear BRscaling) and D (nonlinear BRE scaling) by fitting the above sat- FIG. 6: Comparison of the density dependence of our cal- uration properties. It is encouraging that the results, culated symmetry energies with the empirical results of [7] (dot-dash line) and [10] (shaded area). (0.102.C .0.128) and (0.30.D .0.40), so obtained areactuallyingoodagreementwiththetheoreticalresult D ≃0.35±0.06≃3C given by Eqs.(7-11). Including the above medium modifications, we pro- TABLE I: Comparison of the density exponents for the nu- ceed to calculate the nuclear symmetry energies. We clearsymmetryenergy Esym(ρ). Theexponentsγ andγi are have found that the E (ρ,α) given by our asymmet- defined respectively in Eqs. (14) and (15). sym ric ring-diagram calculations depends on α2 almost per- γ γ i fectly linearly. This is a rather surprising and useful re- Li et al. [10] 0.69-1.1 sult, suggesting that the symmetry energy can be reli- Tsang et al. [7] 0.7 ably obtained from the simple energy difference between non-linear BRE 0.82 1.04 TNF 0.53 0.43 symmetricnuclearmatterandneutronmatter. Oursym- metry energies obtained with the nonlinear BRE scaling agree well with the empirical constraints of [10], and are slightly above the empirical values of [7]. Our results with the TNF force is slightly below the empirical re- IV. SUMMARY AND CONCLUSION sults of both [10] and [7]. The non-linear Ericsonscaling has given satisfactory results for the equations of states EmployingtheVlow−k low-momentuminteractions,we of nuclear matter and nuclear symmetry energies up to havecalculatedthenuclearsymmetryenergyE (ρ)up sym a density of ∼ 5ρ . We believe this scaling provides a 0 to a density of ∼ 5ρ using a ring-diagram framework 0 suitableextensionofthelinearBRscalingtomoderately where pphh ring diagrams are summed to all orders. We high densities of . 5ρ . Our calculated E (ρ) all in- 0 sym firstcalculatetheEOSforsymmetricnuclearmatterand crease monotonically with ρ up to ∼ 5ρ . It may be of 0 neutron matter and compare our results with the corre- interest to carry out further studies about the possibil- sponding empirical constraints of Danielewicz et al. [9]. ity of obtaining a supersoftsymmetry energywhich may To have satisfactory agreements with such constraints, saturate at some low density of ∼1.5ρ [8]. 0 wehavefounditnecessarytoincludecertainmediumcor- rectionstothefree-spaceNNinterations. Inotherwords, Acknowledgement We thank Professor Danielwicz the effective NN interactions in medium are different for sending us the experimental data, and G.E. Brown fromthoseinfreespace,andwhenusingtheminnuclear and E. Shuryak for many helpful discussions. This work many-body problems it may be necessary to include the is supported in part by the U.S. Department of Energy renormalizationeffectsduetothepresenceofothernucle- under Grant Nos. DE-FG02-88ER40388 and DE-FG02- ons. We have considered severalmethods to incorporate 03ER41270(R.M.), andthe U.S. NationalScience Foun- such medium corrections. Although the nuclear matter dation under Grant No. PHY-0099444. [1] V.Baran et al.,Phys. Rep.410, 335(2005). [2] A. W. Steiner, M. Prakash, J. Lattimer and P.J. Ellis, 7 Phys.Rep.411, 325(2005). [27] R. Machleidt, Phys.Rev.C 63 024001 (2001). [3] J. M. Lattimer and M. Prakash, Phys. Rep. 442, [28] V.G.J.Stoks,R.A.M.Klomp,C.P.F.TerheggenandJ.J. 109(2007). de Swart, Phys. Rev.C 49, 2950 (1994). [4] B.A.Li,L.W.ChenandC.M.Ko,Phys.Rep.464,113 [29] R.B.Wiringa,V.G.J.StoksandR.Schiavilla,Phys.Rev. (2008). C 51, 38 (1995). [5] M.DiToroetal.,Prog.Part.Nucl.Phys.62,383(2008). [30] R. Machleidt, Adv.Nucl. Phys. 19, 189-376 (1989). [6] F. Sammarruca, Int. J. Mod. Phys. E 19, 1259 (2010); [31] L. W. Siu, J. W. Holt, T. T. S. Kuo and G. E. Brown, arXiv:1002.0146 [nucl-th]. Phys. Rev.79, 0540004 (2009) [7] M.B. Tsang, Yingxun Zhang, P. Danielewicz, M. Fam- [32] H. Dong, T.T.S. Kuo and R. Machleidt, Phys. Rev. C inao, Zhuxia Li, W.G. Lynch and A.W. Steiner, Phys. 80, 065803(2009). Rev.Lett. 102, 122701(2009). [33] H. Q. Song, S. D. Yang and T. T. S. Kuo, Nucl. Phys. [8] Xiao et al.,Phy.Rev.Lett. 102, 062502 (2009). A462 (1987) 491. [9] P.Danielewicz,R.LaceyandW.G.Lynch,Science298, [34] G.E. Brown and M. Rho, Phys. Rev. Lett. 66, 1592 (2002). 2720(1991). [10] B.A. Li and L.W. Chen, Phys. Rev.C72,064611(2005). [35] T. Hatsudaand S.H.Lee, Phys. Rev.C 46, R34(1992). [11] I. Bombaci and U. Lombardo, Phys. Rev. C44, 1892 [36] R. Rapp, R. Machleidt, J.W. Durso and G.E. Brown, (1991). Phys. Rev.Lett. 82, 1827(1999). [12] W.Zuo, A. Leguene, U. Lombardo, and J. F. Mathiot, [37] J.W. Holt, G.E. Brown, Jason D. Holt and T.T.S. Kuo, Eur. Phys.J. A14, 469(2002). Nucl. Phys. A785, 322(2007). [13] Z. H.Li et al.,Phys. Rev.C74, 047304(2006). [38] J.W. Holt, G.E. Brown, T.T.S. Kuo and R. Machleidt, [14] D.AlonsoandF.Sammarruca,Phys.Rev.C67,054301 Phys. Rev.Lett. 100, 062501(2008). (2003). [39] Y. Nambu and G. Jona-Lasino, Phys. Rev. 122, [15] P.KrastevandF.Sammarruca,Phys.Rev.C74,025808 345(1961). (2006). [40] S.K.Bogner,A.Schwenk,R.J.FurnstahlandA.Nogga, [16] E.N.E. van Dalen, C. Fuchs and A. Faessler, Eur. Phys. Nucl. Phys. A763 (2005) 59. J. A31, 29(2007). [41] M. Ericson, Phys.Lett. B 301,11(1993) [17] R.B. Wiringa et al.,Phys. Rev.C38, 1010(1988). [42] T.D. Cohen, R.J. Furnstahi, D.K. Griegel, Phys. Rev. [18] L.W. Chen et al.,Phys.Rev. C76, 054316(2007). C45,1881(1992). [19] L.W.Chen,C.M.KoandB.A.Li,Phys.Rev.Lett.94, [43] M. Lutz, S. Klimt, W. Weise, Nucl. Phys. 032701(2005). A542,521(1992). [20] A. Akmal, V.R. Pandharipande and D.G. Ravenhall, [44] J. Gasser, H. Leutwyler and M.E. Sainio, Phys. Lett. B Phys.Rev.C58, 1804(1998). 253,252(1991). [21] S.K.Bogner,T.T.S. Kuo,L.Coraggio, A.Covello, Nucl. [45] M.HaradaandK.Yamawaki,Phys.Rept.381(2003)1. Phys.A684, 432(2001) [46] F. Klingl, N. Kaiser, and W. Weise, Nucl. Phys. A624 [22] S.K. Bogner, T.T.S. Kuo, L. Coraggio, A. Covello and (1997) 527. N.Itaco, Phys.Rev.C65, 051301(R)(2002) [47] D. Trnkaet al., Phys.Rev.Lett. 94 (2005) 192303. [23] S.K. Bogner, T.T.S.Kuo and A. Schwenk, Phys. Rep. [48] M. Narukiet al. Phys. Rev.Lett. 96 (2006) 092301. 386,1 (2003) [49] P. Ring and P. Schuck, The Nuclear Many-Body Prob- [24] S.K. Bogner, T.T.S.Kuo, A. Schwenk, D. R. Entem and lem (Springer-Verlag, New York, 1980), and references R.Machleidt, Phys.Lett. B576, 265(2003). quoted therein. [25] J.D.Holt,T.T.S.KuoandG.E.Brown,Phys.Rev.C [50] B.FriedmanandV.R.Pandaripande,Nucl.Phys.A361, 69 (2004) 034329. 502(1981) [26] T.T.S. Kuo,Z.Y.MaandR.VinhMau,Phys.Rev.33, [51] B. A.Brown, Phys. Rev.Lett. 85, 5296 (2000). 717(1986).