Large-scale calculations of supernova neutrino-induced reactions in Z=8-82 target nuclei N. Paar1,∗ H. Tutman1, T. Marketin1,3, and T. Fischer2,3 1Physics Department, Faculty of Science, University of Zagreb, Croatia 2GSI Helmholtzzentrum für Schwerionenforschung, Planckstraße 1, D-64291 Darmstadt, Germany and 3 1 3Institut für Kernhysik, Technische Universität Darmstadt, 0 2 Magdalenenstraße 12, 64289 Darmstadt, Germany n a (Dated: January 4, 2013) J 3 ] h t - l c u n [ 2 v 5 5 6 2 . 0 1 2 1 : v i X r a 1 Abstract Background: In the environment of high neutrino-fluxes provided in core-collapse supernovae or neutron star mergers, neutrino-induced reactions with nuclei contribute to the nucleosynthe- sis processes. A number of terrestrial neutrino detectors are based on inelastic neutrino-nucleus scattering and modeling of the respective cross sections allow predictions of the expected detector reaction rates. Purpose: To provide a self-consistent microscopic description of neutrino-nucleus cross sections involving a large pool of Z = 8 − 82 nuclei for the implementation in models of nucleosynthesis and neutrino detector simulations. Methods: Self-consistent theory framework based on relativistic nuclear energy density functional is employed to determine the nuclear struc- ture of the initial state and relevant transitions to excited states induced by neutrinos. The weak neutrino-nucleus interactionisemployedinthecurrent-current formandacompletesetoftransition operators is taken into account. Results: We perform large-scale calculations of charged-current neutrino-nucleus cross sections, including those averaged over supernova neutrino fluxes, for the set of even-even target nuclei from oxygen toward lead (Z = 8− 82), spanning N = 8− 182 (OPb pool). The model calculations include allowed and forbidden transitions up to J = 5 multipoles. Conclusions: The present analysis shows that the self-consistent calculations result in consid- erable differences in comparison to previously reported cross sections, and for a large number of target nuclei the cross sections are enhanced. Revision in modeling r-process nucleosynthesis based on a self-consistent description of neutrino-induced reactions would allow an updated insight into the origin of elements in the Universe and it would provide the estimate of uncertainties in the calculated element abundance patterns. PACS numbers: 21.30.Fe, 21.60.Jz,23.40.Bw,25.30.-c ∗Electronic address: [email protected] 2 I. INTRODUCTION Neutrino-induced reactions on nuclei play an important role in nuclear astropyhsics, in particular during core-collapse supernova evolution and nucleosynthesis [1, 2]. Elastic neu- trino scattering on nuclei and nucleons determines the neutrino trapping and the diffusion time scale of the outwards streaming neutrinos [3]. A variety of processes contribute to the energy loss in stellar interiors, e.g. pair, photo-, plasma, nucleon-nucleon bremsstrahlung, and recombination neutrino processes [4]. In addition, inelastic neutrino-electron scattering mainly thermalizes the neutrino spectra [5] and inelastic neutral-current neutrino-nucleus scattering has a large contribution to the opacity [6, 7], during the core collapse and subse- quent explosion phase. Additionally, in the environment of exploding massive stars, which has long been explored as a possible site for the r-process nucleosynthesis, charged-current neutrino-nucleus reactions play an important role in the production of chemical elements. As pointed out in Ref. [8], there are many interesting effects of neutrino-induced reactions before, during, and after the r-process. To explore their role in nucleosynthesis reliable neutrino-nucleus reaction rate compilations involving also nuclei with large neutron excess are required. Available data on neutrino-nucleus cross sections are limited to deuterium [9], and 12C and 56Fe target nuclei, obtained by the LSND [10] and KARMEN [11, 12] col- laborations, and at LAMPF [13]. Therefore, only theoretical approaches can provide cross sections for a large number of target nuclei that are involved in various applications of neu- trino physics and astrophysics. Modeling neutrino-nucleus interactions is also important in view of the current research and development of neutrino detectors, e.g., for supernova and solar neutrinos, neutrinos produced in laboratories, and geoneutrinos. The ongoing and planned neutrino detector facilities involve a variety of target materials, induced reac- tions and scientific objectives, e.g., MOON [14], MiniBooNE [15], MINOS [16], SNO+ [17], OPERA [18], LVD (Large Volume Detector)[19], ORLaND experiment proposal at the Spal- lation Neutron Source (SNS) [20], NOvA neutrino experiment [21], Daya Bay reactor neu- trino experiment [22], etc. There is also interesting concept of beta-beamsfor theproduction of neutrinos by using β-decay of boosted radioactive ions [23, 24]. The proposal to establish a beta-beam facility that could produce low-energy neutrino beams in the 100 MeV energy range would allow direct insight into the neutrino-induced reactions in nuclei as well as its underlying structure involving a non-trivial combination of nuclear allowed and forbidden 3 transitions [24]. At low neutrino energies, the neutrino-induced reactions are sensitive to the properties of nuclei involved, i.e. their initial and excited states. Therefore it is necessary to employ a mi- croscopic framework providing reasonable description of nuclear structure properties. Over the past years, a variety of advanced microscopic models have been developed and employed in studies of charged-current neutrino-induced reactions at low energies, also including var- ious particle decay channels. In particular, these include the nuclear shell model [25–31], random phase approximation (RPA) [8, 32–37], continuum RPA (CRPA) [38–42], and the hybrid model which combines the shell model for allowed transitions, with the RPA to ac- count for the forbidden transitions [30, 43–45]. Frameworks based on quasiparticle RPA (QRPA) have also recently been developed, based on Skyrme functionals [34, 46], Brueckner G matrix employed for two-body interaction by solving the Bethe-Salpeter equation using Bonn-CD potential [47–49], and projected QRPA [50–52]. The Fermi gas model [53, 55, 56] andFermiliquidtheory(FLT)[57]have alsobeenemployed instudies oflow-energycharged- current neutrino-nucleus reactions. At finite temperature in stellar environments, thermal population of the excited states may enhance the weak interaction rates and cross sections at low neutrino energies [58, 59]. Despite considerable progress in the development of advanced theoretical frameworks, up to this time only a limited number of microscopic models have been employed in large-scale calculations of neutrino-induced reactions and their implementation in supernova simula- tions. Because the complete modeling of neutrino-induced reactions necessitates the inclu- sion not only of Gamow-Teller transitions but also contributions from forbidden transitions and other higher multipoles, covering a large pool of target nuclei represents a computation- ally very demanding problem. Although the shell modelprovides a very accurate description of ground state wave functions, the description of high-lying excitations necessitates the use of large model spaces which often leads to computational difficulties, making the approach applicable essentially only to allowed transitions in light and medium-mass nuclei. For sys- tematic studies of neutrino-nucleus cross sections throughout the nuclide chart including the heavy nuclei, microscopic calculations must therefore be performed using models based on the QRPA. The first globalcalculations of electron neutrino-nucleus cross sections have been conducted using the gross theory of β-decay [60, 61]. Two microscopic frameworks have been employed inlarge-scalecalculations ofneutrino-induced reactions i)extended Thomas-Fermi 4 plus Strutinsky integral (ETFSI) and continuum quasiparticle random phase approximation (CQRPA) [62], and ii) RPA with Landau-Migdal (LM) force, using Wood-Saxon potential (WS) to determine the single-particle basis of target nuclei [63]. There are no up-to-date systematic calculations conducted in the framework of energy density functional. Recently the framework based on relativistic nuclear energy density functional (RNEDF) has been introduced in modeling the charged-current neutrino-nucleus cross sections [64], and it has also been extended for applications in modeling neutral-current neutrino-induced reactions [65]. In the case of iron group nuclei, comparison of the results for charge exchange reactionsobtainedusingSkyrmefunctionalsandtheRNEDF,aswellaswiththeshellmodel, showed reasonable theoretical uncertainty inherent in modeling neutrino-nucleus cross sec- tions. However, in view of applications in supernova and r-process simulations, which also involve unstable nuclei far from the valley of stability, where no experimental data are avail- able, it is necessary to provide independent insight into relevant neutrino-induced processes from various models and effective interactions. The main objective of this work is conduct- ing large-scale calculations of charged-current neutrino-nucleus cross sections in a large pool of nuclei from oxygen towards lead, based on the RNEDF. In addition to the overall cross sections covering the range of neutrino energies up to 100 MeV, calculations also include the cross sections averaged over neutrino fluxes for the range of temperatures characteristic of various stages of stellar evolution. In addition to the Fermi-Dirac neutrino spectra we also apply supernova spectra from a recent simulation that includes three-flavor Boltzmann neu- trino transport. Particular aim of this work is to emphasize the role of forbidden transitions in modeling neutrino-nucleus cross sections in large pool of nuclei, including both stable and nuclei away from the valley of stability. The relevance of forbidden transitions has already been discussed in several studies, e.g. Refs. [24, 46, 66]. In the study of neutrino capture by r-process waiting point nuclei, first forbidden strength, together with the low-lying Gamow- Teller transitions, increased the rate of neutrino scattering from very neutron-rich nuclei by a factor of at least 2 and in some instances even by a factor of 5 [66]. As pointed out in mi- croscopic calculations based on Skyrme functionals in Ref. [46], the properties of forbidden states are closely related to the neutrino-nucleus cross sections, and could be extracted by using neutrinos from low-energy beta-beams. The paper is organized as follows. Section II contains the basic theoretical background for the neutrino-nucleus cross sections in the charged-current channel based on weak Hamil- 5 tonian and the RNEDF. The results of large-scale calculations of the neutrino-nucleus cross sections are illustrated and discussed in Sec. III. The conclusions of the present work are summarized in Sec. IV. II. THEORETICAL BACKGROUND In the present work we explore the charged-current neutrino-nucleus process, ∗ − ν +X → X +e , (1) e (Z,N) (Z+1,N−1) wheretheincoming electronneutrino (ν )induces charge-exchange reactionintargetnucleus e X(Z,N). The general formalism of the neutrino-nucleus cross sections, derived assuming the Hamiltonian of the weak interaction in the current-current form, is given in details in Refs. [53, 54]. The cross sections include the transition matrix elements between the initial |J i and final nuclear state |J i, for the charge Mˆ , longitudinal Lˆ , transverse electric TˆEL, i f J J J and transverse magnetic TˆMAG multipole operators [54]. In the present work the RNEDF J is employed in calculations of the transition matrix elements contributing to the neutrino- nucleus cross sections. The RNEDF has already been successfully employed in studies of giantresonancesandexoticmodesofexcitation[68–73], β-decayratesofr-processnuclei[74], muon capture [75] and stellar electron capture rates [76], and in constraining the neutron skin in nuclei [77, 78]. More details about the implementation of the RNEDF in modeling charged-current neutrino-nucleus reactions are given in Refs. [64, 67]. The RNEDF based framework employs the self-consistent mean field for nucleons and minimal set of meson fields; isoscalar scalar σ-meson (Jπ = 0+,T = 0), isoscalar vector ω- meson (Jπ = 1−,T = 0) and the isovector vector ρ-meson (Jπ = 1−,T = 1), supplemented with the electromagnetic field. The meson-nucleon interaction is included with a minimal set of the interaction terms, where the vertex functionals include explicit dependence on the vector density. The nuclear ground state properties are described using the relativistic Hartree-Bogoliubov model (RHB), and relevant transitions induced by neutrinos are calcu- lated in the relativistic quasiparticle random phase approximation (RQRPA). More details on the RHB model based on effective density-dependent interactions are given in Ref. [79]. The RQRPA is formulated in the canonical single-nucleon basis of the RHB model [80, 81]. In modeling the neutrino-nucleus cross sections, important advantage of this framework is 6 that it is fully consistent in view of the effective interactions employed. In the particle-hole (ph) and pairing (pp) channels, the same interactions are used in the RHB equations that determine the canonical quasiparticle basis, and in the matrix equations of the RQRPA. In this way, one can employ the same RNEDF in description of the weak processes throughout the nuclide map without any additional adjustments to the properties of specific target nu- clei under consideration. For the model parameters that determine the density-dependent couplings and the meson masses, in this work DD-ME2 parameterization is used [82]. The pairing correlations in open shell nuclei are described by the finite range Gogny interaction, with parameterization D1S [83]. Complete calculation of inelastic neutrino-nucleus reactions spanning the range of neu- trino energies up to ≈ 100 MeV necessitates the inclusion of a number of transitions with various multipoles J [64]. Although higher-order multipoles have rather small contributions at low incoming neutrino energies, these can not be neglected at energies about tens of MeV. In the present study, multipoles up to J = 5 have been taken into account. The large scale calculations of neutrino-nucleus cross sections, involving more than 1000 nuclei and a complete set of all multipoles up to J = 5 with both parities, necessitate considerable com- putational effort. Therefore, for the purposes of the present work, computational framework has been developed using parallel computing methods based on Message Passing Interface (MPI) [84] for the implementation on cluster and grid computer systems. III. RESULTS AND DISCUSSION By employing the model outlined in the previous section, we have conducted large-scale calculations of (ν ,e−) reactions in the OPb pool of even-even target nuclei spanning the e range from oxygen toward lead (Z=8-82) with neutron number N=8-182. The calculations include excitations of all multipoles up to J = 5 and both parities. In the following we explore the systematic behavior of the overall cross sections throughout the pool under consideration, including the cross sections averaged over the Michel neutrino flux obtained from the decay at rest (DAR) of µ+ [13] 96E2 fM(Eν) = ν (mµ −2Eν). (2) m4 µ 7 To simulate the supernova neutrino spectrum, at first instance we use the Fermi-Dirac distribution 1 E2 fFD(Eν) = ν . (3) T3exp[(E /T)−η]+1 ν In Fig. 1 the inclusive neutrino-nucleus cross sections, averaged over the Michel neutrino flux, are shown for the OPb pool of nuclei. Stable nuclei in the pool are denoted by filled circles. In addition, the calculated cross sections for 12C and 56Fe are especially emphasized in comparison to the KARMEN experimental data [11, 12]. One can observe that the pool of nuclei calculated with the same energy density functional without any adjustments of the model parameters fits nicely into two experimental data points. The cross sections increase systematically with increasing neutron number, resulting in values larger up to a factor ≈ 2 − 3 in comparison to those in the valley of stability. We note that sharp edge of the data at large A denotes boundary values for the pool under consideration, obtained for Pb isotope chain. Figure 2 shows the (ν ,e−) cross sections in the OPb pool of nuclei, averaged over the e supernova neutrino flux given by Eq. (3) in the case of T = 4 MeV and η = 0. The results can be compared to those of the stable target nuclei. While the cross sections display a similar pattern as in the previous case with the Michel spectrum, the details of the data, however, depend on (T,η) of the neutrino spectrum. The cross sections for various groups of target nuclei are separately displayed in Fig. 2: stable nuclei, proton-rich nuclei constrained by N/Z < 1, and neutron-rich nuclei with with N/Z > 1.5. The cross sections exhibit a systematic behavior throughout the nuclide map. In the case of neutron rich nuclei, the cross sections are larger in comparison to the stable nuclei, due to the increased number of neutrons that participate in charge-exchange neutrino-induced reactions. For proton-rich nuclei, the reaction pattern is opposite due to þthe smaller number of neutron-proton 2qp configurations, i.e., the (ν ,e−) cross sections are considerably reduced in comparison to e those of stable nuclei. This point is further illustrated in Fig. 3 where we plot the same cross sections versus the difference between the neutron and the proton numbers. For nuclei with N − Z > 1 the cross sections increase significantly with increasing number of excess neutrons. Although some scattering is apparent in the data, most of the results cluster along an almost linear function of the number of excess neutrons. Temperature dependence of neutrino-induced reactions within the OPb pool is illustrated in Fig. 4, where the cross sections averaged over theFermi-Diracdistribution (η = 0)areshown forstable targetnuclei 8 in the range T = 2−8 MeV. At low temperatures, the cross sections appear rather scattered due to the stronger dependence on low-energy excitations in nuclei. However, for higher T, the cross sections are rather smooth and considerably larger due to the inclusion of a number of multipoles contributing at higher neutrino energies. The full set of calculations has been completed for the OPb pool in the range of temperatures T = 0 − 10 MeV and complete tables are available on request. In view of applications in astrophysical models and in predicting the detector response to neutrinos involving various target nuclei, it is crucial to assess the sensitivity of the neutrino- nucleus cross sections on the theoretical frameworks and effective interactions employed. In the case of iron group nuclei, it has been shown that by employing different microscopic models, one can estimate reasonable theoretical uncertainty in neutrino-nucleus cross sec- tions averaged over the Michel spectrum, i.e., for 56Fe < σ > = (258±57)×10−42 cm2 [67]. th Since for nuclei far from the valley of stability various models result in larger differences in nuclear structure properties, one could also expect larger sensitivity in the results of modeling neutrino-induced reactions. In the present analysis theoretical uncertainties are assessed on the basis of the OPb pool of nuclei. Figures 5, 6 and 7 show the dependence of the flux-averaged neutrino-nucleus cross sections per nucleon (T = 4 MeV, η = 0) on the neutron number in the cases of Ni, Sn, and Pb isotopic chains, respectively. In these figures we compare our results to the available sets of cross sections based on two other theoretical frameworks: ETFSI + CQRPA [62] and RPA (WS+LM) [63], where the former contains only contributions from the Gamow-Teller and Fermi transitions, while the latter includes all transitions up to J = 3. The results of the present RQRPA analysis for < σ > /A are νe shown separately for the full calculation including all transitions up to J = 5, and partial cross sections obtained only for the isobaric analog and Gamow-Teller transitions. The RQRPA results for N > Z nuclei, obtained taking into account only the Fermi and the Gamow-Teller transitions, are consistently higher than the corresponding values calculatedusingtheETFSI+CQRPAframeworkforallthreeisotopicchains. Thedeviation israther small intheregionaroundthe valleyofstability inNiandSnisotopes, but increases with additional neutrons. In the case of the Pb isotopic chain, the difference is significant even for the lightest isotopes considered in this study. For nuclei with the proton-to-neutron ratio Z/N < 1, the ETFSI + CQRPA framework predicts a change of the general trend and an anomalous increase of the cross sections with a corresponding decrease of the number of 9 neutrons. Theagrementbetween resultsthattakeintoaccoutforbiddentransitions, RQRPA and RPA(WS+LM), is much better in all three isotopic chains, although for very neutron- rich Ni and Sn isotopes the RQRPA predicts higher values of the flux-averaged cross section. In the case of the Pb isotopic chain, the agreement between the RQRPAand RPA(WS+LM) cross sections is excellent for all the isotopes under consideration. There are several reasons for the shown deviations. Each model employs different effective interactions that result in variations of the excitation pattern contributing to the cross sections, and the present calculations are based on a fully self-consistent approach to the neutrino-induced reactions. The three models agree best for nuclei around the valley of stability, where most of the experimental data on nuclear structure properties are available. In the very neutron rich region the differences between various models increase and grow relatively large. However, in all three isotopic chains the agreement of the RQRPA results is much better with the values obtained using the RPA (WS + LM) model than with the ETFSI + CQRPA, partly indicating the importance of the forbidden transitions. In the case of cross sections averaged over the supernova neutrino flux for (T = 4, η = 0), the impact of the self-consistent and complete calculations including all relevant multipoles isexplored inview ofprevious knowledge onneutrino induced reactions onalarge-scalebasis within the OPb pool. Figure 8 shows the ratio of the cross sections averaged over supernova neutrino flux of the present work and ETFSI + CQRPA [62] model. In order to identify nuclei with pronounced discrepancies, the following groups of nuclei are separately denoted in figure: stable nuclei, proton-rich nuclei defined by N/Z < 1 and neutron-rich nuclei with N/Z > 1.5. For target nuclei where the cross sections are available for both models, the present large-scale calculations result mainly in systematically larger cross sections, for most ofnucleiuptothefactorof4,andforsmallernumber ofnucleiuptothefactorof7. However, for a limited set of medium mass nuclei the cross sections of this work are smaller than the ETFSI+CQRPA ones, and these are mainly limited to neutron deficient nuclei in the range 50 / A / 100. The reason is the anomalous increase of the cross sections in the ETFSI + CQRPA model, when moving along a particular isotope chain from N = Z nucleus toward proton rich nuclei (see Figs. 5 and 6). This behavior has not been observed in the present analysis. Apart from this anomaly, one can observe systematic linear increase of the ratio of cross sections with the nuclear mass. As already discussed in the case of isotopic chains, in addition to the self-consistent implementation of the RNEDF, the model employed in 10