Huther, Lutz (2014)
Nucleosynthesis in neutrino-driven winds.
Technische Universität Darmstadt
Dissertation, Erstveröffentlichung

Kurzbeschreibung (Abstract)

In this thesis the influence of neutrino-induced spallation rates on the nucleosynthesis is investigated. These neutrino-induced rates are studied in various nucleosynthesis processes such as the ν p-process, the r-process and the ν-process. The astrophysical site connecting all these different processes is the core-collapse supernova. Therefore our nucleosynthesis results are dependent on the information obtained by core-collapse supernova simulations and their correct input physics. The production of heavy elements in the core-collapse supernova is dominantly related to the neutrino-driven wind phase, in which matter is ejected from the proto-neutron star due to neutrino-nucleon interactions. Astrophysical parameters in the neutrino-driven wind, such as entropy and electron fraction, will determine which nucleosynthesis process occurs. The traversing shock will, when passing through the outer layers of the star, also have influence on the final nucleosynthesis. The matter in the outer layers of the star will also be subject to the neutrino fluxes from the proto-neutron star. The question that we would like to raise in this thesis is, if neutrino-induced spallation processes have influence on the nucleosynthesis. The neutrino-induced spallation rates have been obtained in a two-step process. In a first step the excitation of the nucleus is described. The second step describes the decay of the excited nucleus. The neutrino- nucleus excitation rates have been calculated for charged-current and neutral-current processes. The nuclear excitation spectra of the target nuclei are calculated with the help of a spherical RPA based on Woods-Saxon single-particle states, adjusted to reproduce the correct separation energies. The used interaction has been of the Landau-Migdal type. The excitation-spectrum includes excitations up to J = 4. For the neutrino energies in a core-collapse supernova, the excitation spectra peaks in the region of the giant resonances, which can be located above the particle threshold. Thus, it it necessary to consider not only the total cross section, but instead to consider partial cross sections for the spallation of light particles from the excited nucleus. This approach is described in a statistical model, where the usual formation cross section is replaced by the neutrino-induced nuclear excitation cross section. The one-particle spallation probability has been calculated with the statistical code MOD-SMOKER which is based on SMOKER – a statistical reaction code with wide application in astrophysics [4, 5, 6, 7]. This code has the advantages that it has explicit knowledge of low lying experimental states, as long as they are experimentally known with its spin and parity value. Furthermore the transmission through the Coulomb potential is treated correctly by confluent hypergeometric functions. The largest disadvantage is, that it only allows for one-particle spallation and does not cover cascade reactions that might occur, if the energies of the neutrinos coming from the proto-neutron star are high enough, or the separation energies are small. Therefore, below the first two-particle emission channel opening, we change the description of the decay probabilities to a model that allows for multi-particle spallation. The code used for this description is the dynamical code ABLA07, that explicitly allows for several particle emissions. Which decay channels will be available for each nucleus is very sensitive to the energy distribution of the neutrinos. We calculated the decay probabilities up to four particle evaporation processes to determine the partial cross section. We made sure that the decay probabilities at the transition between the two models agree and give a smooth transition. The partial cross sections – obtained as a function of incoming neutrino energy – have been folded with a respective supernova neutrino spectrum, parameterized as a Fermi-Dirac function with the parameters neutrino temperature Tν and chemical potential μν . The chemical potential is usually assumed to be zero. This distribution type has been utilized extensively in the past decades, which is the reason why we kept this description also in this work, although different descriptions are available for the neutrino spectra. Changing to one of these alternative descriptions within our model for the neutrino-spectra, such as the α-fit, can be done quite easily. Once our neutrino induced evaporation processes were calculated, we required to implement the neutrino rates in a nuclear reaction network. The utilized nuclear reaction network has so far only been used for the calculation of r-process nucleosynthesis studies including fission cycling [8, 9]. For a better description of the weak interactions close to the proto-neutron star where high densities are available that can influence the electron capture rates, the weak interaction rates from [10] have been included. After generating a working network that accounts for our neutrino-nucleus interactions, we could observe their influence in several nucleosynthesis studies where high neutrino fluxes are expected. These studies were related to various different nucleosynthesis processes assigned to core-collapse supernovae scenarios. These reach from the nucleosynthesis in the neutrino-driven wind to the shock passage through the outer layers of the former star. Within our thesis we have considered different astrophysical parameters for the neutrino-driven wind nucleosynthesis, that actually are based on data from core-collapse supernova simulations. The main input parameter are the neutrino and antineutrino spectra from the cooling of the proto-neutron star, since these will determine the electron fraction. The astrophysical parameters for the neutrino-driven wind of current core simulations are such that the values of entropy and electron fraction are not extreme enough to fulfill the requirements for a successful r-process. Before 2012 it has been shown that the ejecta from the PNS surface are proton rich ( Ye > 0.5) [11]. Due to νp-nucleosynthesis this proton rich matter can contribute to the heavy element nucleosynthesis [48]. Therefore we calculated the nucleosynthesis of this proton rich neutrino-driven wind with the inclusion of our newly determined spallation rates. It turned out that the nucleosynthesis is governed by the antineutrino capture processes on free protons, since the abundance of free protons is high. Furthermore are we able to see, that the partial cross section for neutron emission in neutron deficient nuclei is very small, since proton emissions are more favorable. The second argument against strong changes due to the inclusion of neutrino-nucleus spallation processes is, that the abundance of these heavy elements is substantially lower than the abundance of free protons. Within our study we also posed the question if the observed abundance pattern of r-process poor stars – such as HD 122563 – can be explained with the nucleosynthesis in neutrino-driven wind nucleosynthesis. Therefore we calculated the integrated neutrino-driven wind nucleosynthesis of long-term core-collapse supernovae simulations with different progenitor masses. We were able to observe that with increasing mass of the progenitor, ν p-nucleosynthesis acts more efficiently due to the fact that matter is expanding slower and therefore keeping matter longer in the right conditions for ν p-nucleosynthesis. Thus for an 18 M model an agreement with the metal-poor star’s abundance can be obtained between the elemental abundance pattern and the obtained nucleosynthetic result up to a mass number of roughly Z = 44. The newly included neutrino-spallation rates do not contribute so much that the heavy element nucleosynthesis is changed strongly. Changes can nevertheless be observed in the region of light elements. For example does 7 11the inclusion of neutrinos severely influence the creation of elements as Li and B. However, the total ejected mass of these light elements is too little to contribute to the total nucleosynthesis of the star. In mid of 2012 core-collapse supernova simulations showed that ejecta in the neutrino-driven wind with improved charged-current neutrino nucleon processes, that are treated consistently with the underlying equation of state, can reach values that are initially below 0.5 and are therefore neutron-rich [140]. Later ejecta will become proton rich again, due to the fact that during the cooling of the PNS the spectra become more and more similar. Thus late ejecta will undergo νp-nucleosynthesis, while this is not possible for the early neutron rich ejecta. Within this work we studied the integrated neutrino-driven wind of a current core-collapse supernova model, that includes the above improved treatment, with a progenitor mass of 11.2 M . The electron fraction shows initially neutron rich ejecta ( Ye = 0.474), while later material will again become proton rich. The nucleosynthesis is hereby dominated by the initial neutron rich ejecta, since the mass flux from the PNS decreases very fast. Thus, depending on the mass flux from the proto-neutron star, this will hinder the matter production over the ν p-process. However, the elements usually assigned to the ν p-process – such as 92Mo – are also created in the neutron-rich ejecta, that still produces dominantly neutron deficient material. The neutron rich ejecta show a minor influence on the newly included spallation rates in the nucleosynthesis. In the integrated result these influences are noticeable for the neutron rich stable elements. Large changes due to the inclusion of neutrino-nucleus processes can again be found for the lighter elements, where elements such as 7Li and 11 B are produced due to the inclusion of neutral-current neutrino spallation processes. However, the production factor is too small to be considered as an astrophysical site for the production of these elements. The elemental abundance pattern has been compared to the metal-poor star HD 122563, where it is found that matter can reproduce the abundance pattern up to Z = 42. Heavier elements than Z = 42 could not be produced, due to the neutron shell closure of N = 50. By comparison with the data from the metal-poor star HD 122563, we are only able to gain information on the elemental abundance pattern. As we can see from the isotopic distribution of our nucleosynthesis results in figure 4.17, we see that we overproduce the neutron deficient nuclei in comparison to the solar abundance, whereas neutron rich stable isotopes are underproduced. We are not able to say if this abundance pattern can be also observed in metal-poor stars like HD 122563, because we do not have information on the isotopic abundance. If the abundance pattern would be as it is in our solar system, it would require for another astrophysical site that can explain the creation of the neutron rich elements. The next astrophysical process that we studied in our work was r-process nucleosynthesis in the neutrino-driven wind. Due to the fact that the astrophysical site for r-process nucleosynthesis is still not fully settled and galactic chemical evolution models favor the core-collapse supernova environment as a possible site for the r-process, we utilized a neutrino-driven wind scenario in which an electron fraction of 0.469 at a temperature of T=3 × 109 K was reached. To obtain an r-process that is able to reach mass numbers around A=195, the entropy has been increased by a factor of 2. The luminosities and mean energies for the neutrino spectra have been taken from the original simulation [13]. Since no data has been given for μ- and τ-flavor neutrinos, a variation of these values – within a physical reasonable regime – could be used to observe the influence on neutral-current neutrino-induced particle spallation. To observe only the changes due to the inclusion of neutrino-nucleus processes, we kept for all calculations the rates of free protons and neutrons such, that the electron fraction does not change due to the neglect of these processes. We recognized that the inclusion of charged-current neutrino processes on nuclei do not change the final r-process nucleosynthesis much, whereas neutral-current processes lead to a reduction of the heavy elements produced in this ejecta. Therefore the inclusion of all neutrino rates is dominated by neutral-current processes. The relevant process for neutral-current nucleus processes is the spallation of 4 He, where proton-spallation leads to the reduction of the neutron- to-seed ratio, due to the increased production of heavy elements by subsequent α-captures, starting with the 3H(α, γ)7 Li reaction. We saw that the nucleosynthesis is sensitive to the used neutrino spectra for μ- and τ-flavor, since they have the most influence on neutral-current neutrino processes, as long as the usual energy hierarchy is valid. We should note once more that the astrophysical conditions from state-of-the-art core-collapse simulations favor the production of these light r-process elements (Sr, Y, Zr) over the ν p-processs or slightly neutron rich matter in their wind ejecta, as we have shown above. The question that cannot be explained in current models is, how the abundance pattern of heavy r-process elements (e.g. europium) are created, which are also observed in metal-poor stars such as HD 122563 [36]. Furthermore we studied the hot convective bubble, in which matter is ejected due to neutrino interactions [12, 47]. The matter emission does not occur directly from the proto-neutron star surface, instead it is further away in a region of a few hundred kilometers. The electron-fraction of these ejecta resides between 0.5 and 0.54, where early ejecta have values close to 0.5 and later ejecta become more proton rich. Note that these simulations do not consider the improved description of microphysics in the proto-neutron star. For these ejecta the inclusion of neutrino-nucleus processes do not change the heavy element production tremendously. The only fact that we could observe is that some production of heavy elements over the ν p-process could be made, which however is also very minimal. In a next step we focused on the nucleosynthesis in the outer shells of the exploding star, where the neutrinos, produced by the proto-neutron star, traverse through. The neutrino fluxes are of course less than in the case of the neutrino-driven wind – due to the radial decrease with 1/r 2 . As already shown in previous works [2, 3] the inclusion of neutrino interactions in this region does have influence on the odd-Z elements in the nucleosynthesis. The reason can be found in the elemental abundance ratio, where odd-Z elements show always less abundance than the even-Z nuclei. As it has been noted by [158, 159], if the ratio of the mother nucleus to the daughter nucleus is of the order 103, a production over ν-nucleosynthesis might be possible. Specifically for the light elements 7 Li and 11B the production in this explosive environment has shown to be an important part for the amount of these elements in our solar system. Within this work we utilized a rather phenomenological model for the shock passage and the expansion afterwards. Furthermore the neutrino luminosities have been assumed to be very high, especially for μ and τ flavor neutrinos, which are the dominating contributions for neutral-current processes.10 We used this high values to be able to compare with [2]. Due to this high values it is understandable that the dominating processes in ν -nucleosynthesis are neutral-current processes. However, charged-current processes also contribute and cannot be neglected. The main difference why in this astrophysical process neutrino-nucleus processes show prominent features, whereas in the case of the neutrino-driven wind the difference is barely noticeable, can be explained as follows. The argument is that – contrary to the neutrino- driven wind – in the ν -process the matter is already produced before the expansion starts. Thus the time window for neutrino interaction is longer, than in the neutrino-driven wind. In the neutrino-driven wind heavy elements are created by (n, p) reactions that mimic β + -decays, whereas in ν -nucleosynthesis the elements are created by spallation of the abundant nuclei. The abundance of the elements acting as a target for neutrino-nucleus interactions is in the outer shells substantially higher in comparison to the same elements in the neutrino-driven wind. Thus the relevant quantity of abundance times cross section will be generally higher, which allows converting more matter over spallation reactions in the ν -process. Globally we can say that the inclusion of neutrino-nucleus processes does not have influence on the nucleosynthesis of core-collapse supernovae. For an efficient production of some elements such as 7Li, 9Be, 11 B, but also 138La and 180Ta, neutrino-nucleus interactions are required. However this fact has already been mentioned in [2]. For r-process calcula- tions we can observe that the inclusion of neutrino-processes changes the final abundance. Thus, if the r-process would occur in neutrino-driven winds, at least the neutrino-nucleus interactions on 4He need to be incorporated. However, a full treatment will be always better. Within proton-rich winds on the other hand, the effect of neutrino-nucleus interac- tions on heavy element nucleosynthesis is very small. However, the fact that we now have a full set of rates, allows us to explore the impact of various astrophysical sites and their nucleosynthesis processes in a more sophisticated model, which includes the possible relevant neutrino-nucleus processes in these environments.

Frage zum Eintrag

Redaktionelle Details anzeigen