Towards a more in-depth understanding of Neutron Star Mergers: nucleosynthesis and microphysics modeling

Binary neutron star (BNS) mergers are powerful events of particular interests for modern astrophysics given their intrinsic multi-messenger nature. In addition to gravitational waves (GWs), whose first detection for these systems dates back to 2017, Neutron Star Mergers can also emit both quasi- and non-thermal electromagnetic radiation, as well as neutrino bursts. The possibility of exploiting different kinds of cosmic messengers to make inferences of both physical and astrophysical nature makes these events a very captivating target of study. The scientific potential of BNS mergers is in fact multidisciplinary, as it ranges between a wide variety of different research areas. For example, BNS mergers can be used as unique dense matter laboratories, exploited to probe the (currently unknown) properties of nuclear matter at supranuclear densities. Neutron Star Mergers are also object of Galactic Chemical Evolution (GCE) studies, as they influence the galactic environment in which they occur. In fact, during the coalescence a fraction of the total mass of the system is ejected at moderately relativistic speed towards the outer space. This ensemble of matter, which is typically referred to as the `ejecta', synthesizes new elements during the expansion, which then mix within the interstellar medium on long timescales. Due to the very neutron richness that characterizes the ejected matter at the onset of the expansion, BNS mergers host suitable conditions for the r-process nucleosynthesis to take place. The joint GW-kilonova detection in 2017 currently poses BNS mergers as the only experimentally confirmed astrophysical event able to enrich the galactic environments with heavy r-process nuclides. The radioactive decays of the nuclides synthesized within the ejecta powers a quasi-thermal electromagnetic transient lasting for several weeks after the merger, named as kilonova. The analysis of the observed kilonova signals, together with the measurements of abundance patterns in metal-poor environments, can help to constrain the role of BNS mergers in the GCE. However, for the time being this information remains still elusive due to the rarity of BNS merger events and the large uncertainties in the corresponding theoretical nucleosynthesis models. Finally, an item of interest in its own right is the application of Neutron Star Mergers to fundamental physics studies, such as General-Relativity (GR) validity tests, or also to cosmological inferences, as their multi-messenger nature allows to estimate the cosmic expansion rate, H_0, when the host galaxy is identified. In order to fully exploit the information encoded into the multi-messenger signals and avoid systematical biases, our understanding regarding the dynamics of the system must be complete and robust. Unfortunately, the conditions that characterize BNS mergers are too extreme to be reproduced in any terrestrial laboratory, therefore we have to resort uniquely to theoretical models, that are constructed on the basis of Numerical Relativity (NR) simulations. Ideally, such models should include all the various physical effects that may be relevant for the BNS dynamics. The interplay between these effects and the wide range of spatial scales they are active on, which should be all resolved by the simulation, makes the numerical modeling of Neutron Star Merger systems very demanding from the computational point of view. This leads to the introduction of (sometimes severe) approximations in the implementation of the various effects, e.g. the microphysics framework, which is of particular interest in the context of this thesis. Additional uncertainty sources are associated to the only partial knowledge of the underlying physics for some effects, as in the case of the Equation of State (EOS) describing the neutron star fluid, or the complete omission of potentially important degrees of freedom (DOFs), e.g. in terms of the particle content. For the latter, an example is represented by muons, which recent studies found to be possibly impactful in the merger aftermath. All these aspects limit the precision and the realism in the description of the system. One of the main physical ingredients that simulations should account for are neutrinos. Neutrinos play a crucial role in the dynamics of BNS coalescences as they can affect the properties and the stability of the remnant by redistributing energy and momentum while they diffuse, possibly influencing the emission of GWs in the merger aftermath. Moreover, neutrino-matter interactions set the neutron-to-proton content of ejected matter, with direct consequences on the nucleosynthesis yields and on the associated kilonova signal. In particular, neutrino irradiation from the central remnant leads to ejecta leptonization, boosting the production of light elements in the ejected matter. This effect is proportional to the lifetime of the central object, therefore it is particularly relevant when binaries produce a long-lived remnant after the merger. As a consequence, in some cases the distribution of heavy $r$-process element in the ejecta can be complemented by the presence a non-negligible amount of lighter nuclides. Despite this possibility, the production of light elements in the context of BNS mergers has not been particularly investigated in literature so far. Understanding quantitatively how and when such elements are formed in this context could help us to expand and strengthen our understanding about BNS mergers and the associated signals in a complementary way to heavy r-process nuclides. Because of their relevance, the description of neutrinos in BNS mergers must be accurate. In numerical simulations neutrinos are evolved through the Boltzmann equation, which is approximately solved using appositely designed transport schemes. Recently, many efforts have been devoted at improving the precision and reliability of the employed transport schemes. Along with a robust transport scheme, feeding accurate neutrino-matter interaction rates into the source term of the transport equation is of paramount importance for modeling the correct neutrino dynamics. Nonetheless, the characterization of neutrino interactions in BNS merger simulations is typically oversimplified and incomplete. In fact, neutrino rates are usually implemented adopting too simplistic approximations and important microphysics effects are not accounted for. Moreover, the set of reactions that are considered is often not exhaustive, as there are processes whose impact in BNS merger conditions could be significant, but have not been explored yet in detail. As an example, the inelastic scattering of neutrinos off electrons and positrons (proven to be relevant for the dynamics of CCSNe and not usually included in BNS merger simulations), has been found by recent studies to be possibly impactful in the merger aftermath. My PhD thesis tries to address part of the problems discussed above by presenting original results and newly developed tools. In particular, the aim of the thesis is twofold. On one hand it expands the current understanding of BNS mergers, by exploring an aspect that has not been particularly investigated in literature so far, i.e. the nucleosynthesis of light elements in the ejecta. On the other, it contributes to increase the physical realism of the microphysics framework that is usually implemented into numerical simulations by providing new tools that improve the EOS and neutrino rates description. The thesis is organized with the following structure. Chapter 2 introduces the matter by providing a general overview of the Neutron Star Merger dynamics, focusing in particular on the ejection of matter from the system and the associated nucleosynthesis, as well as the role of neutrinos in the coalescence. In Chapter 3, I detail the general methods employed for the analysis, such as the setup of the General-Relativistic Hydrodynamics simulations, the nucleosynthesis calculations and the way in which neutrino rates fit within the formalism of gray M1 transport schemes. The next chapters are then intended to present the original results of the thesis, which I summarize in the following. Chapter 4 focuses on the production of light elements in the BNS ejecta, following the indication from recent works featuring state-of-the-art BNS simulations that the electron fraction distribution can feature a significant tail around 0.4 and beyond that. For such high Ye values, the final nucleosynthesis yields are dominated by nuclides that are lighter than the ones in the three r-process peaks. These elements are good candidates to be studied for expanding our understanding of BNS mergers, as they are easier to be identified in kilonova spectra and they show a much larger abundance variability in relation with the binary parameters. Motivated by this aspect, the production of light elements in BNS ejecta is quantitatively studied, extending the analysis presented by Perego et al. (2022) to elements with atomic number (Z) smaller than 38. First, employing SkyNet as a full nuclear reaction network, the typical thermodynamic conditions that favor the production of these elements are discussed. Then, the actual ejecta yields are computed by combining the network data with the ejecta properties extracted from numerical BNS simulations targeting GW170817-like binaries. With the exception of Hydrogen and Helium, the findings show that the abundances for Z smaller than 20 are severely suppressed, while elements between Calcium and the first r-process peak are way more synthesized. The highest abundances in this range are around 0.001 and are found in the dynamical ejecta of (soft) equal-mass BNS mergers; for which the shock-heated, leptonized component of the ejecta is more prevalent. Neutrino irradiation on the spiral-wave wind enhances even more the nucleosynthesis of light nuclides at the expense of the heavier ones. The enhanced production of Iron-group elements by Neutron Star Mergers may also question the proposed origin of some live radioisotopes detected on Earth. This could be the case for the recent detection of 60Fe and 244Pu in a deep-sea crust sample dating back to the past 3–4 Myr. Since 60Fe is usually attributed to standard core-collapse supernovae, previous works concluded that a single BNS merger cannot explain the 60Fe-over-244Pu ratio measured in the crust. In the last part of Chapter 4 this scenario is revived on the basis of nucleosynthesis models computed for a set of numerical BNS simulations producing long-lived massive remnants and expelling spiral-wave wind ejecta over a 100–200 ms timescale. In fact, assuming an inefficient large-scale mixing, the mid-high latitude ejecta feature an isotopic ratio compatible with observations when a merger happening 80–150 pc away from the Earth and between 3.5 and 4.5 Myr ago is considered. Expected isotopic ratios for additional live radioactive nuclides are then presented and compared with other proposed scenarios to allow to discriminate among them once additional detections will be available. Chapter 5 and 6 instead have a different focus, as they present newly developed tools aimed at improving the accuracy of the microphysics modeling in numerical BNS simulations. In particular, Chapter 5 discusses the strategy to be followed to properly include muons in the BNS merger description, focusing specifically on the modification of the EOS. To this end, (anti)muons are described as an ideal Fermi gas of massive particles and the analytical expressions describing their contribution to the total EOS quantities are explicitely reported. These formulas have been then included into a C++ class designed to implement the contribution of massive leptons into a general-purpose multi-species EOS, starting from a tabulated baryonic table. Different kinds of implementations have been explored and tested for such a purpose, targeting different levels of accuracy and numerical performance. Chapter 6 is dedicated to the presentation of 'BNS_NURATES', a novel open-source numerical library designed for the efficient on-the-fly computation of neutrino interactions, with particular focus on regimes relevant to BNS mergers. The library targets an higher level of accuracy and realism in the implementation of neutrino rates by accounting for relevant microphysics effects on the interactions, such as weak magnetism and mean field effects, and by including the contributions of some relevant reactions that are usually omitted, like inelastic $\nu e^\pm$ scattering and (inverse) nucleon decays. As a first application case, energy-dependent and energy-integrated neutrino emissivities and opacities are computed for conditions extracted from a BNS merger simulation with M1 transport scheme. A qualitative difference in the results that is ascribable to the level of sophistication of the rates is observed, proving that an improved treatment of neutrino-matter interactions in BNS mergers is necessary to properly model the neutrino dynamics. In particular, neutrino scattering off electrons/positrons proves to be important for the energy exchange of heavy-type neutrinos, as they do not undergo beta processes when muons are not accounted for. Moreover, microphysics effects can significantly modify the contribution of beta processes for electron-type (anti)neutrinos, increasing at the same time the importance of (inverse) neutron decays. The improved treatment also modify the conditions at which neutrinos decouple from matter in the system, potentially affecting their emission spectra. All the main results and findings presented in the thesis are finally summarized in Chapter 7, where caveats and further possible expansions are also discussed.