| Popis: |
Mass-accreting electron-degenerate stellar cores that are composed primarily of the carbon-burning ashes 16-O and 20-Ne (ONe cores) appear in several astrophysical scenarios. On the one hand, they can be formed during the late evolution of intermediate-mass stars with 8 to 10 solar masses. On the other hand, they can occur in the context of the accretion-induced collapse (AIC) of ONe white dwarfs, where the collapse is induced by mass transfer from a companion star in a binary system or due to cooling of the outer layers of a white dwarf-white dwarf merger remnant. Their evolution is critically depending on electron capture (EC) reactions on nuclei with mass number A ~ 20 that become relevant above a density of 10^9 g/cm^3. Besides removing electrons (the main pressure support) from the core, EC are also responsible for releasing or absorbing heat. In the canonical picture, the accreting ONe core will undergo compression until EC on the abundant 20-Ne are activated. As a consequence, the core becomes gravitationally unstable. Simultaneously, ECs release sufficient heat to ignite oxygen in a thermonuclear runaway, launching an outwards traveling deflagration wave. Nevertheless, the energy release from oxygen fusion is insufficient to halt gravitational collapse. Otherwise, the star would be destroyed by a thermonuclear explosion. Subsequent to collapse, a neutron star is formed and the stellar envelope explodes in an electron-capture supernova (ECSN). Recent 3D hydrodynamic simulations of the oxygen deflagration in ONe cores have suggested that the outcome of such events (either ECSNe or thermonuclear explosions) depends critically on the ignition density of oxygen. Depending on the treatment of convection, the ignition density is estimated to be ~ 2x10^10 g/cm^3 in case core convection sets in prior to ignition and ~ 10^10 g/cm^3, if not. It has been suggested that models corresponding to the first case lead to a collapse, while models corresponding to the second case result in a thermonuclear explosion of the star. We study accreting ONe cores in the AIC scenario, focusing on open questions regarding the evolution of the core prior to ignition. By including the secondary carbon-burning products 23-Na and 25-Mg in the initial models, new insights can be gained concerning Urca cooling. While it seems well established that EC processes do not trigger convection in the ONe core, the poorly understood phenomenon of overstable convection could alter this picture and will be assessed by us, in detail. Furthermore, modifications at high densities to the standard set of nuclear reactions, responsible for neon and oxygen burning, are investigated. Previously, reaction channels that become possible due to the presence of 20-O, formed by double-EC on 20-Ne, have not been considered. Neon burning is modified by the reaction 20-O(a,g)24-Ne and oxygen burning can additionally proceed by the fusion involving neutron-rich oxygen isotopes: 20-0 + 16/20-O -> 36/40-S*. In neither case experimental data is available. Further investigations are dedicated to exploring the origin and the consequences of an off-center ignition of the flame, due to EC processes. For our simulations, we make use of the ``Modules for Experiments in Stellar Astrophysics'' stellar evolution code (MESA). In order to determine the EC and beta(-) decay rates, we utilize the recently implemented capability of MESA to evaluate weak reaction rates with very high accuracy by solving the phase-space integral directly, only requiring matrix elements and excitation energies of all contributing transitions, either known experimentally or originating from shell-model calculations. Additionally, it is our aim to study the oxygen deflagration that is potentially followed by an ECSN. This is done by combining the 1D shock-capturing core-collapse-supernova code AGILE-IDSA with a level-set-based flame description, using laminar and turbulent flame velocities based on microscopic calculations. All relevant weak processes on the oxygen-burning ashes are considered, in order to correctly predict the deleptonization and energy generation. We confirm the accuracy of the direct determination of weak rates in MESA and extend it to the Urca nuclei 23-Na and 25-Mg. The MESA models of accreting ONe cores show that especially the abundance of 25-Mg and the related Urca cooling can affect the ignition density by 10%, resulting in values between 8.7 - 9.7 x 10^9 g/cm^3. Additionally, we investigate the previously reported off-center ignition (~ 50 km) of the oxygen deflagration, caused by including the second-forbidden transition between the 20-Ne and 20-F ground states. We conclude that in this case, EC heating on 20-Ne acts on much longer timescales (~ years), giving the core sufficient time to expand and shift the ignition away from the center. Furthermore, we report that overstable convection, treated as a diffusive process, does not affect the evolution of the ONe core. However, applying Kato's linear growth analysis suggests that instabilities could grow on a timescale of around 10-100 s. This would give enough time for instabilities to develop, as the timescale between the onset of semiconvection in the core, due to EC on 20-Ne, and collapse is around 100 years. Also, we find that including a larger reaction network, together with the aforementioned modified set of reactions during neon and oxygen burning, has no impact on the evolution of the ONe core, at least if the burning is initiated by a thermonuclear runaway. Using spatially high-resolution ONe core models that develop a thermonuclear runaway in the center, we show preliminary results of the oxygen deflagration, simulated with AGILE-IDSA. We further demonstrate the capability of AGILE-IDSA to perform self-consistent ECSN simulations with Nomoto's canonical progenitor model. We want to point out that this approach can efficiently complement expensive 3D simulations by performing parameter studies, allowing for a better targeted use of computer resources in 3D simulations. |
