Computational power is an issue when it comes to simulating complex phenomena like neutron star mergers in three dimensions. Hence, it is important to strive for novel numerical methods that allow to capture the violent dynamics of a neutron star collision and resolve the different spatial scales, on which different processes take place. A promising way forward are moving-mesh tools, where the computational grid moves with the fluid flow and thus the numerical resolution is adaptively placed where it is needed. Apart from these obvious advantages, such schemes also promise to be less prone to numerical artifacts because the advection across cell bounadries is minimized. Within a long-term development we have adapted and extended the moving-mesh code AREPO for simulations of relativistic systems. Our paper in MNRAS presents the first simulations of neutron star mergers with a moving-mesh code. Developing such a tool was one of the major goals of the GreatMoves project.

Hyperons are baryons that contain at least one strange quark. But they are not only strange, they are puzzling as well. The "hyperon puzzle " raised the question whether these articles should occur in the center of neutron stars. If so one could generally expect a softening of the neutron star equation of state. This however seems somewhat in tension with the observation of neutron stars with masses above two solar masses, which requires a certain stiffness of the high-density equation of state. In collaboration with colleagues from Barcelona we recently presented a comprehensive study of hyperons in neutron star mergers. We did not solve the hyperon puzzle, but we outlined possibilities to potentially identify the presence of hyperons in neutron stars by observing gravitational waves of the postmerger stage. If hyperons occur, the thermal pressure for a given amount of thermal energy is reduced. As consequence the merger remnant is more compact and gravitational wave frequencies slightly increased, which may then point to the presence of hyperons.

A kilonova is the electromagnetic emission produced in the rapidly expanding explosion following a neutron star merger. What a telescope observes as kilonova many million light years away is the result of numerous atomic emission and absorption processes. Understanding the kilonova by modeling the emission requires to simulate all these atomic processes taking into account the properties of all different chemical elements in the expanding cloud. Shingles et al., ApJL 954, L41 presents such a simulation of radiation in three spatial dimensions. It includes the vast number of more than 40 million atomic transitions. By this detailed approach it is possible to trace back which chemical elements shape which part of the observed spectrum. For instance, the chemical elements strontium, yttrium and cerium seem to be quite important. It is remarkable that model reproduces prominent features of the kilonova AT2017gfo without fine-tuning the model to this particular observation. The calculation also highlights the important of three-dimensional kilonova models and makes predictions how kilonovae should look like from different observer angles.

Neutron star mergers produce heavy elements like gold and uranium through the rapid neutron-capture process. In fact, they are to date the only confirmed source where this process takes place. The nucleosynthesis occurs in the material ejected from the merger site; obviously only gravitationally unbound matter can be mixed into the interstellar medium and form new stars and planets. The ejection process during a neutron star coalescence is rather complex occuring over different time scales with different ejection mechanisms being involved. Simulations of neutron star mergers typically do not cover all mass ejection episodes, which obviously yields an only incomplete picture of the nucleosynthetic yields. In ApJL 951, L12 (2023) we have consistently included all mass ejection processes and computed the resulting element production and kilonova properties. For the specific models considered here, we observe that in comparison to a solar elemental distribution too little heavier elements are produced. This may suggest that this particular binary configuration may not be the most representative one and may only contribute a subdominant fraction to the total abundance (see left panel). The properties of the resulting kilonova are in remarkable agreement with the observed event in 2017 (see right panel).

The kilonova AT2017gfo associated with GW170817, the first neutron star merger detected by gravitational waves, has been a treasure trove for Nuclear Astrophysics. A kilonova emerges from the electromagnetic emission of the hot explosion of the material that is escaping the merger site. In these outflows heavy elements are formed and the nucleosynthesis process creating these elements heats the matter producing the electromagnetic emission. In a recent study, which was lead by colleagues from the Niels Bohr Institute, we investigated the geometry of the explosion. It came as a big surprise that the outflow is very spherical. This is a priori not obvious considering the large amount of angular momentum in the system, which stems from the orbital motion of the neutron stars. Hence, naively one would expect that the outflow features a pronounced symmetry axis instead of having the geomerty of a sphere. This observation is not only helpful to better understand kilonovae, but it may also be used as a new probe for cosmology to determine the Hubble constant. The analysis recently appeared in Nature. (See also the press release and our popular article in Sterne und Weltraum.)