The impact of electron-capture (EC) cross sections on neutron-rich nuclei on the dynamics of corecollapse during infall and early post-bounce is studied performing spherically symmetric simulations in general relativity using a multigroup scheme for neutrino transport and full nuclear distributions in extended nuclear statistical equilibrium models. We thereby vary the prescription for EC rates on individual nuclei, the nuclear interaction for the EoS, the mass model for the nuclear statistical equilibrium distribution and the progenitor model. In agreement with previous works, we show that the individual EC rates are the most important source of uncertainty in the simulations, while the other inputs only marginally influence the results. A recently proposed analytic formula to extrapolate microscopic results on stable nuclei for EC rates to the high densities and temperatures and the neutron rich region, with a functional form motivated by nuclear-structure data and parameters fitted from large scale shell model calculations, is shown to lead to a sizable (16%) reduction of the electron fraction at bounce compared to more primitive prescriptions for the rates, leading to smaller inner core masses and slower shock propagation. We show that the EC process involves ≈ 130 different nuclear species around 86 Kr mainly in the N = 50 shell closure region, and establish a list of the most important nuclei to be studied in order to constrain the global rates.

We perform simulations of the Kelvin-Helmholtz cooling phase of proto-neutron stars with a new numerical code in spherical symmetry and using the quasi-static approximation. We use for the first time the full set of charged-current neutrino-nucleon reactions, including neutron decay and modified Urca processes, together with the energy-dependent numerical representation for the inclusion of nuclear correlations with random-phase approximation. Moreover, convective motions are taken into account within the mixing-length theory. As we vary the assumptions for computing neutrino-nucleon reaction rates, we show that the dominant effect on the cooling timescale, neutrino signal and composition of the neutrino-driven wind comes from the inclusion of convective motion. Computation of nuclear correlations within the random phase approximation, as compared to mean field approach, has a relatively small impact.