Impact of in situ nuclear networks and atomic opacities on neutron star merger ejecta dynamics, nucleosynthesis, and kilonovae

Modeling binary neutron star merger (BNSM) ejecta evolution requires simulations involving hydrodynamics, nuclear reactions, and radiative processes. The impact of nuclear burning and atomic opacity is poorly understood and often treated with simplified prescriptions. We systematically investigate different treatments of nuclear heating, thermalization, and opacities in radiation-hydrodynamics simulations of BNSM ejecta and kilonova light curves. Ejecta from long-term numerical-relativity simulations are evolved to ~30 days using a 2D ray-by-ray approach. We compare simplified heating-rates, thermalization prescriptions, and gray opacities with in-situ nuclear networks (NN) that track energy deposition, and include a composition-dependent thermalization scheme and frequency-dependent, atomic-physics-based opacities. Coupling NN and hydrodynamics affects nucleosynthesis and kilonova emission. Assuming homologous expansion alters the abundance evolution and produces a narrower second $r$-process peak and a third peak shifted to higher mass numbers. Nuclear heating back-reaction delays and reddens the early emission. A constant thermalization underestimates the early luminosity and overestimates the late emission. Analytical opacities yield dimmer and redder kilonovae at early times ($t\lesssim$ hour) and a prolonged emission at $t\gtrsim5$ days. Resolving the first hundreds of milliseconds of hydrodynamics is essential for robust nucleosynthesis calculations, and composition-dependent thermalization and frequency-dependent, atomic opacities are needed to accurately capture the ejecta temperature and kilonova brightness and color evolution. Analytic nuclear-power fits with simplified thermalization and opacities can reproduce the density and temperature evolution of the ejecta. [Abridged].