Constraining P-process nucleosynthesis through crucial cross-section measurements

This thesis navigates the complexities of the p-process, crucial in understanding the formation of stable, proton-rich nuclei beyond iron, commonly created in explosive stellar events. The modeling of this intricate process relies on theoretical calculations of nuclear reaction rates, involving reaction cross-sections. The primary focus of this thesis will be on measuring relevant reaction cross-sections, aiming to facilitate the reconciliation of existing disparities between theoretical calculations and experimental observations. Five cross-section measurements of (p,γ) and (p,n) reactions have been conducted using the activation method on germanium isotopes, specifically ^70,72,73,76Ge, at 2.5 MeV using the Tandetron of the Horia Hulubei National Institute. The main objective was to address discrepancies noticed in the abundance of ⁷⁴Se, the lightest p-nucleus. The analysis provided the first experimental data for two reaction cross sections: ⁷³Ge(p,γ)⁷⁴As and ⁷³Ge(p,n)⁷³As. Furthermore, a new measurement of the ⁷²Ge(p,γ)⁷³As reaction cross section revealed a notable discrepancy of approximately 38% when compared to another study that employed the in-beam γ-summing technique. Additionally, the dissertation articulates plans for an upcoming α-elastic scattering experiment at IJCLab Orsay using the SplitPole magnetic spectrometer. The primary objective is to refine our comprehension of the poorly known α-nucleus interactions, fundamental in governing the reactions dictating the abundance patterns. This experiment will focus on elucidating the mechanisms behind the formation of high-mass p-nuclei beyond the mass range of A > 140. Finally, Machine Learning concepts are introduced in order to bridge the gap between experimental data and the refinement of nuclear models. Bayesian and Frequentist approaches are explored to disentangle the existing correlations among various inputs affecting reaction cross-section simulations. (author)