Thèse de doctorat
Résumé : In the mid-eighties, the development of radioactive-ion beams enabled the exploration ofregions of the nuclear landscape away from the valley of stability. Close to the neutrondripline, in the light neutron-rich region, halo nuclei were observed. These nuclei exhibit asurprisingly large matter radius and a strongly clusterized structure. These two featurescan be explained by the weak binding of one or two neutrons which allows them to tunnelfar from the rest of the nucleons, surrounding the nucleus by a diffuse halo. These nuclearstructures have challenged the usual description of the nucleus, described as a compactmany-body object with nucleons piling up into well defined orbitals. Because they areshort-lived, these nuclei are often studied through reaction processes, such as elasticscattering, breakup and knockout. To infer precise information from the experimentaldata, an accurate reaction model coupled with a realistic description of the nucleus isneeded.Compared to other state-of-the-art methods, the eikonal approximation is very cheapfrom a computational viewpoint. This model assumes that the projectile-target relativemotion does not differ much from the initial plane wave. It also makes the adiabaticapproximation, which sees the internal coordinates of the projectile as frozen during thecollision. These two assumptions hold for reactions occurring at high energy, i.e., above60 MeV/nucleon, in which the deflection of the projectile by the target is small and thecollision time is brief.In this thesis, I focus on improvements of the eikonal approximation. First, I studythe extension of the validity of the eikonal model down to 10 MeV/nucleon, in the energyrange of the facilities HIE-ISOLDE at CERN and ReA12 at the upcoming FRIB. To thisend, I analyse different corrections to the eikonal approximation, which account for thedeflection of the projectile by the target. I assess their accuracy for the elastic-scatteringand breakup observables of one-neutron halo nuclei at 10 MeV/nucleon. Next, I developa dynamical correction to the eikonal approximation, which applies to both nuclear andCoulomb interactions while conserving the eikonal numerical cost. I study this correctionin the cases of breakup reactions of one-neutron halo nuclei on light and heavy targets.Then, I investigate which nuclear-structure information can be inferred from knockoutreactions of one-neutron halo nuclei. To do so, I conduct a sensitivity analysis of theirobservables to the nuclear structure of the projectile, described within a halo effectivefield theory. In particular, I study the influence onto the cross sections of the ground-statewave function, the presence of subthreshold bound states and resonances.