Résumé : One-nucleon halo nuclei are exotic nuclei which can be seen as a core around which orbits a loosely-bound valence nucleon. They are usually studied through reactions such as elastic scattering and breakup. The ratio method has been developed as a tool to study one-neutron halo nuclei at high energies. It consists of the ratio of angular cross sections, breakup and elastic scattering, which removes most of the sensitivity to the reaction mechanism and to the reaction model. In the simple recoil excitation and breakup (REB) model, the ratio simplifies to a form factor dependent solely on the wave function of the projectile. By measuring this observable and comparing it to the REB form factor, i.e. in the ratio method, more detailed information on the structure of the halo could be obtained. For neutron-halo nuclei at high energy, the ratio observable obtained from accurate CDCC and DEA theoretical calculations follows its REB prediction. I study the extension of this method to lower energies of the reaction which could make the measurement appropriate to facilities such as SPIRAL2 (GANIL, Caen, France) and ReA12 at FRIB (Michigan State University) and to proton halos. This is done by comparing the REB form factor to dynamical calculations of the ratio. The reactions investigated are the reaction of 11Be, the archetypical one-neutron halo nucleus, on 12C, 40Ca and 208Pb targets at 20 MeV/nucleon and of 8B, the archetypical one-proton halo nucleus, on 12C, 58Ni and 208Pb targets at44 MeV/nucleon.For these reactions, the adiabatic assumption is no longer valid due to the effect of the Coulomb interaction. This effect is mainly visible at forward angle for 11Be and is aggravated for 8B by the fact that the halo is charged. The ratio works less well than for neutron-halos at intermediate and high energies. Nevertheless, the ratio is shown to be very sensitive to the orbital angular momentum l0 in which the halo is bound and its binding energy E0, i.e. the single-particle structure of the projectile. Variations of l0 and E0 induce visible changes in shape and in magnitude (up to several orders) of the ratio. Also, the agreement of the ratio with its REB prediction is best when the projectile is loosely-bound and for low l0, i.e. for s and p waves. The validity of the method is not affected by the use of energy ranges—or bins— in the projectile continuum. These tend to increase the cross section without changing the agreement of the ratio with its REB prediction. The applicability of the method is finally explored at high energy for proton-rich nuclei 17F, 25Al and 27P. I show that the ratio method works the latter since this nucleus is bound by a mere 0.870 MeV in the s-wave. For the other nuclei, although the agreement of the ratio with its REB prediction is less good than for neutron-halo nuclei at high energy, it still provides estimates of nuclear-structure features, such as l0 and E0 and could be applied in what can be called an approximate application of the ratio method. Heavy nuclei exhibit a neutron skin, i.e. a thin layer around the nucleus where only neutrons are found. The thickness of the skin is highly correlated with the slope of the symmetry energy. The process of coherent neutral-pion photoproduction is used to extract the nuclear density and hence the neutron-skin thickness of heavy nuclei. In order to analyse recent data on the photoproduction on 12C, 40,48Ca, 116, 120, 124Sn and 208Pb, I build a reaction code. My model uses the formalism of Kerman, McManus and Thaler (KMT) which allows to build the photoproduction matrix on a nucleus from the ones describing the elementary process on a single nucleon. Within the impulse approximation, the photoproduction is seen as the coherent sum of the photoproduction on each of the nucleons. In the plane wave impulse approximation (PWIA), no rescattering of the pion is considered after its production and the cross section is directly proportional to the Fourier transform of the density. Such process is taken into account at the distorted wave impulse approximation (DWIA) by considering a potential simulating the pion-nucleus interaction and built from the KMT formalism.The agreement of my model with the data is good, especially for 208Pb. The distortion has a significant impact on the photoproduction process. The sensitivity of the process to the density of the target is analysed by performing the calculations with several different densities calculated in different structure models. The distortion has the effect of deteriorating this sensitivity. In the particular case of a 208Pb target, the impact of variations of the neutron-skin thickness of around 0.1 fm on the photoproduction cross section is ten times smaller than the size of the error bars on the experimental data. These results, although less dramatic, hold for the tin targets, for which preliminary data exists. In the light of these results, the coherent neutral-pion photoproduction process does not seem to be suited in the study of the neutron-skin thickness. This conclusion goes in contrast to the results of recent measurements on 208Pb, for which the method was shown to be sensitive to fine details of the density.