Single-particle and collective properties around closed shells probed by in-source laser spectroscopy

Resonant laser ionisation is a very versatile tool in nuclear physics, used for the production of clean radioactive ion beams as well as for the study of ground-state and isomer properties. In this Ph.D. work, many aspects of resonant laser ionisation are investigated, from improving the performance of laser ion sources at ISOL facilities to the measurement of magnetic dipole moments and charge radii. The LISOL gas catcher ion source relies on resonant laser ionisation for increased efficiency and selectivity. Using a $^{252}$Cf fission source, the element dependence of the non-resonant contribution to the ion beam has been investigated. The efficiency of extraction for a non-laser-ionised element ranges from 0.03% for krypton to 74% for ceasium. A relationship with the ionisation potential is proposed, although a few elements like rubidium and cerium do not verify this relationship. In order to suppress those non-resonantly-ionised elements, two new approaches are proposed. First, the dual-chamber gas catcher is investigated. This gas catcher is separated in two volumes, one for the stopping and neutralising of the nuclear recoils and one for the laser ionisation. Both volumes are joined by a channel but no direct line of sight is possibe from one to the other. The reduced density of charges in the second volume increases the chances of survival of laser ions and also permits the use of DC electical fields inside the gas catcher to collect ions surviving the neutralisation process. The gas catcher has been characterised in off-line conditions with nickel and on-line conditions with rhodium. The ion collector has been found to perform as expected. However, another source of non-resonant ionisation has been identified in the form of decay of implanted activity on the surface of the gas catcher and of the SPIG rods, producing singly- and doubly-charged ions. A second approach that has been investigated is that of the laser ion source trap (LIST) coupled to a gas catcher. By applying an electrical potential between the SPIG and the gas catcher, it is possible to repel any ion coming from the gas cell and get a pure atom beam into the SPIG. The lasers are then overlapped with this beam in order to ionise the element of interest. The LIST mode has been achieved both off-line and on-line, although various restrictions in geometrical overlap and duty factor do not allow for the use of this technique for efficient radioactive ion beam production at the LISOL facility. The same effect of decay from the surface of the SPIG rods is responsible for non-resonant ion contaminants. A thorough study of the atomic transition linewidth has been performed inside the gas cell and in the super sonic jet using the list approach in order to determine in which limits in-source laser spectroscopy in a gas catcher is possible. It concluded that the conditions in a gas catcher, in spite of the pressure broadening, are more favorable than in a hot-cavity ISOL facility. The study of the isotope shift of the stable nickel isotopes showed however that no information on the changes in the mean-square charge radii can be extracted for the light and medium heavy nuclei. The hyperfine structure of the copper isotopes, however, is very large and can be resolved with in-gas-cell laser spectroscopy. The magnetic dipole moment of the neutron-deficient copper isotope $^{57}$Cu is a key parameter in challenging nuclear models as it should be determined by the outermost proton outside the N = Z = 28 closed core $^{56}$Ni. This measurement could not be performed in a hot-cavity ISOL facility as this short-lived isotope (T$_{1/2}$ = 199 ms) decays before it can diffuse out of the target matrix. A new value for the magnetic dipole moment of $\mu = +2.582(7)_{\mu N}$ is proposed, in disagreement with the previous $\beta$-NMR value but in good agreement with all presently available calculations. The measurement has been repeated 68 times to ensure its accuracy and systematic effects on the stable isotope $^{63}$Cu are thoroughly discussed. A more precise value for the magnetic moment of $^{58}$Cu is also proposed to be $\mu = +0.479(13)_{\mu N}$, consistent with a $\pi$ p$_{3/2} \otimes$ νp$_{3/2}$ configuration. Similarly as for the nickel isotopes, no information could be extracted from the isotope shifts. At the CERN ISOLDE facility, the polonium isotopes have been studied. Three laser ionisation schemes have been characterised and the yields for $^{193−204}$Po have been measured. Comparing the yields to the estimated production rates with the ABRABLA code concluded on an ionisation efficiency of at least 0.4%. The contaminants at mass A = 200 have been estimated to be less than 5%. Using these new beams, in-source laser spectroscopy has been performed on the isotopes $^{191−204,206−211,216,218}$Po with counting rates ranging from 0.01 ion/s in $^{ 191}$Po to over 10$^{7}$ ions/s in $^{208}$Po, and with half lives ranging from 33 ms in $^{192}$Po to 102 years in $^{209}$Po. In this Ph.D. work, the analysis of the even-A isotopes is reported. Large-scale atomic calculations were compared to the King plot made from the two transitions studied for $^{200−210}$Po. While high confidence is found on the electronic F-factors, the specific mass shift constants M are in disagreement. This result in a large systematic uncertainty in the extraction of the $\delta$(r$^{2}$). Those show nonetheless a large departure from the spherical droplet model for A < 200. This departure is not reproduced by including predicted static deformation parameters and the Beyond Mean Field calculations also fail to reproduce the trend of the most neutron-deficient isotope $^{192}$Po. Finally, relative charge radii are compared between the neighbouring even-Z elements $_{78}$Pt, $_{80}$Hg, $_{82}$Pb, $_{84}$Po and $_{86}$Rn. It shows that the magnitude of the departure from sphericity is much larger than for Z < 82 and highlights the importance that the specific proton shells are playing above Z = 82.