Artificial-intelligence-supported shell-model calculations for light Sn isotopes

Abstract

The region around the doubly magic nuclide $^{100}Sn$ is very interesting for nuclear physics studies in terms of structure, reaction and nuclear astrophysics. The main ingredients in nuclear structure studies using the shell model are the single-particle energies and the two-body matrix elements. To obtain the former, experimental data of $^{101}Sn$ isotope spectrum are necessary. Since there is not enough experimental data, different approaches are used in the literature to obtain single-particle energies. In sn100pn interaction, the hole excitation spectrum was used in $^{131}Sn$ to determine neutron single-particle energies. The other approach is the use of the lightest isotope, $^{107}Sn$, which figures the model space orbitals. In this study, we estimated the spectrum of the $^{101}Sn$ isotope by artificial neural network method in order to obtain neutron single-particle energies. After the training was carried out by using the experimental spectra of the nuclei around $^{100}Sn$ isotope, the $^{101}Sn$ spectrum was obtained. Subsequently, neutron SPEs of the model space orbitals are defined. Shell model calculations for $^{102-108}Sn$ isotopes are carried out and results are compared to the experimental data and results obtained using the widely used interaction in the region, sn100pn. According to the results, it is seen that the Sn isotope spectra obtained with the new SPE values are more compatible with the experimental data.