November 19, 02:00 PM, Paris Time (GMT +1).
Paving the way for the ab initio description of FeSe
The major role played by strong correlations in iron-based superconductors (IBSC) has been widely assessed, as pointed out by the successful predictions of local magnetic moments and mass differentiations in these materials. Nevertheless, cumulating evidence shows that local correlations alone cannot capture the finer details of their electronic structure, yielding Fermi pocket way too large when compared with experiments. This issue is most striking in FeSe, where single-site dynamical mean-field theory (DMFT) predicts Fermi pockets from five to six times larger than experiments, hampering the theoretical description of the related electronic instabilities. On general grounds, this deficiency has been related to the lack of non-local interactions in the theoretical treatment. In this seminar, I will show how the interplay between non-local exchange and local interactions is responsible for the main low-energy features of FeSe by means of DFT + slave-spin mean-field calculations. Within this framework, a net shrinking of the Fermi pockets accompanies with an overall improvement of the quasiparticle properties, as documented by the comparison with angle-resolved photoemission spectroscopy (ARPES) and transport measurements. Additionally, the relatively low computational cost of our approach opens the way to future application to more challenging scenarios, e.g. the surface properties of correlated materials.