Ozone: A failure of FLOSIC or standard DFT-like symmetry-breaking?

There are well known cases where DFT requires spin disproportionation to determine qualitatively correct bond breaking and cases where such disproportionation remains at the equilibrium geometry. Ozone (O₃) is an example of the latter based on conventional many-electron methods. The structure of the O₃ molecule has been of great importance for over four decades due to its crucial role on life on planet earth. From a theoretical point of view, it offers a unique and challenging problem though its structure seems so simple for the advanced theories of today. In this talk we will present the results of the electronic structure study of O₃ using density functional based approximations and show that for the first time, the ground state of the molecule is correctly described by DFT only when self-interaction-corrections (SIC) are taken into account; in particular Fermi-Löwdin Orbital-SIC. Using these results, we will show that while a broken spin-symmetry is found on the O₃ molecule, the overall symmetry of the molecule, i.e. the symmetry of density, is preserved and the misconception of FLOSIC’s failure to correctly describe ozone [1] is resolved. [1] Hahn et al., J. Chem. Theory Comput. 13, 5823-5828 (2017).