Computational chemistry currently lacks ad hoc tools for probing the nature of chemical bonds in heavy and superheavy-atom systems where the consideration of spin-orbit coupling (SOC) effects is mandatory. We report an implementation of the Quantum Theory of Atoms-In-Molecules in the framework of two-component relativistic calculations. Used in conjunction with the topological analysis of the Electron Localization Function, we show for astatine (At) species that SOC significantly lowers At electronegativity and boosts its propensity to make charge-shift bonds. Relativistic spin-dependent effects are furthermore able to change some bonds from mainly covalent to charge-shift type. The implication of the disclosed features regarding the rationalization of the labeling protocols used in nuclear medicine for (211)At radioisotope nicely illustrates the potential of the introduced methodology for investigating the chemistry of (super)heavy elements.

The potential energy surfaces of the group 17 XF3 (X = Cl, Br, I, At) fluorides have been investigated for the first time with multiconfigurational wave function theory approaches. In agreement with experiment, bent T-shaped C(2v) structures are computed for ClF3, BrF3, and IF3, while we predict that an average D(3h) structure would be experimentally observed for AtF3. Electron correlation and scalar relativistic effects strongly reduce the energy difference between the D(3h) geometry and the C(2v) one, along the XF3 series, and in the X = At case, spin-orbit coupling also slightly reduces this energy difference. AtF3 is a borderline system where the D(3h) structure becomes a minimum, i.e., the pseudo-Jahn-Teller effect is inhibited since electron correlation and scalar-relativistic effects create small energy barriers leading to the global C(2v) minima, although both types of effects interfere.