The importance of tight $f$ basis functions for heavy p-block oxides and halides: a parallel with tight $d$ functions in the second row
It is well-known that both wavefunction ab initio and DFT calculations on second-row compounds exhibit anomalously slow basis set convergence unless the basis sets are augmented with additional `tight' (high-exponent) $d$ functions, as in the cc-pV($n+d$)Z and aug-cc-pV($n+d$)Z basis sets. This has been rationalized as being necessary for a better description of the low-lying $3d$ orbital, which as the oxidation state increases sinks low enough to act as a back-donation acceptor from chalcogen and halogen lone pairs. This prompts the question whether a similar phenomenon exists for the isovalent compounds of the heavy p-block. We show that for the fourth and fifth row, this is the case, but this time for tight $f$ functions enhancing the description of the low-lying $4f$ and $5f$ Rydberg orbitals, respectively. In the third-row heavy $p$ block, the $4f$ orbitals are too far up, while the $4d$ orbitals are adequately covered by the basis functions already present to describe the $3d$ subvalence orbitals.