Comment on "Multireference configuration-interaction calculations for positronium halides" [J. Chem. Phys. 122, 054302 (2005)]

Large-scale configuration–interaction calculations of the binding energies and annihilation rates of the positronium halides, PsF, PsCl, PsBr, and PsI [S. L. Saito, J. Chem. Phys. 122 054302 (2005)], have made erroneous predictions about the structures of these atoms. The predictions were based on small annihilation rates, which result from using a small basis and additionally invalid estimates of the contributions from single-particle orbitals with >8.