Pseudopotential approaches to Ca, Sr, and Ba hydrides. Why are some alkaline earth MX2 compounds bent?

Quasirelativistic and nonrelativistic 10-valence-electron pseudopotentials for Ca, Sr, and Ba are presented. Results of calculations with 6s6p5d basis sets for MH, MH+, and MH2 are compared with all-electron and 2-valence-electron pseudopotential calculations with and without core-polarization potentials. The 10-valence-electron pseudopotential approach agrees well with all-electron calculations. It circumvents problems for the 2-valence-electron pseudopotentials arising from an incomplete separation of valence and subvalence shells in polar molecular systems due to strongly contracted occupied (n−1)-d orbitals. All higher-level calculations show SrH2 and BaH2 to be bent with angles of ∼140° and 120°, respectively, while CaH2 is linear with a flat potential-energy surface for the bending motion. The use of a core-polarization potential together with the 2-valence-electron pseudopotential approach allows an investigation of the relative importance of core-polarization vs direct d-orbital bonding participation as reasons for the bent structures. The calculations strongly suggest that both contribute to the bending in SrH2 and BaH2. Even at the Hartree–Fock level of theory 10-valence-electron pseudopotential calculations given reasonable angles when the potential-energy surface is not exceedingly flat, and only moderately contracted basis sets including both compact d functions and diffuse p functions are used. The effect of core-valence correlation and the importance of f functions also are discussed.