Second-order Møller–Plesset perturbation theory for computing molecular-field splitting: application to the S2p3/2 level in C2H2n+1SF5, n=0, 1, and 2

High-resolution molecular X-ray photoelectron spectra of second-row atoms may reveal broadening or even splitting of the 2p 3/2 peak as compared to the 2p 1/2 component of the spectrum. This splitting reflects the lifted degeneracy of atomic 2p orbitals at sites of less than cubic symmetry and is referred to as molecular-field splitting (MFS). The prospect of using second-order Moller–Plesset theory (MP2) for ab initio calculation of the MFS for sulfur 2p 3/2 levels is examined. This method is subsequently applied to compute the MFS in ethynyl, ethenyl, and ethyl sulfur pentafluoride, resulting in values of 80, 116 and 121 meV, respectively. The initial-state contribution to the splitting is analyzed in terms of asymmetry in the electron density at sulfur, on the one hand, and in terms of the sulfur-ligand overlap density and electron density at the groups bonded to sulfur, on the other. At the Koopmans' theorem level of theory, the main source of splitting in the title compounds is found to be s–d hybridization of sulfur. At post-SCF levels of theory, core-valence electron correlation contributes substantially to the MFS, in proportion to the occupational asymmetry of the S3p shell. The realization that mixing of sulfur atomic orbitals of the same parity may add important contributions to the electric-field gradient at the core, resolves the observed diversity in the relative importance of core-valence electron correlation between different classes of sulfur compounds.