|Description||Photodetachment of O- (2p5)2P° has been investigated with a new R-matrix (close-coupling) code  that uses a B-spline basis for the continuum functions. The completeness of the B-spline basis ensures that no Buttle correction is needed for the R-matrix elements. Another distinguishing feature of the present work is the use of non-orthogonal orbital sets in the construction of the target and the scattering wavefunctions. This allows us to employ a highly correlated target description without possibly introducing inconsistencies between the continuum and the bound parts of the trial wavefunction.
Our close-coupling expansion includes the lowest 25 bound and autoionizing states of neutral oxygen. Emphasis was placed on the accuracy of the target wavefunctions by describing each state (including the initial O- state) with a different set of orbitals, determined separately from state-specific multi-configuration Hartree-Fock calculations. Note that using non-orthogonal orbital sets directly accounts for relaxation effects during 2s or 2p excitation.
Our absolute values of the photodetachment cross sections exceed the experimental values  by 40%. The latter measurements were carried out 45 years ago, but the data are still being used today to normalize current photodetachment measurements. Our previous calculations of B- and C- photodetachment  also show excellent agreement with experiment for the energy dependence, but the theoretical absolute values exceed the experimental data by 35%. Note, however, that both the measured B- and C- photodetachment cross sections were normalized to the O- data from . Such very similar discrepancies between the absolute values from three independent calculations with a single experiment suggest a possible systematic error in .
 O. Zatsarinny, Comp. Phys. Commun. 174 (2006) 273.
 S.J. Smith, Proc. 4th ICIPG, Uppsala (1959) IC219
 O. Zatsarinny and T.W. Gorczyca, Abstracts of XXIII ICPEAC, Stockholm (2003); We026, Th006.||en