NSC23766 The shape and the location

The shape and the location of this peak were examined, and the parameters necessary for constructing the Fano profile (6) were computed (see Table 2). The computed peak location is in good agreement with the experiment; unfortunately, however, since it is narrow and highly intense, no specific conclusions can be drawn regarding the coincidence of the height and the shape. The computed resonance lines and their respective Fano profiles are shown in Fig. 5 Very narrow and closely spaced peaks caused by the resonance of the 3s→np channel with discrete 2p→ns and 2p→nd transitions are observed in the 33.0–38.5eV region. Above the NSC23766 of 38.5eV corresponding to the bonding energy of the 2p-shell, the ionization of this shell starts occurring in two channels, 2p→ns and 2p→nd. A number of peaks are observed in the 38.5–45.5eV energy range in the experimental cross-section, which can be attributed to two-electron excitations characterized by the simultaneous excitation of the 2p and the 3s-shell electrons [14]. Furthermore, above 70.9eV, the photoionization of the 2s-shell starts. The excitations of the 2s-electrons in the continuous spectrum contribute insignificantly to the total photoionization cross-section; however, the discrete 2s→3p and 2s→4p transitions cause autoionization resonance peaks to emerge in the 63–71 energy range. Numerical computations were performed; the shapes and the locations of the autoionization resonance peaks caused by the interactions with the discrete 2s→3p and 2s→4p transitions were examined. It can be seen from the experimental curve [14] that each of these peaks is in turn split into two lines corresponding to a specific spin state, i.e. the triplet and the singlet. Because of this, the locations of these peaks were computed within SP RPAE. The computations performed in this approximation showed that the discrete 2s→3p(2P), 2s→3p(1P), 2s→4p(2P) and 2s→4p(1P) transition energies were equal to 70.53, 71.75, 74.45 and 74.34eV, respectively. Consequently taking into account the dynamic polarization potential allowed shifting the location of the 2s-level from the energy of –76.11eV to–71,95eV, i.e., the transition energies decreased by 4.16eV and became equal to 66.37, 67.59, 70.29 and 70.175 эВ,eV, respectively. Remarkably, according to the computations performed, the singlet-state energy for the 2s→4p transition proved to be lower than the triplet-state one. This fact needs to be rechecked using the multi-configuration Hartree–Fock approximation. There are discrepancies in the experimental and the computed peak shapes. The computed ones turned out to be more intense, and there is a characteristic dip in the photoabsorption cross-sections after each peak. This is possibly because an insufficiently small energy step was taken when the experiment was carried out in this energy range. In particular, this would explain the absence of the autoionization resonance peaks caused by the discrete 2s→5p, 2s→6p, etc., transitions on the experimental curve. It should be noted that taking into account the polarization corrections does not lead to a significant improvement in the agreement between the experimental and the computed cross-sections in the 30–100eV energy range; it does, however, allow to refine the locations of the autoionization resonance peaks and the ionization thresholds of shells. Introducing the polarization corrections using the dynamic polarization potential in the 5.14–20.14eV range turned out to be justified as it improved the agreement between the computed and the experimental results. Therefore, we can conclude that the effects related to atomic core polarization have a significant effect on the transition amplitudes only at low photoelectron energies. Unfortunately, the obtained computed photoabsorption cross-section is different from the experimental one, and this discrepancy grows with an increase of the photon energy. The agreement between the cross-sections is 80–90% in the 45–64eV energy range, and decreases to 70% in the 70–100eV range. Presumably, this could be due to the fact that the computations were performed in a non-relativistic approximation. It is possible that using the Hartree–Fock–Dirac approximation for computing electron wave functions could improve the agreement between the computations and the experiment.