Journal of Physical Chemistry A, Vol.103, No.37, 7496-7505, 1999
The electron affinities of the selenium fluorides SeFn (n=1-7)
The molecular structures, electron affinities, and dissociation energies of the SeFn/SeFn- (n = 1-7) species were examined using hybrid Hartree-Fock/density functional theory (DFT). The three different types of electron affinities reported in this work are the adiabatic electron affinity (EA(ad)), the vertical electron affinity (EA(vert)), and the vertical detachment energy (VDE). The first Se-F dissociation energies of the SeFn and SeFn- species were also been reported. The basis set used in this work is of double-zeta plus polarization quality with additional s- and p-type diffuse functions, and is denoted as DZP++. Four different density functionals (BHLYP, B3LYP, BP86, and BLYP) were used in this work. Among these, the best for predicting molecular structures and energies was found to be BHLYP, whereas other methods generally overestimated bond lengths. Neutral SeF7 was found to have no structures that were significantly bound with respect to dissociation. SeF7-structures with D-5h, C-4 upsilon, and C-3 upsilon symmetry were found to lie very close in energy. The most reliable adiabatic electron affinities, obtained at the DZP++ BHLYP level of theory, are 1.99 eV (Se), 2.37 eV (SeF), 2.21 eV (SeF2), 3.39 eV (SeF3), 2.50 eV (SeF4), 5.23 eV (SeF5), and 3.13 eV (SeF6). The BHLYP adiabatic electron affinities of the Se atom, SeF5, and SeF6 molecules predicted by this work are in good agreement with the experimental results, but the predicted electron affinities for SeF3 are much larger than the experimental value (1.7 +/- 0.1 eV) obtained by the electron impact appearance energy (EIAE) method, which usually gives lower EA(ad) values. The other molecular electron affinities (SeFn, n = 1, 2, 3, 7) are unknown experimentally. The predicted vertical detachment energy for SeF7- is very large, 8.01 eV. The neutral bond dissociation energies D-e(Fn-1Se-F) are largely unknown experimentally. For SeF5, the DFT methods predict D-e(F4Se-F) = 0.88-1.67 eV, which is lower than the experimental estimated value of 2.8 eV. The DZP++ BLYP bond dissociation energy value, D-e(F5Se-F) = 3.15 eV, is slightly lower than the dissociation energies predicted by the other methods (DZP++ BHLYP, 3.34 eV; DZP++ B3LYP, 3.31 eV; DZP++ BP86, 3.44 eV). Except for the DZP++ BP86 result, theory matches the experimental estimate 3.15 +/- 0.2 eV based on thermochemical data. Excluding the DZP++ BHLYP results, the dissociation energy for diatomic SeF ranges from 3.4 to 3.80 eV among which the DZP++ B3LYP result (3.40 eV) is in best agreement with the experimental value (3.5 eV). For the bond dissociation value of the anion D-e(SeFi(5)(-)-F) the DZP++ BHLYP method gives D-e(SeF5- -F) = 1.23 eV, whereas the DZP++ B3LYP, DZP++ BP86, and DZP++ BLYP methods predict dissociation energies (B3LYP, 1.83 eV; BP86, 2.26 eV; BLYP, 2.13 eV) that are larger than experiment (1.09 +/- 0.1 eV). It is concluded that the density functional methods, although very useful in establishing trends, must be used very carefully. Moreover, additional (SeFn-SeFn-) experiments are required to precisely establish the reliability of the different density functional methods.