Electron correlation and the reality of xenon difluoride

Abstract

The electronic structure of xenon difluoride has been studied using ab initio theoretical methods. The primary goal was to determine whether current theoretical methods are capable of yielding a reasonable value of the dissociation energy of XeF2. A Slater function basis set of slightly better than "double [ plus polarization" quality was employed. Four different types of wave functions were investigated: two-configuration SCF, full valence configuration interaction (CI), the first-order wave function, and a larger 1234 configuration wave function including all double excitation from the loa, orbital. Although the TCSCF symmetric stretching potential curve has both a minimum and maximum, the minimum lies above the comparable energy of separated Xe + 2F. However, the two most complete wave functions predict dissociation energies of 1.97 and 2.14 e", in qualitative agreement with experiment, 2.78 eV. All four wave functions provide good predictions of the Xe-F equilibrium bond distance. As was the case for KrF2, the bonding in XeF2 is found to conform quite closely to Coulson's model F Xe+ F--F-Xe+ F near the equilibrium geometry, The role of the "outer orbitals" 5d and 4f appears to be a quantitative rather than qualitative one.Xenon difluoride appears to be the simplest known Xecontaining molecule, although there is still some controversy2 concerning the existence of the XeF radical. As such, XeF2 plays a special role in the chemistry of the noble gases.' XeF2 was first prepared4 in 1962, shortly after Bartlett's discovery5 of XePtF6, and several relatively simple methods of preparation are now a~a i l a b l e .~ The dissociation energy for the process XeF2 -Xe + 2F is -64 kcal/mo16 = 2.78 eV. Assuming the value' 38.8 f 2.3 kcal/mol for the dissociation energy De of F2, the molecular dissociation energy for the process XeF2 -Xe + F2 is found to be -25 kcal/mol. For comparison, the smaller KrF2 molecule is known to lie energetically aboue (by -15 kcal/mol) the analogous dissociation limit Kr + F2. This difference between KrF2 and XeF2 explains the transient nature of the former as compared to the relative stability of the latter. The geometrical structure of XeF2 is known from infrared and Raman studies to be linear and ~y m m e t r i c ,~ corresponding to point group Dmh. Reichman and Schreinerg have determined the gas-phase Xe-F bond distance to be 1.977 f 0.002

An attempt on the calculation of relativistic potential energy curves: Noble gas difluorides

“…The large change compared with the minimal basis set calculation shows that the result is largely basis set dependent, but the trend of the potential energy curve especially at large distances is in the right direction. It is wellknown from non-relativistic calculations that a larger basis increases the minimum whereas the non-physical maximum in the outer region disappears [13,14]. The same trend can be observed here.…”

“…The reason for this choice was the theoretical importance of these compounds which did not exist until the early 60's [11][12]. Furthermore, very elaborate non-relativistic calculations have been published [13][14]. The only noble gas difluorides which are known are KrF 2 and XeF 2 with a bond length of 3.5 a.u.…”

The essential role of charge-shift bonding in hypervalent prototype XeF2

“…Computational background. Since its synthesis 2 in 1962, the XeF 2 molecule has been the subject of a number of ab initio theoretical studies 13,[24][25][26][27][28][29][30][31] . These studies show a consensus that electron correlation effects are essential [25][26][27][28][29][30] .…”

NF5Viable or Not?

“…85,90 On the basis of the 3c-4e model, Rundle, 94 Musher 95 (who introduced nonorthogonal sp x hybrids), 96 and 85 concluded that due to the partial ionic "fractional bonds" 85 of the 3c-4e bond model the octet rule is not violated. The advances in computational quantum chemistry in the 1970s made the computation of reliable polyatomic wave functions possible 97,98 and, as reviewed by Kutzelnigg,32 reinforced the 3c-4e picture of bonding. Interpretations of computed wave functions by means of shared electron number and occupation number analysis, 99-101 electron density analysis, [102][103][104] and natural population analysis (NPA) [105][106][107][108][109] reveal in the 1980s that the d orbital contribution to bonding in hypervalent molecules is far less than the traditional sp 3 d and sp 3 d 2 models demand.…”