Experimental and computational study of neutral xenon halides (XeX) in the gas phase for X=F, Cl, Br, and I

Abstract

We report a combined experimental and theoretical study of the xenon monohalide radicals XeX • ͑XϭF, Cl, Br, and I͒ together with their cationic and anionic counterparts XeX ϩ and XeX Ϫ. In brief, the XeX ϩ cations are characterized by reasonably strong chemical bonds with significant charge-transfer stabilization, except for XϭF. In contrast, the neutral XeX • radicals as well as the XeX Ϫ anions can mostly be described in terms of van der Waals complexes and exhibit bond strengths of only a few tenths of an electron volt. For both XeX • and XeX Ϫ the fluorides ͑XϭF͒ are the most strongly bound among the xenon halides due to significant covalency in the neutral radical, and to the large charge density on fluoride in the XeX Ϫ anion, respectively. Mass spectrometric experiments reveal the different behavior of xenon fluoride as compared to the other halides, and in kiloelectron-volt collisions sequential electron transfer according to XeX ϩ →XeX • →XeX Ϫ can be achieved allowing one to generate neutral XeX • radicals with lifetimes of at least a few microseconds for XϭF and I.

An ab initio study of the mono- and difluorides of krypton

“…Early ab initio calculations predicted dissociative ground electronic states for these molecules; [7][8][9][10][11] however, crossed molecular-beam studies clearly show bound ground states, 12,13 though the binding energies are small. Recently published results of high-level calculations on the halides of xenon have yielded ground-state potentials in remarkably good agreement with experimental ones; 14 these are the first such results to be reported for any of the noble-gas monohalides.…”

“…27 As has been noted elsewhere in the context of XeF, 14 inclusion of the effect of triples greatly improves the quality of the calculated potential over that of CCSD. Additionally, the calculated energies were usually corrected by subtracting the BSSE calculated using the counterpoise method.…”

“…Schröder et al 14 observed similar disagreement in their CCSD͑T͒ calculations on XeF ͑using a relativistic effective core potential basis set for Xe, and an augmented, correlation-consistent basis set for F͒. They suggested that this reflects a failure of the coupled-cluster technique to accurately reflect the multiconfigurational nature of the wave function.…”

“…7,9,10 The studies by Dunning and Hay, 9-11 however, did reveal a great deal of information about the excited charge-transfer states of this and other noble-gas monohalides, much of which is still quite valuable. Until very recently, 14 there have been no attempts to apply more modern techniques to this general problem. Our objective is to find an ab initio technique capable of producing a groundstate potential for KrF which is bound.…”

“…They suggested that this reflects a failure of the coupled-cluster technique to accurately reflect the multiconfigurational nature of the wave function. 14 These workers 14 then applied the multireference-averaged coupledpair functional ͑MR-ACPF͒ technique in order to incorporate some of the charge-transfer states into the ground-state wave function; this resulted in a 28% increase in the dissociation energy of XeF over CCSD͑T͒, uncorrected for BSSE. Then, the effects of spin-orbit coupling, quite large for Xe ϩ , were also included in the calculation.…”