Are ther known compounds of Argon? What is their molecular geometry and hybridisation?

Question

Can Argon hybridize orbitals and/or form covalent-like/ionic-like compounds? Is there any study of that? I would be happy to read concrete references.

What kind of molecular geometries for argon compounds are known or hypothesized at the current time?

Is there some peculiarity to apply something like VSEPR or molecular orbital theory to theoretical or experimentally known argon compounds?

Answer 1

Argon is a quite unreactive element, however very few compounds of it are known.

The most popular, and currently the only neutral compound, is argon hydroflouride, $\ce{HArF}$, which was reported by L. Khriachtchev et.al..^[1] In the same issue there is also a comment on the matter by my former supervisor G. Frenking.^[2]
Another interesting read on the subject is the prediction of the metastable $\ce{HHeF}$, by M. W. Wong.^[3] It generally treats all compounds of the $\ce{HNgF}$ family.

All calculations done so far indicate, that the molecule is linear and only kinetically stable. The highly accurate CCSD(T)/aug-cc-pV5Z level predicts, that the $\ce{HArF}$ molecule is $5.87~\mathrm{eV}$ higher in energy than its fragments $\ce{HF}$ and $\ce{Ar}$. The minimum barrier on this level of theory is a mere $0.35~\mathrm{eV}$ along the stretching coordinate of the hydrogen argon bond. The bending is predicted to be higher in energy. The reported bond distances for $\ce{H-Ar}$ and $\ce{Ar-F}$ are $133~\mathrm{pm}$ and $197~\mathrm{pm}$, respectively.
An NBO analysis on the MP2/aug-cc-pVTZ level of theory shows a strong ionic character of the molecule, it is reported, that there is a strong $\ce{(HAr)+F- }$ charge transfer in the electronic structure of the molecule.
Wong performed accurate calculations on the QCISD/6-311+G** level of theory and combined it with a Quantum Theory of Atoms in Molecules approach. He calculated a positive value of the Laplacian of the electron density $\nabla^2\rho_\mathrm{b}$ for the bond critical point of the argon fluoride bond, which indicates a mainly electrostatic interaction. The contrary is the case for the argon hydrogen bond, where the Laplacian is found to be negative, i.e. charge concentration, and therefore indicating a dominantly covalent interaction.
The following picture is a snapshot of this behaviour at the BP86/cc-pVTZ level of theory.^[4]

Solid lines are regions of charge depletion, $\nabla^2\rho_\mathrm{b}>0$, while dashed lines are regions of charge concentration $\nabla^2\rho_\mathrm{b}<0$. Red lines are zero flux surfaces, where $\nabla\rho_\mathrm{b}=0$. And red spheres indicate bond critical points.

This molecule is quite interesting for many reasons, however, since it is only kinetically stable it has to be treated with fullest force of quantum chemical calculations. The calculations I did on a low level DFT method are not sufficient at all on its own, as has been investigated by J. Panek et.al..^[5]]

Hybridisation is in a wider sense the mixing of s and p orbitals;^[6] and upon further examination of the electron density this will be observed in this molecule. Most likely for the hydrogen argon covalent bond. As hybridisation is just a mathematical transformation of the original atomic orbitals there is no real interpretative value in this case. For example, the NBO analysis at the BP86/cc-pVTZ level of theory give only a negligible 3% s contribution of argon to the $\ce{H-Ar}$ bond.

The VSEPR theory is only a observational concept for well-behaved molecules. It basically lacks a solid scientific basis and should not be used any more, but especially in cases like this, it must not be used.

---

Literature and Notes

1. Leonid Khriachtchev, Mika Pettersson, Nino Runeberg, Jan Lundell, and Markku Räsänen, Nature, 2000, 406 874-876.
2. Gernot Frenking, Nature, 2000, 406, 836-837.
3. Ming Wah Wong, J. Am. Chem. Soc., 2000, 122 (26), 6289-6290.
4. To avoid copyright issues I performed a quick calculation to illustrate the above raised point.
5. Jarosław Panek, Zdzisław Latajka, and Jan Lundell, Phys. Chem. Chem. Phys., 2002, 4, 2504-2510.
6. Hybridisation is a concept that dates back to the beginnings of Valence Bond (VB) Theory. It was originally implemented in the strict fashion of creating $s^{1/(n+1)}p^{n/(n+1)}$ or $sp^n$ hybrid orbitals for $n = \{1,2,3\}$. H. Bent later went on suggesting fractional hybridisation schemes, see Henry A. Bent, Chem. Rev., 1961, 61 (3), 275–311. Nowadays the term s-p-mixing in Molecular Orbital theory is almost synonymous with the VB theory term hybridisation.