Existence of noble gas-inserted phosphorus fluorides: FNgPF2 and FNgPF4 with Ng-P covalent bond (Ng = Ar, Kr, Xe and Rn).

Abstract

The scarce literature on noble gas (Ng)-phosphorous chemical bonding and our recent theoretical prediction of the FNgP molecule motivate us to explore a unique novel class of neutral noble gas-inserted phosphorus trifluoride and pentafluoride molecules, i.e., FNgPF2 and FNgPF4 (Ng = Ar, Kr, Xe, and Rn). The predicted molecules have been designed by inserting an Ng atom between the F and P atoms in the PF3 and PF5 molecules. The minima and saddle point geometries of all the FNgPFn (n = 2 and 4) molecules have been optimized using density functional theory (DFT) and second-order Møller-Plesset perturbation theory (MP2). The coupled cluster theory (CCSD(T)) method is also used to optimize the FNgPF2 molecules to test the performance of the above-mentioned methods. The predicted FNgPF2 and FNgPF4 molecules are found to be energetically stable with respect to all the probable 2-body and 3-body dissociation channels, except for the one leading to the global minimum products (Ng + PF3 and Ng + PF5). The existence of large barrier heights corresponding to the saddle point geometries is responsible for the kinetic stability of the metastable FNgPFn (n = 2 and 4) molecules, which prevents them from dissociating into their global minima products. The optimized structural parameters, energetics and harmonic vibrational frequency analysis suggest that the Ng-P bond is covalent in nature, while the F-Ng bond is mostly ionic in nature with some degree of covalency in the predicted molecules. In fact, the Ng-P bond length in the experimentally observed Ng-PF3 van der Waals complex is reduced significantly in the isomeric FNgPF2 molecule, almost leading to a conventional covalent Ng-P bond (cf. 4.152 vs. 2.413 Å for the Kr-P bond). Furthermore, the charge distribution and the AIM analysis also confirm the above-mentioned conclusion and indicate that the predicted FNgPF2 and FNgPF4 molecules can be represented as [F]δ-[NgPF2]δ+ and [F]δ-[NgPF4]δ+, respectively. All the computational results strongly reinforce the possible existence of these predicted FNgPFn (n = 2 and 4) molecules and clearly indicate that it may be possible to synthesize and characterize these molecules under suitable experimental technique(s).