Theoretical predictions of nuclear magnetic resonance parameters in a novel organo-xenon species: chemical shifts and nuclear quadrupole couplings in HXeCCH

We calibrate the methodology for the calculation of nuclear magnetic resonance (NMR) properties in novel organo-xenon compounds. The available state-of-the-art quantum-chemical approaches are combined and applied to the HXeCCH molecule as the model system. The studied properties are (129)Xe, (1)H, and (13)C chemical shifts and shielding anisotropies, as well as (131)Xe and (2)H nuclear quadrupole coupling constants. The aim is to obtain, as accurately as currently possible, converged results with respect to the basis set, electron correlation, and relativistic effects, including the coupling of relativity and correlation. This is done, on one hand, by nonrelativistic correlated ab initio calculations up to the CCSD(T) level and, on the other hand, for chemical shifts and shielding anisotropies by the leading-order relativistic Breit-Pauli perturbation theory (BPPT) with correlated ab initio and density-functional theory (DFT) reference states. BPPT at the uncorrelated Hartree-Fock level as well as the corresponding fully relativistic Dirac-Hartree-Fock method are found to be inapplicable due to a dramatic overestimation of relativistic effects, implying the influence of triplet instability in this multiply bonded system. In contrast, the fully relativistic second-order Moller-Plesset perturbation theory method can be applied for the quadrupole coupling, which is a ground-state electric property. The performance of DFT with various exchange-correlation functionals is found to be inadequate for the nonrelativistic shifts and shielding anisotropies as compared to the CCSD(T) results. The relativistic BPPT corrections to these quantities can, however, be reasonably predicted by DFT, due to the improved triplet excitation spectrum as compared to the Hartree-Fock method, as well as error cancellation within the five main BPPT contributions. We establish three computationally feasible models with characteristic error margins for future calculations of larger organo-xenon compounds to guide forthcoming experimental NMR efforts. The predicted (129)Xe chemical shift in HXeCCH is in a novel range for this nucleus, between weakly bonded or solvated atomic xenon and xenon in the hitherto characterized molecules.