Pure and Applied Chemistry

Abstract: Since density functional theory (DFT) achieved a remarkable break-through in computational chemistry, the important general question "How reliable are quantum chemical calculations for spectroscopic properties?" should be answered anew. In this project, the most successful density functionals, namely the Becke B3LYP functionals, and the correlation-consistent polarized valence quadruple zeta basis sets (cc-pvqz) are applied to small molecules. In particular, the complete set of experimentally known diatomic molecules formed by the atoms H to Ar (these are 214 species) is uniformly calculated, and calculated spectroscopic properties are compared with experimental ones. Computationally demanding molecules, such as open-shell systems, anions, or noble gas compounds, are included in this study. Investigated spectroscopic properties are spectroscopic ground state, equilibrium internuclear distance, harmonic vibrational wavenumber, anharmonicity, vibrational absolute absorption intensity, electric dipole moment, ionization potential, and dissociation energy. The same computational method has also been applied to the ground-state geometries of 56 polyatomic molecules up to the size of benzene. Special sections are dedicated to nuclear magnetic resonance (NMR) chemical shifts and isotropic hyperfine coupling constants. Each set of systems for a chosen property is statistically analyzed, and the above important question "How reliable...?" is mathematically answered by the mean absolute deviation between calculated and experimental data, as well as by the worst agreement. In addition to presentation of numerous quantum chemically calculated spectroscopic properties, a corresponding updated list of references for experimentally determined properties is presented.