Michael J. G. Peach, John A. Kattirtzi, Andrew M. Teale, David J. Tozer, J. Phys. Chem. A, 114, 7179–7186, 2010.
This project was one of the first applications of the optimised effective potential (OEP) methodology of Yang and Wu to Coulomb-attenuated functionals, and was the first to combine CAM-B3LYP and transition-metal NMR. It reiterates the importance of using the OEP procedure when evaluating magnetic response parameters from functionals containing exact orbital exchange, and evaluates the practicability of several approaches for avoiding unphysical potentials.
Further information, including details of subsequent work in this area, can be found on the research page. For the abstract, and access to the full text, see below.
The influence of the optimized effective potential (OEP) and Coulomb-attenuation on shielding constants and chemical shifts is investigated for three disparate categories of molecule: main group, hydrogen bonded, and transition metal systems. Expanding the OEP in the orbital basis leads to physically sensible exchange-correlation potentials; OEP generalized gradient approximation results provide some indication of the accuracy of the expansion. OEP uncoupled magnetic parameters from representative hybrid and Coulomb-attenuated functionals can be a dramatic improvement over conventional results; both categories yield similar accuracy. Additional flexibility is introduced by expanding the OEP in an extensive even-tempered basis set, but this leads to the well-known problem of unphysical, oscillatory potentials. Smooth potentials are recovered through the use of a smoothing norm, but deficiencies in the procedure are highlighted for transition metal complexes. The study reiterates the importance of the OEP procedure in magnetic response calculations using orbital-dependent functionals, together with the need for careful attention to ensure physically sensible potentials. It also illustrates the utility of Coulomb-attenuated functionals for computing short-range molecular properties.