Validation of Relativistic DFT Approaches to the Calculation of NMR Chemical Shifts in Square-Planar Pt(2+) and Au(3+) Complexes.

Recently implemented hybrid density functional methods of calculating nuclear magnetic shielding using the two-component zeroth-order regular approximation approach (J. Phys. Chem. A2009, 113, 11495) have been employed for a series of compounds containing heavy transition-metal atoms. These include Pt(2+), Pd(2+), and Au(3+) organometallics and metal complexes with azines, some of which exhibit interesting biological and catalytic activities. In this study we investigate the effects of geometry, exchange-correlation functional, solvent, and scalar relativistic and spin-orbit corrections on the nuclear magnetic shielding-mainly for (13)C and (15)N atoms connected to a heavy-atom center. Our calculations demonstrate that the B3LYP method using effective core potentials and a cc-pwCVTZ-PP/6-31G** basis set augmented with the polarizable continuum model of the dimethylsulfoxide solvent provides geometries for the complexes in question which are compatible with the experimental NMR results in terms of both the trends and the absolute values of the (13)C shifts. The important role of the exact exchange admixture parameter for hybrid functionals based on B3LYP and PBE0 is investigated systematically for selected Pt(2+) and Au(3+) complexes. The (13)C and (15)N NMR chemical shifts are found to be best reproduced by using a B3LYP or PBE0 approach with 30% and 40-50% exact exchange admixtures for the Pt(2+) and Au(3+) complexes, respectively. The spin-orbit contributions to the (15)N NMR chemical shifts reflect metal-ligand bonding that is much more ionic for the Au(3+) than for the Pt(2+) complex. Finally, an optimized density functional method is applied to a series of transition-metal complexes to estimate the scope and the limitations of the current approach.