TD-DFT Benchmark on Inorganic Pt(II) and Ir(III) Complexes

We report in the present paper a comprehensive investigation of representative Pt(II) and Ir(III) complexes with special reference to their one-photon absorption spectra employing methods rooted in density functional theory and its time dependent extension. We have compared nine different functionals ranging from generalized gradient approximation (GGA) to global or range-separated hybrids, and two different basis sets, including pseudopotentials for 4 iridium and 7 platinum complexes. It turns out that hybrid functionals with the same exchange part give comparable results irrespective of the specific correlation functional (i.e., B3LYP is very close to B3PW91 and PBE0 is very close to MPW1PW91). More recent functionals, such as CAM-B3LYP and M06-2X, overestimate excitation energies, whereas local functionals (BP86 -GGA-, M06-L -Meta GGA-) strongly underestimate transition energies with respect to experimental results. As expected, basis set effects are weak, and the use of a triple-ζ polarized (def2-TZVP) basis set does not significantly improve the computed excitation energies with respect to a classical double-ζ basis set (LANL2DZ) augmented by polarization functions, but it significantly raises the computational effort.

We report in the present paper a comprehensive investigation of representative Pt(II) and Ir(III) complexes with special reference to their one-photon absorption spectra employing methods rooted in density functional theory and its time dependent extension. We have compared nine different functionals ranging from generalized gradient approximation (GGA) to global or range-separated hybrids, and two different basis sets, including pseudopotentials for 4 iridium and 7 platinum complexes. It turns out that hybrid functionals with the same exchange part give comparable results irrespective of the specific correlation functional (i.e., B3LYP is very close to B3PW91 and PBEO is very close to MPW1PW91). More recent functionals, such as CAM-B3LYP and M06-2X, overestimate excitation energies, whereas local functionals (BP86 -GGA-, M06-L -Meta GGA-) strongly underestimate transition energies with respect to experimental results. As expected, basis set effects are weak, and the use of a triple-C polarized (def2-TZVP) basis set does not significantly improve the computed excitation energies with respect to a classical double-C basis set (LANL2DZ) augmented by polarization functions, but it significantly raises the computational effort.