Performance of the Bethe–Salpeter Equation for Electronic Excitations in First-Row Transition Metal Complexes

The performance of the Green's function-Bethe-Salpeter
Equation
(BSE@GW) approach has been systematically tested on a series of first-row
transition metal complexes. Special attention was paid to the interplay
between metal-centered (MC) and charge transfer (CT) transitions and
their ordering in the excited singlet and triplet state manifolds.
Here, the commonly used time-dependent density functional theory (DFT)
reaches its limits and in particular shows a strong dependence on
the exchange-correlation functional. In contrast, the present BSE@GW
ultraviolet–visible (UV–vis) spectra show a good agreement
with experimental results as far as the absorption spectra and their
assignment are concerned. This is achieved independently of the exchange-correlation
functional underlying the initial guess of the GW iteration. These
conclusions are drawn on the basis of an investigation of a test set
including nearly pure MC transitions in triazacyclononane with different
metal centers and mixed MC-CT transitions in Fe(II) complexes with
different numbers of N-heterocyclic carbene ligands. Furthermore,
important aspects of the calculations such as the dependence on the
ground state equilibrium geometry or the influence of the Tamm-Dancoff
approximation are highlighted.