Benchmarking Structures and UV-Vis Spectra of Iron Complexes Against Experimental Data

This benchmark study focuses on the evaluation of theoretical methodologies for geometry and ultraviolet-visible (UV-vis) spectral prediction of mononuclear iron coordination complexes. For this purpose, 17 structurally diverse iron complexes with experimentally determined X-ray structures and UV-vis absorption spectra were selected from the literature. For ground-state geometry, different computational approaches were evaluated: GFN1-xTB, BP86(D4), PBE(D4), revPBE(D4), OPBE(D4), TPSS(D4), r²SCAN, B97(D4), B3LYP/G(D4), TPSSh(D4), MN15, revM11, ωB97X(D4), HF-3c, r²SCAN-3c, and PBEh-3c. The meta-hybrid functional TPSSh(D4) delivers the best performance, establishing it as the preferred method for geometry optimizations of iron coordination complexes. For the prediction of UV-vis absorption spectra, time-dependent density functional theory (TD-DFT) calculations were performed on the optimized structures, at the TPSSh(D4)/def2-TZVP/CPCM level of theory, using 13 density functionals (TPSS, r²SCAN, revM06-L, TPSSh, O3LYP, B97, B3LYP/G, PBE0, MN15, revM11, ωPBE, CAM-B3LYP and ωB97X). The functionals were ranked based on their ability to reproduce both the excitation energies and the overall spectral shape of the experimental spectra after using optimized Gaussian broadening and energy shifts on the calculated spectra. The hybrid functional O3LYP provided the most accurate excitation energies, with the lowest average energy shift, while the meta-GGA functional revM06-L demonstrated exceptional performance for reproducing the spectral shape, with the highest median similarity to the experimental spectra.