Probing the performance of DFT in the structural characterization of [FeFe] hydrogenase models

Abstract

In this work, a series of DFT and DFT-D methods is combined with double-ζ basis sets to benchmark their performance in predicting the structures of five newly synthesized hexacarbonyl diiron complexes with a bridging ligand featuring a μ-S₂C₃ motif in a ring-containing unit functionalized with aromatic groups. Such complexes have been considered as [FeFe] hydrogenase catalytic site models with potential for eco-friendly energetic applications. According to this assessment, r²SCAN is identified as the density functional recommended for the reliable description of the molecular and crystal structures of the herein studied models. However, the butterfly (μ-S)₂Fe₂ core of the models demonstrates a minor deformation of its optimized geometry obtained from both molecular and periodic calculations. The FeFe bond length is slightly underestimated while the FeS bonds tend to be too long. Adding the D3(BJ) correction to r²SCAN does not lead to any improvement in the calculated structures.

Keywords: H‐cluster; benchmarking; density functional theory; hydrogenase; molecular crystal.