Exploring the high sensitivity of DFT thermochemistry for protonation states of a ferredoxin model complex [CH3S4Fe2IIIS2H]−.

This article focuses on evaluating the relative energies of four protonation isomers of a [2Fe–2S] ferredoxin model using density functional theory (DFT), coupled-cluster (CC), and phaseless auxiliary-field quantum Monte Carlo (ph-AFQMC) methods. The study finds that DFT results for iron–sulfur clusters are highly sensitive to the amount of exact Hartree–Fock exchange (HFX) in the exchange–correlation functional, with pure (meta-)GGA functionals underestimating and some hybrids overestimating the stability of iron-hydride isomers. Among tested DFT methods, r²SCAN12-D4, B3LYP-D4, and B97-1-D3(OP) (with 10%–21% HFX) best reproduce reference energies from advanced ab initio calculations. The authors also demonstrate that density-corrected DFT using Kohn–Sham CCSD electronic densities and direct random phase approximation (RPA) on top of certain hybrid functionals yield reliable results consistent with many-body benchmarks. Spin-projection corrections and dispersion effects are important for achieving chemical accuracy, and the study highlights the challenges posed by strong electron correlation and self-interaction errors in modeling FeS clusters.