Assessment of DFT functionals for a minimal nitrogenase [Fe(SH)4H]- model employing state-of-the-art ab initio...

Area of Science:

• Computational Chemistry
• Quantum Mechanics
• Theoretical Chemistry

Background:

• Accurate calculation of electronic structure and energy differences is crucial for understanding chemical systems.
• The [Fe(SH)4H]- model presents a challenge for standard quantum-mechanical (QM) methods due to proton tautomerism.
• Density functional theory (DFT) methods often show significant variability in results for such systems.

Purpose of the Study:

• To determine a reliable energy difference (∆E) between two isomers of the [Fe(SH)4H]- model.
• To evaluate the performance of various quantum-mechanical (QM) methods, particularly density functional theory (DFT) and coupled-cluster (CC) approaches.
• To identify the most accurate DFT functionals for this specific chemical system.

Main Methods:

• High-level coupled-cluster (CC) calculations, including extrapolated CC series and semistochastic heat-bath configuration interaction (SHCI).
• Comparison of results from 35 different density functional theory (DFT) methods.
• Assessment of single-reference CC calculations against multireference methods.

Main Results:

• The energy difference (∆E) for the [Fe(SH)4H]- isomers was determined to be 101 kJ/mol using converged high-level QM methods.
• DFT methods showed large variations in ∆E, ranging up to 140 kJ/mol, influenced by the inclusion of exact Hartree-Fock exchange.
• M06 and B3LYP-D3 functionals provided the most accurate DFT results for this system, with errors around 4-11 kJ/mol compared to benchmark calculations.

Conclusions:

• High-level coupled-cluster and configuration interaction methods are essential for obtaining converged and accurate energy differences for challenging systems like [Fe(SH)4H]-.
• Among the tested DFT methods, M06 and B3LYP-D3 emerge as the most reliable choices for this specific system.
• Single-reference CC methods can offer reasonable accuracy for systems with moderate multi-reference character, despite potential challenges indicated by diagnostics.