Performance of Effective Core Potentials for Density Functional Calculations on 3d Transition Metals

Area of Science:

• Computational Chemistry
• Quantum Chemistry
• Materials Science

Background:

• Hartree-Fock-based effective core potentials (ECPs) are widely used for calculations involving transition metals.
• Assessing the accuracy of ECPs and basis sets is crucial for reliable computational chemistry results.

Purpose of the Study:

• To evaluate the performance of popular ECPs in Hartree-Fock and density functional theory (DFT) calculations for 3d transition metals.
• To compare ECP accuracy against all-electron calculations using established basis sets.

Main Methods:

• Basis-set convergence studies were performed for ten different cases involving 3d transition metals.
• Calculations included equilibrium bond dissociation energies and excitation energies.
• Results were compared against highly accurate relativistic (NCBS-DK) and nonrelativistic (NCBS-NR) all-electron calculations.

Main Results:

• All tested ECP calculations, on average, performed worse than relativistic all-electron def2-TZVP calculations.
• The best-performing ECP (CEP-RECP with MDF10) showed a significantly larger average basis-set incompleteness error (3.7 kcal/mol) compared to def2-TZVP (0.9 kcal/mol).
• ECPs with hybrid functionals performed better at achieving the basis-set limit than with local or meta-GGA functionals.

Conclusions:

• Relativistic all-electron def2-TZVP calculations are recommended for accurate DFT studies of 3d transition metals.
• While ECPs have limitations, specific combinations with hybrid functionals show improved performance.
• The choice of ECP and functional significantly impacts the accuracy of transition metal calculations.