Benchmarking isotropic hyperfine coupling constants using (QTP) DFT functionals and coupled cluster theory

Area of Science:

• Computational Chemistry
• Quantum Chemistry
• Theoretical Chemistry

Background:

• Accurate prediction of isotropic hyperfine coupling constants is crucial for fitting experimental model Hamiltonians.
• Previous studies have focused on wavefunction and Density Functional Theory (DFT) approaches, but few have examined DFT functionals satisfying the Bartlett ionization potential (IP) condition.
• There is a need to evaluate DFT functionals for their predictive power on hyperfine coupling constants across various radical types.

Purpose of the Study:

• To benchmark the performance of 24 commonly used DFT functionals for predicting isotropic hyperfine coupling constants.
• To specifically assess the Quantum Theory Project (QTP) functionals (CAM-QTP00, CAM-QTP01, CAM-QTP02, QTP17) for this property.
• To compare DFT predictions with coupled-cluster singles and doubles (CCSD) and CCSD(T) results for organic radicals.

Main Methods:

• Prediction of isotropic hyperfine coupling constants for 56 radicals using 24 DFT functionals.
• Inclusion of small and large organic radicals, as well as transition metal complexes.
• Validation using coupled-cluster singles and doubles (CCSD) and CCSD with perturbative triples [CCSD(T)] calculations for select organic radicals.

Main Results:

• The QTP17 and CAM-QTP00 functionals underperformed, despite being parameterized for an IP eigenvalue condition.
• The CAM-QTP01 functional demonstrated the highest accuracy for both organic radical datasets.
• CAM-QTP01 and CAM-QTP02 were the most accurate functionals for the transition metal complex dataset.
• Hybrid DFT functionals generally provided an optimal balance between accuracy and precision across all datasets.

Conclusions:

• CAM-QTP01 emerges as a highly accurate functional for predicting isotropic hyperfine coupling constants in organic radicals.
• CAM-QTP01 and CAM-QTP02 are recommended for studies involving transition metal complexes.
• Hybrid DFT functionals offer a reliable choice for general applications requiring a balance of accuracy and computational efficiency.