A new nonempirical tuning scheme with single self-consistent field calculation: Comparison with global and IP-tuned...

Area of Science:

• Computational Chemistry
• Quantum Chemistry
• Theoretical Chemistry

Background:

• Range-separated functionals are crucial for minimizing delocalization errors in electronic structure calculations.
• Current nonempirical tuning methods for these functionals are computationally expensive, requiring multiple ΔSCF calculations.
• Accurate prediction of highest occupied molecular orbital (HOMO) energies and HOMO-LUMO gaps is essential for understanding chemical reactivity.

Purpose of the Study:

• To develop a novel, computationally efficient scheme for nonempirical tuning of range-separated functionals.
• To determine the optimal tuning parameter (μ) using a single self-consistent field (SCF) calculation.
• To validate the proposed method's accuracy in reproducing HOMO energies and HOMO-LUMO gaps.

Main Methods:

• A new tuning scheme based on the evaluation of the spherically symmetric average Electron Localization Function (ELF) region was developed.
• The radius of the average ELF region serves as a measure of the functional's long-range behavior dominance.
• The method determines the optimal μ value through a single SCF calculation, avoiding iterative ΔSCF procedures.

Main Results:

• The proposed ELF-tuned range-separated functional significantly improves the reproduction of HOMO energies and HOMO-LUMO gaps.
• Maximum error in HOMO energy prediction is considerably smaller compared to global and IP-tuned functionals.
• The deviation of HOMO energies from ΔSCF IP calculations is substantially reduced by the ELF-tuned functional.

Conclusions:

• The developed ELF-based tuning scheme offers a more efficient and accurate alternative for range-separated functionals.
• This method provides a reliable way to minimize delocalization errors with reduced computational cost.
• The improved accuracy in predicting key electronic properties makes this approach valuable for various chemical applications.