A Benchmark Database for Spin-Flip Gap Calculations in Single- and Multireference Systems Using ΔDFT and Beyond.

Area of Science:

• Computational Chemistry
• Quantum Chemistry
• Theoretical Chemistry

Background:

• Calculating spin-flip gaps (SFGs) in molecular systems is difficult due to limited benchmark data and the complex nature of open-shell singlet states.
• Existing methods often struggle with accuracy and scalability for these challenging systems.

Purpose of the Study:

• To establish a comprehensive benchmark database for SFGs, aiding in the development and assessment of theoretical methods.
• To evaluate the performance of various density functional theory (DFT) functionals and a novel method, hierarchically correlated orbital functional theory (HCOFT), for SFG calculations.

Main Methods:

• Construction of a 419-data point SFG benchmark database, divided into single-reference (SFG-SR) and multireference (SFG-MR) subsets.
• Assessment of 32 DFT functionals using the ΔDFT approach for the SFG-SR subset.
• Evaluation of HCOFT variants, particularly 1-HCOFT, for the SFG-MR subset, focusing on accuracy and basis set dependence.

Main Results:

• Hybrid DFT functionals significantly outperform semilocal ones for SFG-SR, with Hartree-Fock exchange content being crucial for accuracy and scaling.
• Optimized ΔDFT methods provide a practical and accurate alternative to (SF-)TDDFT for SFG prediction.
• 1-HCOFT demonstrates excellent accuracy and low basis set dependence for singlet diradicals, showing promise for large multireference systems.

Conclusions:

• The developed benchmark database provides a reliable foundation for advancing theoretical methods in calculating SFGs.
• Optimized DFT functionals and 1-HCOFT offer scalable and accurate strategies for modeling spin states in diverse molecular systems.
• This work bridges the gap between benchmark data and emerging theoretical approaches for complex electronic structures.