Benchmarking London dispersion corrected density functional theory for noncovalent ion-π interactions.

Area of Science:

• Computational Chemistry
• Quantum Chemistry
• Biophysical Chemistry

Background:

• Ion-π interactions are vital noncovalent forces in biological systems, influencing protein structure and molecular recognition.
• Accurate computational modeling of these interactions is essential for understanding biological processes.

Purpose of the Study:

• To benchmark various computational methods for their accuracy in describing ion-π interactions.
• To evaluate the performance of density functional approximations (DFAs) and semiempirical methods against high-level wave function theory.
• To identify reliable computational approaches for ion-π interactions in large molecular systems.

Main Methods:

• Compilation of the IONPI19 molecular test set for inter- and intramolecular ion-π interactions.
• Benchmarking 19 density functional approximations (DFAs) with different dispersion corrections.
• Comparison with wave function based methods (e.g., MP2) and semiempirical QM methods (e.g., GFNn-xTB).

Main Results:

• Dispersion-uncorrected DFT significantly underestimates ion-π interactions.
• The charge-dependent D4 dispersion model outperforms the standard D3 correction.
• Double-hybrid functionals, particularly PWPB95-D4/QZ and revDSD-PBEP86-D4/QZ, demonstrate superior reliability.

Conclusions:

• Accurate description of ion-π interactions requires appropriate treatment of dispersion.
• Double-hybrid functionals with advanced dispersion models provide the most accurate results for ion-π interactions.
• These findings enable more feasible computational studies of complex systems where coupled cluster methods are intractable.