GGA-Level Subsystem DFT Achieves Sub-kcal/mol Accuracy Intermolecular Interactions by Mimicking Nonlocal Functionals.

Area of Science:

• Computational Chemistry
• Quantum Chemistry
• Density Functional Theory (DFT)

Background:

• Nonlocal kinetic energy functionals are essential for subsystem DFT simulations, particularly for approximating nonadditive kinetic energy.
• These functionals should correctly reduce to Thomas-Fermi and von Weizsäcker functionals in specific density regimes.
• Existing nonlocal functionals can be computationally expensive, limiting their widespread application.

Purpose of the Study:

• To propose a new generalized gradient approximation (GGA) nonadditive kinetic energy functional.
• To develop a functional that retains the computational efficiency of semilocal functionals while mimicking nonlocal behavior.
• To ensure the functional accurately accounts for inter-subsystem density overlap.

Main Methods:

• Development of a novel GGA functional for nonadditive kinetic energy.
• The functional's dependence on inter-subsystem density overlap was a key design feature.
• Validation against Kohn-Sham DFT and coupled cluster benchmark calculations (CCSD(T)) for weakly interacting dimers.

Main Results:

• The proposed GGA functional successfully reproduces the desired behavior of nonlocal functionals.
• It demonstrates high accuracy in predicting interaction energies for dimers in the S22-5 and S66 test sets.
• The mean absolute deviation was significantly below 1 kcal/mol, indicating excellent performance.

Conclusions:

• The new GGA functional provides an accurate and computationally efficient method for subsystem DFT.
• It effectively captures the nonadditive kinetic energy contributions in weakly interacting systems.
• This development offers a practical advancement for electronic structure calculations in complex molecular systems.