Subsystem-DFT potential-energy curves for weakly interacting systems.

Area of Science:

• Computational chemistry
• Quantum chemistry
• Materials science

Background:

• Standard Kohn-Sham density-functional theory (DFT) approximations like LDA and GGA struggle to accurately describe London dispersion interactions, often failing to predict bound potential-energy surfaces for van der Waals complexes.
• Subsystem-based DFT has shown potential for improved accuracy in interaction energies for weakly bound systems compared to standard DFT without dispersion corrections, possibly due to error cancellation.

Purpose of the Study:

• To thoroughly investigate the accuracy of subsystem DFT for weakly interacting systems, with a specific focus on the shape of potential-energy surfaces (PESs).
• To evaluate the performance of different functionals within subsystem DFT for describing van der Waals interactions.

Main Methods:

• Utilized extensive benchmark data sets, including S22x5 and S66x8, for evaluating subsystem DFT performance.
• Analyzed potential-energy surfaces (PESs) generated by subsystem DFT using various functionals, including LDA, PW91, and BP86 GGAs.

Main Results:

• Subsystem DFT potential-energy surfaces (PESs) exhibit significant variation depending on the chosen functional.
• Local-density approximation (LDA) results in subsystem DFT are generally good but differ from Kohn-Sham (KS) counterparts.
• Perdew-Wang (PW91) generalized gradient approximations (GGAs) in subsystem DFT often lead to overbinding, while Becke-Perdew (BP86) GGAs show similar issues to standard KS-DFT.

Conclusions:

• Subsystem DFT offers a viable alternative for studying van der Waals interactions, but functional choice is critical.
• Empirical corrections are being explored to mitigate overbinding issues observed with PW91 and BP86 functionals in subsystem DFT.