Non-perturbative calculation of molecular magnetic properties within current-density functional theory.

Area of Science:

• Computational Chemistry
• Quantum Chemistry
• Density-Functional Theory

Background:

• Accurate calculation of molecular magnetic properties is crucial for understanding chemical phenomena.
• Standard density-functional theory (DFT) methods often struggle with strong external magnetic fields.
• Current-density functionals (CDFs) offer a potential improvement but require robust implementations.

Purpose of the Study:

• To present a novel, fully self-consistent implementation of Kohn-Sham DFT using London atomic orbitals for molecules.
• To enable non-perturbative treatment of external magnetic fields for studying magnetic response properties and strong-field effects.
• To investigate the impact of current-density functionals on the accuracy of magnetic property calculations.

Main Methods:

• Implementation of Kohn-Sham DFT with London atomic orbitals.
• Non-perturbative treatment of external magnetic fields.
• Application of both standard DFT and current-density functionals (CDFs).
• Finite-field calculations of magnetizabilities, hypermagnetizabilities, and NMR shielding constants.

Main Results:

• The implementation is the first fully self-consistent CDF approach for molecules.
• Existing CDFs show sensitivity to numerical implementation details and offer no improvement over standard DFT when regularized.
• The method successfully handles very strong magnetic fields where perturbative approaches fail.
• Inadequacies of CDF approximations worsen with increasing magnetic field strength, while standard DFT remains stable but less accurate.

Conclusions:

• The developed implementation provides a robust tool for studying molecules in strong magnetic fields.
• Current CDFs require improvement for accurate prediction of magnetic properties, especially under strong fields.
• Further development of current-dependent density-functionals is necessary and can be tested with this new implementation.