All-electron real-time and imaginary-time time-dependent density functional theory within a numeric atom-centered...

Area of Science:

• Computational Chemistry
• Quantum Dynamics
• Materials Science

Background:

• Real-time time-dependent density functional theory (RT-TDDFT) models quantum dynamics without linear response approximation.
• Imaginary-time time-dependent density functional theory (it-TDDFT) offers robust convergence for ground-state calculations.
• All-electron numerical atom-centered orbital (NAO) basis sets provide high precision in electronic structure calculations.

Purpose of the Study:

• To present high-precision all-electron RT-TDDFT and it-TDDFT implementations within the FHI-aims code using NAO basis functions.
• To validate RT-TDDFT against linear-response TDDFT and analyze basis set convergence.
• To demonstrate the capabilities of these methods for periodic systems, core-level spectra, and large-scale simulations.

Main Methods:

• Implementation of RT-TDDFT and it-TDDFT using an all-electron NAO basis set framework.
• Validation of RT-TDDFT against linear-response TDDFT for small molecules.
• Application of velocity-gauge formalism for periodic boundary conditions and core-level spectra analysis.
• Benchmarking of computational scaling for systems up to ~500 atoms.

Main Results:

• RT-TDDFT results show good agreement with linear-response TDDFT, highlighting the importance of augmentation basis functions for convergence.
• The all-electron NAO formalism enables successful convergence of challenging systems with it-TDDFT.
• The implementation demonstrates excellent performance for core-level spectra and periodic systems.
• The computational approach exhibits nearly linear weak and strong scaling behavior for large systems.

Conclusions:

• The presented all-electron RT-TDDFT and it-TDDFT implementations in FHI-aims offer a robust and scalable approach for quantum dynamics and ground-state calculations.
• The methods are validated and demonstrate high accuracy and efficiency, particularly for complex systems and large-scale simulations.
• This work provides a valuable tool for advancing computational studies in chemistry and materials science.