Time-Dependent Density Functional Theory for X-ray Absorption Spectra: Comparing the Real-Time Approach to Linear...

Area of Science:

• Computational Chemistry
• Quantum Mechanics
• Spectroscopy

Background:

• Time-dependent density functional theory (TDDFT) is a powerful tool for simulating X-ray absorption spectra.
• Two formulations exist: conventional linear-response and explicitly time-dependent (real-time).
• Real-time TDDFT offers broadband calculations without core/valence separation but may introduce artifacts.

Purpose of the Study:

• To compare the accuracy and efficiency of real-time and linear-response TDDFT for simulating X-ray absorption spectra.
• To identify and address artifacts in real-time TDDFT simulations.
• To evaluate the necessity of core/valence separation in TDDFT calculations.

Main Methods:

• Simulations of X-ray absorption spectra at elemental K-edges using both real-time and linear-response TDDFT.
• Analysis of spectral contamination from continuum transitions in real-time TDDFT.
• Development and application of a filtering technique to decompose spectra.
• Comparison of computational efficiency between the two TDDFT approaches.

Main Results:

• Real-time TDDFT spectra are often contaminated by continuum transitions, especially with triple-ζ basis sets.
• These artifacts can obscure spectral interpretation due to large dipole oscillator strengths.
• A filtering technique successfully recovers meaningful spectra and separates contributions from different elements and edges.
• Linear-response TDDFT avoids these artifacts and is computationally more efficient.

Conclusions:

• While real-time TDDFT can be useful for assessing core/valence separation approximations, linear-response TDDFT is generally preferred for accurate and efficient X-ray absorption spectra simulations.
• The filtering technique is crucial for interpreting real-time TDDFT spectra.
• Basis set choice significantly impacts the severity of continuum transition artifacts in real-time TDDFT.