Charge transfer excitations in TDDFT: A ghost-hunter index.

Area of Science:

• Computational Chemistry
• Quantum Chemistry
• Theoretical Chemistry

Background:

• Time-dependent density-functional theory (TD-DFT) is widely used for electronic excitation calculations.
• Ghost charge transfer (CT) states are a known artifact in TD-DFT, complicating accurate predictions.
• Existing methods for CT state identification have limitations, especially for delocalized systems.

Purpose of the Study:

• To develop a novel, computationally inexpensive index for the on-the-fly detection of ghost CT states in TD-DFT.
• To provide a robust method for identifying CT states across various molecular systems and functionals.
• To overcome the limitations of traditional Mulliken-based approaches for CT state analysis.

Main Methods:

• Introduction of the M_AC index, a modification of the Mulliken estimation for CT excitations.
• Computation of an effective CT distance using a density-based index (D_CT).
• Inclusion of orbital-weighted Ionization Potential and Electron Affinity in the index calculation.

Main Results:

• The M_AC index successfully detects ghost CT states in various model systems.
• The method is robust across different classes of TD-DFT functionals (GGA, global hybrids, range-separated hybrids).
• Accurate identification of ghost states was achieved even in intramolecular excitations with significant donor-bridge-acceptor delocalization.

Conclusions:

• The M_AC index offers an efficient and reliable tool for identifying ghost CT states in TD-DFT.
• This new index extends the applicability of CT state analysis to complex, delocalized systems.
• The M_AC index provides a valuable advancement for accurate electronic structure calculations.