Final Report Summary - TDFTBBC (Taking DFT Beyond the Boundaries of Chemistry: Solvation and Charge Transport in Biology)
The aim of this project was to develop new theory and computational algorithm aimed at a computationally fast and accurate prediction of electronic excitation energies and excited state densities of charge transfer (CT) character of large molecules and molecular aggregates.
After careful assessment of current algorithms based on time-dependent density-functional theory (TD-DFT), we realized that an accurate and reproducible method for CT excitations should depart from the TD-DFT scheme. We, therefore, opted for a more direct algorithm based on subsystem-DFT which completely by-passes the problems associated with using TD-DFT.
The developed code was implemented in the Amsterdam Density Functional computer software and has proven:
(1) high scalability wth the system size,
(2) fast computation of the needed quantities, and
(3) high accuracy.
These qualities allow our code to outplay the commonly used TD-DFT-based alternatives both in speed and in accuracy for the calculation of CT-type electronic excitations in molecular systems.
Three high-profile journal articles have been published so far on this method and a third one is expected by end of year.
After careful assessment of current algorithms based on time-dependent density-functional theory (TD-DFT), we realized that an accurate and reproducible method for CT excitations should depart from the TD-DFT scheme. We, therefore, opted for a more direct algorithm based on subsystem-DFT which completely by-passes the problems associated with using TD-DFT.
The developed code was implemented in the Amsterdam Density Functional computer software and has proven:
(1) high scalability wth the system size,
(2) fast computation of the needed quantities, and
(3) high accuracy.
These qualities allow our code to outplay the commonly used TD-DFT-based alternatives both in speed and in accuracy for the calculation of CT-type electronic excitations in molecular systems.
Three high-profile journal articles have been published so far on this method and a third one is expected by end of year.