Using force-matched potentials to improve the accuracy of density functional tight binding for reactive conditions

Area of Science:

• Computational Chemistry
• Materials Science
• Chemical Physics

Background:

• Accurate modeling of chemical reactivity in condensed phases is crucial for understanding material properties and behavior.
• Existing methods like Kohn-Sham density functional theory (DFT) face limitations in simulating long-term chemical processes.
• Parametrization of empirical potentials, such as in density functional tight binding (DFTB), requires accurate repulsive energy terms.

Purpose of the Study:

• To develop and validate a force matching approach for determining accurate empirical repulsive energies within the DFTB framework.
• To enhance the capability of DFTB for simulating chemical reactivity in condensed phase systems.
• To enable longer timescale simulations of chemical properties compared to conventional DFT methods.

Main Methods:

• Employed force matching to derive empirical repulsive energy parameters for the DFTB method.
• Applied the developed parametrization to simulate chemical reactivity in molten liquid carbon.
• Evaluated the method's performance for a phenolic polymer under combustion conditions.

Main Results:

• The force matching approach successfully determined accurate empirical repulsive energies for DFTB.
• Achieved improved simulation results for molten liquid carbon compared to previous DFTB parametrizations.
• Demonstrated enhanced accuracy for predicting the behavior of a phenolic polymer during combustion.

Conclusions:

• Force matching provides a robust method for accurate DFTB parametrization, particularly for repulsive energies.
• The improved DFTB method enables reliable predictions of chemical properties over extended timescales.
• This advancement offers a valuable tool for studying chemical reactivity in condensed phases, bridging the gap between empirical and ab initio methods.