Mass-zero constrained dynamics for simulations based on orbital-free density functional theory.

Area of Science:

• Computational Chemistry
• Materials Science
• Theoretical Physics

Background:

• First-principles molecular dynamics (MD) simulations are crucial for understanding material properties.
• Orbital-free density functional theory (OFDFT) offers computational efficiency but faces challenges in accurate nuclear dynamics.
• Existing methods struggle to maintain time-reversibility and enforce Born-Oppenheimer approximations.

Purpose of the Study:

• To develop an efficient and time-reversible integration algorithm for OFDFT-based MD.
• To ensure accurate sampling of nuclear dynamics under the Born-Oppenheimer approximation.
• To preserve the quasilinear scaling of OFDFT with various functionals.

Main Methods:

• Introduced a new algorithm adapting Mass-Zero (MaZe) constrained dynamics for OFDFT.
• Combined standard MD integrators (Verlet, velocity Verlet) with a modified SHAKE algorithm for electronic constraints.
• Enforced full adiabatic separation between nuclear and electronic degrees of freedom.

Main Results:

• The algorithm ensures exact Born-Oppenheimer probability sampling for nuclei.
• Demonstrated preservation of OFDFT's quasilinear scaling, including nonlocal functionals.
• Validated efficiency and accuracy through static and dynamic property calculations of liquid sodium.

Conclusions:

• The developed MaZe-based algorithm provides an efficient and accurate method for OFDFT molecular dynamics.
• This approach enables reliable simulations of materials properties by correctly handling nuclear and electronic dynamics.
• The method is applicable to a wide range of OFDFT functionals, enhancing its versatility.