Nonlocal kinetic energy functionals by functional integration.

Area of Science:

• Computational Materials Science
• Quantum Chemistry
• Condensed Matter Physics

Background:

• Density-Functional Theory (DFT) seeks accurate electron density functionals.
• Approximating the noninteracting kinetic energy functional (T_s[ρ]) is a significant challenge.
• Current methods often derive potentials from the first functional derivative, limiting accuracy.

Purpose of the Study:

• To develop a new nonlocal functional for T_s[ρ] that overcomes limitations of existing methods.
• To construct the kinetic energy potential via functional integration of the second functional derivative.
• To introduce the Mi-Genova-Pavanello (MGP) functional with a density-independent kernel.

Main Methods:

• Developed a new nonlocal functional (MGP) for T_s[ρ].
• Constructed the potential from the second-functional derivative via functional integration.
• MGP satisfies three exact conditions: nonzero kinetic electron, reduction to inverse Lindhard function, and potential derived from second derivative integration.

Main Results:

• MGP accurately reproduces equilibrium volumes, bulk moduli, total energy, and electron densities for silicon and III-V semiconductors.
• The functional demonstrates numerical stability, converging within 12 iterations.
• MGP exhibits low computational cost and memory requirements, comparable to existing functionals.

Conclusions:

• The MGP functional provides an accurate and efficient approach for calculating the noninteracting kinetic energy in DFT.
• This method challenges the traditional paradigm by deriving potentials from the second functional derivative.
• MGP shows promise for orbital-free and subsystem DFT simulations, particularly for metallic and semiconducting materials.