Adiabatic Connection Correlation Functionals in Metallic Solids from Hartree-Fock Gaussian Basis Set Ground State

Area of Science:

• Computational chemistry
• Condensed matter physics
• Materials science

Background:

• Adiabatic Connection Integrand Interpolation (ACII) methods blend Görling-Levy second-order perturbation theory (GL2) with strong interaction density functionals.
• ACII has shown success in molecular, strongly correlated systems, and the uniform electron gas (UEG).
• The GL2 term diverges in real metallic solids, posing challenges for existing ACII methods.

Purpose of the Study:

• To adapt and test ACII methods for metallic solids, specifically addressing the divergence of the GL2 term.
• To evaluate different ACII approaches for accuracy in UEG correlation and strong interaction functionals.
• To compare calculated metallic properties with reference data and state-of-the-art Density Functional Theory (DFT) methods.

Main Methods:

• Calculations utilized Hartree-Fock (HF) ground states optimized with a novel derivative-free approach and Gaussian Type Orbitals.
• Tested various ACII strategies, including Density Parameter Interpolation (DPI) for UEG correlation.
• Employed the Point-charge-plus-Continuum (PC) model for strong interaction functionals.

Main Results:

• The Density Parameter Interpolation (DPI) ACII approach, combined with a strong interaction functional matching the Wigner crystal PC model, accurately reproduces the second-order gradient-expansion correlation coefficient.
• This optimized ACII method achieves accuracy comparable to leading DFT approaches for lattice constants and bulk correlation energies in metals.
• Combining DPI with GL2 correlation for atoms yields accurate cohesive energies, avoiding common DFT error cancellation.

Conclusions:

• The developed ACII method, specifically DPI with an accurate strong interaction functional, provides a reliable tool for calculating properties of metallic solids.
• This approach overcomes limitations of previous methods by accurately handling the GL2 term divergence in metals.
• The findings offer a pathway to more accurate and reliable computational materials science for metallic systems.