An inversion technique for the calculation of embedding potentials.

Area of Science:

• Computational Chemistry
• Quantum Chemistry
• Materials Science

Background:

• Accurately modeling large molecular systems requires efficient methods to include electronic correlation.
• Existing density functional theory (DFT)-based embedding methods have limitations in describing local correlation effects.

Purpose of the Study:

• To develop and validate a new embedding method for incorporating local electronic correlation in large chemical systems.
• To provide an alternative to existing DFT-based embedding techniques.

Main Methods:

• The proposed method partitions the total system density into subsystem and environment components.
• An iterative embedding potential is derived using density differences, driving the calculation of localized molecular orbitals.
• This potential is incorporated into the Fock equation to capture local correlation effects within the subsystem.

Main Results:

• The method was successfully applied to hydrogen chains and their van der Waals interactions with H(2).
• Results showed excellent agreement with exact calculations performed on the entire system.
• The approach effectively includes local electronic correlation in the subsystem of interest.

Conclusions:

• The new embedding method offers a promising route for accurately treating electronic correlation in large-scale molecular simulations.
• It provides a viable alternative to current DFT-based embedding strategies, particularly for systems where local correlation is significant.