Dissociation Energies via Embedding Techniques.

Area of Science:

• Computational Chemistry
• Quantum Mechanics
• Materials Science

Background:

• Calculating interaction energies for large systems like liquids and crystals is computationally expensive.
• Fragmentation into smaller subsystems (many-body expansion) is a promising approach to reduce computational cost.
• Accurate and efficient methods are needed to determine interaction energies and gradients for complex systems.

Purpose of the Study:

• To test and evaluate various subsystem (embedding) approaches for calculating interaction energies.
• To explore the limits and behaviors of different embedding methods.
• To identify favorable methods for accurately describing systems with varying interaction strengths.

Main Methods:

• Mechanical embedding, point charges, polarizable embedding, polarizable density embedding, and density embedding were investigated.
• Nonembedded fragmentation, quantum mechanics/molecular mechanics (QM/MM), and quantum mechanics/quantum mechanics (QM/QM) embedding theories were evaluated.
• Symmetry-adapted perturbation theory with density functional theory was used for interpretation.

Main Results:

• Different embedding methods and schemes show varying degrees of accuracy depending on interaction strength.
• Embedding approaches reduced interaction energy errors by up to a factor of 20 compared to unembedded methods.
• Errors in interaction energies were reduced to below 0.1 kJ/mol with the presented embedding approaches.

Conclusions:

• Subsystem and embedding methods offer a computationally efficient alternative for calculating interaction energies in large systems.
• The choice of embedding method is crucial and depends on the specific system and interaction characteristics.
• These advanced embedding techniques significantly improve accuracy while reducing computational burden.