Geometry-dependent atomic multipole models for the water molecule.

Area of Science:

• Computational chemistry
• Molecular modeling
• Quantum chemistry

Background:

• Accurate representation of molecular electric potentials is crucial for understanding intermolecular interactions.
• Previous models often lacked sufficient complexity or geometric adaptability.

Purpose of the Study:

• To develop and optimize atomic electric multipole models for the water molecule.
• To reproduce ab initio computed electric potentials with high fidelity.
• To investigate the geometry dependence of these multipole models.

Main Methods:

• Ab initio calculations using coupled cluster theory with triple excitations.
• Optimization of atomic multipole models (charges, dipoles, quadrupoles).
• Fitting multipole components as a function of geometry using Taylor series expansion.

Main Results:

• Successfully optimized atomic multipole models reproducing accurate electric potentials.
• Models of increasing complexity were developed, from atomic charges to charges, dipoles, and quadrupoles.
• Geometry dependence was analyzed, with multipole components fitted using fourth-order Taylor series.

Conclusions:

• The developed atomic multipole models provide a robust and accurate description of the water molecule's electric potential.
• The geometry-dependent fitting enhances the models' applicability across various molecular configurations.
• These optimized models are valuable for simulations and studies involving water-water interactions.