From dimers to the solid-state: Distributed intermolecular force-fields for pyridine

Area of Science:

• Computational chemistry
• Materials science
• Crystallography

Background:

• Accurate modeling of molecular crystals is crucial for predicting material properties.
• Existing force fields often struggle to capture subtle structural variations and phase transitions.

Purpose of the Study:

• To develop and validate an anisotropic atom-atom force-field for pyridine.
• To assess the force-field's performance in predicting experimental crystal structures and identifying new phases.
• To compare the anisotropic model with isotropic potentials and empirical methods.

Main Methods:

• Development of a distributed intermolecular force-field (DIFF) incorporating anisotropic atomic multipoles, polarizabilities, dispersion coefficients, and repulsion models derived from DFT dimer calculations.
• Modeling of pyridine crystal structures and comparison with experimental data.
• Crystal structure prediction studies to identify polymorphs and high-pressure phases.

Main Results:

• The DIFF model accurately reproduces experimental pyridine crystal structures, comparable to isotropic potentials fitted to experimental data.
• The DIFF model successfully predicted an unreported high-pressure phase of pyridine, which empirical potentials failed to identify.
• Differences in modeled structures were comparable to temperature, pressure, and zero-point vibrational effects.

Conclusions:

• The anisotropic DIFF model provides a more realistic representation of the pyridine pair potential energy surface for crystalline phases.
• While successful, improvements in modeling many-body terms and addressing slight over-binding are needed.
• The study highlights the complexity of crystallization modeling and the limitations of empirical potentials in capturing all relevant physical phenomena.