Room-temperature stacking disorder in layered covalent-organic frameworks from machine-learning force fields.

Area of Science:

• Materials Science
• Computational Chemistry
• Crystallography

Background:

• Layered covalent-organic frameworks (COFs) often exhibit local structures that differ from average crystal structures determined by X-ray diffraction.
• Theoretical calculations suggest that eclipsed stacking in some COFs is not energetically favorable, favoring inclined arrangements.

Purpose of the Study:

• To investigate the structural disorder and dynamic behavior of layered COFs at room temperature.
• To explore the energy landscape of stacking configurations in prototype COFs using advanced simulation techniques.

Main Methods:

• Employing molecular dynamics (MD) simulations with an on-the-fly machine learning force field (MLFF) to model COF structures at 300 K.
• Performing mesoscale MD simulations on large systems (155,000 atoms) to validate findings from smaller-scale simulations.

Main Results:

• MD simulations showed that eclipsed stacking configurations spontaneously transform into zigzag arrangements, leading to lower free energy.
• Simulated diffraction patterns derived from MLFF MD trajectories closely matched experimental observations.
• The dynamic disorder observed in MLFF MD simulations was consistent across different simulation scales.

Conclusions:

• The stacking behavior in layered COFs is more complex than previously assumed, with dynamic disorder playing a significant role.
• The zigzag configuration represents a lower free energy state compared to eclipsed stacking.
• MLFF-driven MD simulations provide a reliable method for understanding the structural intricacies of COFs.