An Efficient Reactive Force Field without Explicit Coordination Dependence for Studying Caustic Aluminum Chemistry

Area of Science:

• Computational chemistry
• Materials science
• Chemical engineering

Background:

• Reactive force fields (RFFs) offer efficient sampling of chemical reaction pathways in complex systems compared to density functional theory.
• Improving RFF efficiency is crucial for systems with slow dynamics, such as viscous and superconcentrated solutions.

Purpose of the Study:

• To present a novel RFF that enhances computational efficiency.
• To model aluminate reactivity in sodium hydroxide solutions, focusing on aluminum hydroxide nucleation.

Main Methods:

• Developed a new RFF by omitting explicit atom coordination dependence and using a small parameter set (Lennard-Jones, Gaussian, Stillinger-Weber potentials).
• Parametrized the RFF using AIMD simulation data.
• Validated the model against experimental radial distribution functions, free energy profiles for oligomerization, and formation energies.

Main Results:

• The new RFF successfully models aluminate reactivity and early-stage Al(OH)3 nucleation.
• Achieved a computational cost reduction of one order of magnitude compared to ReaxFF.
• Demonstrated validation against experimental data and computed thermodynamic properties.

Conclusions:

• The developed RFF provides a more efficient approach for simulating chemical reactions in complex, viscous systems.
• This method is highly relevant for industrial aluminum processing, particularly in understanding nucleation phenomena.