A fully quantum-mechanical treatment for kaolinite

Area of Science:

• Materials Science
• Computational Chemistry
• Mineral Physics

Background:

• Developing accurate computational models for minerals like kaolinite is crucial for understanding their behavior.
• Density functional theory (DFT) calculations provide a foundation for creating these models.

Purpose of the Study:

• To develop and evaluate neural network potentials for kaolinite using different DFT functionals.
• To investigate the impact of nuclear quantum effects (NQEs) on mineral properties.

Main Methods:

• Fitting neural network potentials to DFT data from revPBE + D3 and revPBE + vdW functionals.
• Calculating static and dynamic properties of kaolinite using these potentials.
• Comparing potential accuracy against experimental infrared (IR) spectra.
• Assessing the influence of NQEs via a fully quantum nuclear treatment.

Main Results:

• The revPBE + vdW functional better reproduced static properties.
• The revPBE + D3 functional showed superior agreement with experimental IR spectra.
• NQEs had a negligible effect on static properties.
• NQEs significantly altered the dynamic properties of kaolinite.

Conclusions:

• Neural network potentials offer a viable route for modeling kaolinite.
• The choice of DFT functional impacts the accuracy of static versus dynamic properties.
• NQEs are essential for accurately describing the dynamic behavior of kaolinite.