Transferability of interatomic potentials with insights into the structure–property relationship of SiO₂–CaO–MgO–Al₂O₃ melts

ABSTRACT

Molecular dynamics (MD) simulations were carried out to validate the transferability of Matsui, Kawamura, Miyake, and Guillot potentials on the composition–structure–property relationship of liquid SiO₂–CaO–MgO–Al₂O₃ (CMAS) melts in a wide range of chemical compositions at 1773K. Results revealed that Matsui, Miyake, and Guillot potentials were able to reproduce experimentally observed local atomic structures and properties while Kawamura potential cannot. Even though an accurate trend can be obtained, the values of properties obtained with Miyake, Matsui and Guillot potentials varied obviously with each other. Viscosity obtained using Einstein–Stokes method is about one magnitude lower than experimental viscosity, while viscosities obtained with Green–Kubo and reverse nonequilibrium molecular dynamics methods underestimate the viscosity by about two magnitudes. The increase in CaO/SiO₂ ratio under fixed MgO/Al₂O₃ ratio and the increase in MgO/Al₂O₃ ratio under fixed CaO/SiO₂ ratio can both decrease the polymerisation degree of system, which finally decreases the viscosity. The influence of adding a Morse term to BMH potential is very slightly. This paper provides the foundation to further simulate the atomic structure of CMAS system and further optimise the Born–Mayer–Higgins potential for various oxide systems.

KEYWORDS:

• Aluminosilicates melts
• molecular dynamics
• interatomic potentials
• structure–property relationship

Acknowledgements

Computations were performed on the Niagara supercomputer at the SciNet HPC Consortium in the Compute/Calcul Canada national computing platform. SciNet is funded by the Canada Foundation for Innovation under the auspices of Compute Canada, the Government of Ontario, Ontario Research Fund – Research Excellence, and the University of Toronto. The authors acknowledge the technical support and suggestions of Prof Mansoor Barati of the University of Toronto.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

The authors acknowledge the financial support of the National Science Foundation of China for Young Scientists [grant number 51804025], the National Science Foundation of China [grant numbers 51974019 and 51774032], the Chinese Fundamental Research Funds for the Central Universities (FRF-TP-17-086A1), and the National Key Research and Development Program of China [grant numbers 2017YFB0304300 and 2017YFB0304303].