A deep potential molecular dynamics study on the ionic structure and transport properties of NaCl-CaCl2 molten salt

• Published in: Ionics 2024, Vol.30 (1), p.285-295
• Main Authors: Gegentana, Cui, Liu, Zhou, Leping, Du, Xiaoze
• Format: Article
• Language: English
• Subjects: Calcium chloride; Chemistry; Chemistry and Materials Science; Chloride ions; Condensed Matter Physics; Coordination numbers; Density functional theory; Diffusion coefficient; Dynamic structural analysis; Electrochemistry; Energy Storage; High temperature; Machine learning; Molecular dynamics; Molecular structure; Molten salts; Optical and Electronic Materials; Performance prediction; Renewable and Green Energy; Self diffusion; Simulation; Storage systems; Temperature; Temperature dependence; Thermal stability; Thermal storage; Transport properties; Viscosity
• ISSN: 0947-7047; 1862-0760
• DOI: 10.1007/s11581-023-05265-8

Chloride salts that have abundant reserves own good thermal stability in high-temperature components of solar systems and thermal storage devices. However, the high corrosiveness makes their experimental study difficult, especially at high temperatures. In this work, the ionic structure and transport properties of the binary chloride NaCl-CaCl 2 system are investigated by applying density functional theory, machine learning, and molecular dynamics simulation methods. The maximum errors of energy and force for deep potential molecular dynamics are 5.02 × 10 −4 eV/atom and 0.02 eV·Å –1 , respectively. The ionic structure of the NaCl-CaCl 2 system is mainly determined by the coordination requirements of calcium and chloride ions. As the temperature decreases, the coordination number increases, and the network structure of ions becomes more compact, resulting in a positive temperature dependence of the self-diffusion coefficient of ions in the system and a negative temperature dependence of the viscosity. The simulation results are in good agreement with the experimental data. The deep potential molecular dynamics simulation method can effectively solve the problem of finite scale effects faced by the ab initio molecular dynamics (AIMD) simulation method in predicting transport properties, while achieving the computational accuracy of AIMD and thus can better predict transport properties such as self-diffusion coefficient and viscosity. This is of great importance in accurately predicting the thermal performance of molten salts and their enhancement for solar and thermal storage systems.