SiO2 nanoparticles in molten LiCl-KCl eutectic for enhanced heat storage capacity: Insight from molecular dynamics with machine learning potentials

Abstract

The Gen 3 concentrating solar power (CSP) system, featuring a high-temperature molten chloride salt reheating device, can continuously generate electricity with high efficiency. Adding nanoparticles to molten salts can improve the suspension’s thermal conductivity and heat capacity. However, nanoparticle agglomeration can impair heat transfer and storage performance, potentially leading to device failure. This study employed molecular dynamics (MD) simulations with machine learning (ML) potentials to explore the enhancement mechanism of SiO₂ nanoparticles on the thermal properties of a molten LiCl-KCl eutectic. Specific heat capacity and suspension density were monitored and compared with experiments, considering molten salt temperature, nanoparticle sizes, and loadings as control variables. Adding 13–28 Å SiO₂ at 1–6% w/w increased the specific heat capacity of the molten eutectic. Interfacial adsorption of molten salt ions onto the nanoparticles created a unique microstructure, preventing particle agglomeration and enhancing inter-particle heat transport rates. While the specific surface area of nanoparticles is a primary factor in determining thermal performance, its impact diminishes with smaller nanoparticles due to agglomeration. Detailed comparisons of ion distributions around the nanoparticles revealed that strong interfacial adsorption of Li⁺ primarily regulates the electrical field near the interface, leading to excess surface adsorption and increased heat capacity. The molecular configurations for LiCl-KCl nanofluids were detailed and presented for the first time to suggest practical ways to enhance their heat storage performance. Strategies to minimize particle agglomeration in the molten LiCl-KCl eutectic are discussed based on the study’s conclusions.