Comparative Studies of the Structural and Transport Properties of Molten Salt FLiNaK Using the Machine-Learned Neural Network and Reparametrized Classical Forcefields

First-Principles Investigation of the Effects of UF4 and ThF4 Fuels on the Structural, Dynamic, and Thermodynamic Properties of LiF-NaF

An in-depth understanding and characterization of molten salt properties are necessary for the optimized design, efficient operation, and safety assurance of molten salt reactors (MSRs). Investigating molten salt properties in experimental settings can be challenging and time-consuming due to the high temperatures of interest, the salt's corrosiveness, purity and composition control, and health and safety concerns. Therefore, it is beneficial to perform computational screening to assist in the ultimate experimental measurements. Herein, we used first-principles molecular dynamics simulations to calculate several thermophysical, structural, and dynamic properties of eutectic LiF-NaF with fuel additives UF4 and ThF4. We found that with the incorporation of uranium or thorium, a prepeak appears in the structure factor, indicative of a medium-range structural ordering. Furthermore, we explore the mechanism through which these structural changes enhance shear stress correlations, thereby increasing the salt's viscosity. This work highlights the importance of studying the atomic-scale structure of molten salts and how the addition of fuel elements can substantially affect it.

Improving the reliability of machine learned potentials for modeling inhomogeneous liquids

The atomic-scale response of inhomogeneous fluids at interfaces and surrounding solute particles plays a critical role in governing chemical, electrochemical, and biological processes. Classical molecular dynamics simulations have been applied extensively to simulate the response of fluids to inhomogeneities directly, but are limited by the accuracy of the underlying interatomic potentials. Here, we use neural network potentials (NNPs) trained to ab initio simulations to accurately predict the inhomogeneous responses of two distinct fluids: liquid water and molten NaCl. Although NNPs can be readily trained to model complex bulk systems across a range of state points, we show that to appropriately model a fluid's response at an interface, relevant inhomogeneous configurations must be included in the training data. In order to sufficiently sample appropriate configurations of such inhomogeneous fluids, we develop protocols based on molecular dynamics simulations in the presence of external potentials. We demonstrate that NNPs trained on inhomogeneous fluid configurations can more accurately predict several key properties of fluids-including the density response, surface tension and size-dependent cavitation free energies-for liquid water and molten NaCl, compared to both empirical interatomic potentials and NNPs that are not trained on such inhomogeneous configurations. This work therefore provides a first demonstration and framework to extract the response of inhomogeneous fluids from first principles for classical density-functional treatment of fluids free from empirical potentials.

First-principles molten salt phase diagrams through thermodynamic integration

Precise prediction of phase diagrams in molecular dynamics simulations is challenging due to the simultaneous need for long time and large length scales and accurate interatomic potentials. We show that thermodynamic integration from low-cost force fields to neural network potentials trained using density-functional theory (DFT) enables rapid first-principles prediction of the solid-liquid phase boundary in the model salt NaCl. We use this technique to compare the accuracy of several DFT exchange-correlation functionals for predicting the NaCl phase boundary and find that the inclusion of dispersion interactions is critical to obtain good agreement with experiment. Importantly, our approach introduces a method to predict solid-liquid phase boundaries for any material at an ab initio level of accuracy, with the majority of the computational cost at the level of classical potentials.

Modeling Chemical Reactions in Alkali Carbonate-Hydroxide Electrolytes with Deep Learning Potentials

We developed a deep potential machine learning model for simulations of chemical reactions in molten alkali carbonate-hydroxide electrolyte containing dissolved CO₂, using an active learning procedure. We tested the deep neural network (DNN) potential and training procedure against reaction kinetics, chemical composition, and diffusion coefficients obtained from density functional theory (DFT) molecular dynamics calculations. The DNN potential was found to match DFT results for the structural, transport, and short-time chemical reactions in the melt. Using the DNN potential, we extended the time scales of observation to 2 ns in systems containing thousands of atoms, while preserving quantum chemical accuracy. This allowed us to reach chemical equilibrium with respect to several chemical species in the melt. The approach can be generalized for a broad spectrum of chemically reactive systems.

First-principles molecular dynamics simulations of UCln-MgCl2 (n = 3, 4) molten salts

Molten chlorides are a preferred choice for fast-spectrum molten salt reactors. Molten MgCl₂ forms eutectic mixtures with NaCl and is considered as a promising dilutant to dissolve fuel salts such as UCl₃ and UCl₄. However, the structure and chemical properties of UCl_n (n = 3, 4) in molten MgCl₂ are not well understood. Here we use first-principles molecular dynamics to investigate the molten salt system UCl_n-MgCl₂ (n = 3, 4) at various concentrations of U³⁺ and U⁴⁺. It is found that the coordination environment of Cl^- around U³⁺, especially in the first coordination shell, varies only slightly with the uranium concentration and that both the 7-fold coordinate (UCl₇^4-) and 6-fold coordinate (UCl₆^3-) structures dominate at ∼40%, leading to an average coordination number of 6.6-6.7. A network or polymeric structure of U³⁺ cations sharing Cl^- ions is extensively formed when the mole fraction of UCl₃ is greater than 0.2. In contrast, the average coordination number of Cl^- around U⁴⁺ is about 6.4 for a mole fraction of UCl₄, x(UCl₄), of 0.1 but decreases to 6.0 for x(UCl₄) = 0.2 and then stays at about 6.0-6.2 with the uranium concentration. The 6-fold coordinate structure (UCl₆^2-) is the most populous in UCl₄-MgCl₂, at about 60%. U-Cl network formation becomes dominant (>50%) only when x(UCl₄) > 0.5. Unlike Na⁺, Mg²⁺ forms a network structure with Cl^- ions and when x(UCl₃) or x(UCl₄) < 0.5, over 90% of Mg²⁺ ions are part of a network structure, implying the complex influences from Mg²⁺ on the coordination of Cl around U. The present work reveals the impact of MgCl₂ as a solvent for UCl_n (n = 3, 4) on the U-Cl coordination and structure, and motivates further studies of their transport properties and the tertiary systems containing MgCl₂-UCl_n.

Computational methods to simulate molten salt thermophysical properties

Molten salts are important thermal conductors used in molten salt reactors and solar applications. To use molten salts safely, accurate knowledge of their thermophysical properties is necessary. However, it is experimentally challenging to measure these properties and a comprehensive evaluation of the full chemical space is unfeasible. Computational methods provide an alternative route to access these properties. Here, we summarize the developments in methods over the last 70 years and cluster them into three relevant eras. We review the main advances and limitations of each era and conclude with an optimistic perspective for the next decade, which will likely be dominated by emerging machine learning techniques. This article is aimed to help researchers in peripheral scientific domains understand the current challenges of molten salt simulation and identify opportunities to contribute.

Adding MgCl2 to Molten NaCl-UCl n (n=3, 4): Insights from First Principles Molecular Dynamics

Molten chlorides are proposed for fast-spectrum molten salt reactors. Molten MgCl 2 with NaCl forms eutectic mixtures and is considered as a promising dilutant to dissolve fuel salts such as UCl 3 and UCl 4 . Previous study suggests the formation of U-Cl network at U:Na=1:1 binary salt. However, it is unclear how the structure of UCl n (n = 3, 4) in NaCl will change after adding MgCl 2 in the salt. Here we use first-principles molecular dynamics to investigate the molten ternary salts NaCl-MgCl 2 -UCl n (n=3, 4) at various concentrations of Mg 2+ in NaCl-UCl n with a fixed ratio of Na:U at 1:1. It is found that the addition of Mg 2+ in NaCl-UCl 3 leads to higher coordination number (from 6.5 to 6.7) of Cl around U while the U-Cl network structure slightly decreases with the Mg concentration. Adding MgCl 2 to NaCl-UCl 4 , however, breaks down the U-Cl network more completely. We attribute the different behavior of adding Mg 2+ into NaCl-UCl 3 and NaCl-UCl 4 to the difference between U(III) and U(IV) in attracting Cl - ions to form the first coordination shell. The present work reveals the impact of MgCl 2 as a dilutant solvent to the NaCl-UCl n fuel salts, which will be helpful in further studies and understanding of the thermophysical and transport properties of the ternary systems.