This study describes an original approach to determine the structure of molten fluorinated salts at high temperatures. This approach focused on model mixtures of the type MF-AlF₃ (M=Na, K), MF-AlF₃-Al₂O₃ (M = Na, K) compositions industrial interest such as NaF-AlF₃- Al₂O₃-CaF₂ used as electrolytes in the production of aluminum by electrolysis. Thus, it is necessary to know the electroactive species present, depending on the constituents’ nature, composition, and bath temperature. We use a method based on the coupling between high-temperature nuclear magnetic resonance (HT-NMR) measurements and molecular dynamics (MD) simulations with ab-initio NMR parameters. Based on density functional theory (DFT), the latter enables the compute of the interaction parameters involved in the NMR experiments between the different species constituting the bath and evaluating the model’s quality derived from MD by comparison with experimental NMR data. A model of interaction potentials easily transferable over wide ranges of compositions capable of accounting for the microscopic structure and phenomena related to these liquids’ dynamics was developed to implement these simulations. The use of NMR measurements combined with molten mixtures’ simulations depicts an excellent agreement between the two results.

本研究描述了一种在高温下确定熔融氟化盐结构的原始方法。这种方法侧重于 MF-AlF ₃ (M=Na, K)、MF-AlF ₃ -Al ₂ O ₃ (M = Na, K) 组合物的模型混合物，例如 NaF-AlF ₃ - Al ₂ O ₃ -CaF ₂用作电解生产铝的电解质。因此，有必要根据成分的性质、组成和浴温了解存在的电活性物质。我们使用一种基于高温核磁共振 (HT-NMR) 测量和分子动力学 (MD) 模拟与 ab-initio NMR 参数之间耦合的方法。基于密度泛函理论 (DFT)，后者能够计算构成浴的不同物种之间的 NMR 实验中涉及的相互作用参数，并通过与实验 NMR 数据进行比较来评估源自 MD 的模型质量。为了实现这些模拟，开发了一种可以在很宽的成分范围内轻松转移的相互作用势模型，该模型能够解释与这些液体动力学相关的微观结构和现象。使用核磁共振测量结合熔融混合物的模拟描绘了两个结果之间的极好一致性。