Direct Link between Structure, Dynamics, and Thermodynamics in Molten Salts

Abstract

The strongly increased use of molten salts in the energy industry necessitates knowledge of their physicochemical properties on a microscopic scale to guide the development of new technology. Here, we focus on the eutectic LiCl−KCl mixture and unravel the links between structure, dynamics, and thermodynamics and investigate both mixing and temperature effects. In the mixture, for K−Cl, an elongation of the ionic bond length is accompanied by faster ion dynamics and lower Gibbs energy of activation; the opposite is true for Li−Cl. This leads to the counterintuitive result of retarded dynamics of the lighter ion, while the heavier ion diffusion is accelerated. In contrast, higher temperatures lead to a shortening of the cation− anion distances accompanied by faster dynamics despite an increase in the Gibbs energy of activation. Ionic bond breaking happens on the picosecond timescale and is suggested to proceed through an associative substitution mechanism because of a negative entropy of activation.

Microscopic origins of conductivity in molten salts unraveled by computer simulations

“…Simulations including explicit polarizability of the ions effectively model three body interactions, going beyond the pair potential approximation underlying classical force fields 28 . It has been shown that the model predicts physicochemical properties accurately over a wide range of temperatures for the gas, liquid, and solid phases, compared to both other force fields and experimental data [16][17][18]29 .…”

Systematically improved melting point prediction: a detailed physical simulation model is required

“…The force field performs well for molten salts, both pure alkali halides and mixtures, with improved results for equilibrium and dynamical properties in comparison to both experiments and other force fields. 6,16,17 In contrast to molecular mechanics, ab initio quantum chemistry calculations are in general considered to provide the most accurate approach to calculate physicochemical properties via computer simulations. Melting points have been determined using ab initio Molecular Dynamics (MD), however this is computationally demanding, in particular for strongly ionic substances, 18 and the accuracy is limited by sampling and system size convergence.…”

“…In our latest contribution we showed that our polarisable force field is indeed able to accurately predict physicochemical properties, such as the conductivity, of such a mixture. 17 As an alternative to using polarisable models, Leontyev and Stuchebrukhov [22][23][24][25] introduced ions with effective charges that are scaled down in order to mimic the electronic dielectric screening in condensed matter. Such models work in conjunction with other effective-charge empirical models such as popular water models.…”

“…30 Using our new force field we were able contribute to this issue, by directly connecting structural, dynamical and thermodynamical properties in molten salts. 16,17 Two main definitions have been used to investigate melting: 4 (i) the melting point is considered to be the point where the Gibbs free energy of the solid equals that of the melt; (ii) there are characteristic changes in the solid as T m is approached, e.g. the formation of a critical concentration of Schottky defects.…”

“…A polarisable model comes at extra computational costs, however, especially when investigating properties that rely on an accurate description of dynamic properties the inclusion of polarisability is the road ahead. 16,17 The mechanisms behind melting and phase-transitions at large are still not completely understood but we are confident that with improved accurate force fields in general it will be possible to gain further insights in these processes. Such theoretical methods can contribute to the prediction of phase-transition processes that are challenging to capture experimentally.…”